From 3fd81c73cfd8bad36b2a1cf7955006e35c1d9db5 Mon Sep 17 00:00:00 2001 From: Camil Staps Date: Fri, 23 Oct 2015 16:44:24 +0200 Subject: Assignment 3: code, plots --- Assignment 3/packages/pydot.py | 2037 ++++++++++++++++ Assignment 3/packages/treeViewer.py | 17 + Assignment 3/packages/xlrd/__init__.py | 467 ++++ Assignment 3/packages/xlrd/biffh.py | 663 ++++++ Assignment 3/packages/xlrd/book.py | 1420 ++++++++++++ Assignment 3/packages/xlrd/compdoc.py | 473 ++++ Assignment 3/packages/xlrd/doc/compdoc.html | 69 + Assignment 3/packages/xlrd/doc/xlrd.html | 2064 +++++++++++++++++ Assignment 3/packages/xlrd/examples/namesdemo.xls | Bin 0 -> 22528 bytes .../packages/xlrd/examples/xlrdnameAPIdemo.py | 179 ++ Assignment 3/packages/xlrd/formatting.py | 1262 ++++++++++ Assignment 3/packages/xlrd/formula.py | 2179 ++++++++++++++++++ Assignment 3/packages/xlrd/info.py | 1 + Assignment 3/packages/xlrd/licences.py | 77 + Assignment 3/packages/xlrd/sheet.py | 2424 ++++++++++++++++++++ Assignment 3/packages/xlrd/timemachine.py | 52 + Assignment 3/packages/xlrd/xldate.py | 213 ++ Assignment 3/packages/xlrd/xlsx.py | 804 +++++++ 18 files changed, 14401 insertions(+) create mode 100755 Assignment 3/packages/pydot.py create mode 100755 Assignment 3/packages/treeViewer.py create mode 100644 Assignment 3/packages/xlrd/__init__.py create mode 100644 Assignment 3/packages/xlrd/biffh.py create mode 100644 Assignment 3/packages/xlrd/book.py create mode 100644 Assignment 3/packages/xlrd/compdoc.py create mode 100644 Assignment 3/packages/xlrd/doc/compdoc.html create mode 100644 Assignment 3/packages/xlrd/doc/xlrd.html create mode 100644 Assignment 3/packages/xlrd/examples/namesdemo.xls create mode 100644 Assignment 3/packages/xlrd/examples/xlrdnameAPIdemo.py create mode 100644 Assignment 3/packages/xlrd/formatting.py create mode 100644 Assignment 3/packages/xlrd/formula.py create mode 100644 Assignment 3/packages/xlrd/info.py create mode 100644 Assignment 3/packages/xlrd/licences.py create mode 100644 Assignment 3/packages/xlrd/sheet.py create mode 100644 Assignment 3/packages/xlrd/timemachine.py create mode 100644 Assignment 3/packages/xlrd/xldate.py create mode 100644 Assignment 3/packages/xlrd/xlsx.py (limited to 'Assignment 3/packages') diff --git a/Assignment 3/packages/pydot.py b/Assignment 3/packages/pydot.py new file mode 100755 index 0000000..104c817 --- /dev/null +++ b/Assignment 3/packages/pydot.py @@ -0,0 +1,2037 @@ +# -*- coding: utf-8 -*- +"""Graphviz's dot language Python interface. + +This module provides with a full interface to create handle modify +and process graphs in Graphviz's dot language. + +References: + +pydot Homepage: http://code.google.com/p/pydot/ +Graphviz: http://www.graphviz.org/ +DOT Language: http://www.graphviz.org/doc/info/lang.html + +Programmed and tested with Graphviz 2.26.3 and Python 2.6 on OSX 10.6.4 + +Copyright (c) 2005-2011 Ero Carrera + +Distributed under MIT license [http://opensource.org/licenses/mit-license.html]. +""" + +__revision__ = "$LastChangedRevision$" +__author__ = 'Ero Carrera' +__version__ = '1.0.%d' % 29 # int( __revision__[21:-2] ) +__license__ = 'MIT' + +import os +import re +import subprocess +import tempfile +import copy +try: + import dot_parser +except Exception, e: + print "Couldn't import dot_parser, loading of dot files will not be possible." + + + +GRAPH_ATTRIBUTES = set( ['Damping', 'K', 'URL', 'aspect', 'bb', 'bgcolor', + 'center', 'charset', 'clusterrank', 'colorscheme', 'comment', 'compound', + 'concentrate', 'defaultdist', 'dim', 'dimen', 'diredgeconstraints', + 'dpi', 'epsilon', 'esep', 'fontcolor', 'fontname', 'fontnames', + 'fontpath', 'fontsize', 'id', 'label', 'labeljust', 'labelloc', + 'landscape', 'layers', 'layersep', 'layout', 'levels', 'levelsgap', + 'lheight', 'lp', 'lwidth', 'margin', 'maxiter', 'mclimit', 'mindist', + 'mode', 'model', 'mosek', 'nodesep', 'nojustify', 'normalize', 'nslimit', + 'nslimit1', 'ordering', 'orientation', 'outputorder', 'overlap', + 'overlap_scaling', 'pack', 'packmode', 'pad', 'page', 'pagedir', + 'quadtree', 'quantum', 'rankdir', 'ranksep', 'ratio', 'remincross', + 'repulsiveforce', 'resolution', 'root', 'rotate', 'searchsize', 'sep', + 'showboxes', 'size', 'smoothing', 'sortv', 'splines', 'start', + 'stylesheet', 'target', 'truecolor', 'viewport', 'voro_margin', + # for subgraphs + 'rank' ] ) + + +EDGE_ATTRIBUTES = set( ['URL', 'arrowhead', 'arrowsize', 'arrowtail', + 'color', 'colorscheme', 'comment', 'constraint', 'decorate', 'dir', + 'edgeURL', 'edgehref', 'edgetarget', 'edgetooltip', 'fontcolor', + 'fontname', 'fontsize', 'headURL', 'headclip', 'headhref', 'headlabel', + 'headport', 'headtarget', 'headtooltip', 'href', 'id', 'label', + 'labelURL', 'labelangle', 'labeldistance', 'labelfloat', 'labelfontcolor', + 'labelfontname', 'labelfontsize', 'labelhref', 'labeltarget', + 'labeltooltip', 'layer', 'len', 'lhead', 'lp', 'ltail', 'minlen', + 'nojustify', 'penwidth', 'pos', 'samehead', 'sametail', 'showboxes', + 'style', 'tailURL', 'tailclip', 'tailhref', 'taillabel', 'tailport', + 'tailtarget', 'tailtooltip', 'target', 'tooltip', 'weight', + 'rank' ] ) + + +NODE_ATTRIBUTES = set( ['URL', 'color', 'colorscheme', 'comment', + 'distortion', 'fillcolor', 'fixedsize', 'fontcolor', 'fontname', + 'fontsize', 'group', 'height', 'id', 'image', 'imagescale', 'label', + 'labelloc', 'layer', 'margin', 'nojustify', 'orientation', 'penwidth', + 'peripheries', 'pin', 'pos', 'rects', 'regular', 'root', 'samplepoints', + 'shape', 'shapefile', 'showboxes', 'sides', 'skew', 'sortv', 'style', + 'target', 'tooltip', 'vertices', 'width', 'z', + # The following are attributes dot2tex + 'texlbl', 'texmode' ] ) + + +CLUSTER_ATTRIBUTES = set( ['K', 'URL', 'bgcolor', 'color', 'colorscheme', + 'fillcolor', 'fontcolor', 'fontname', 'fontsize', 'label', 'labeljust', + 'labelloc', 'lheight', 'lp', 'lwidth', 'nojustify', 'pencolor', + 'penwidth', 'peripheries', 'sortv', 'style', 'target', 'tooltip'] ) + + +# +# Extented version of ASPN's Python Cookbook Recipe: +# Frozen dictionaries. +# http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/414283 +# +# This version freezes dictionaries used as values within dictionaries. +# +class frozendict(dict): + def _blocked_attribute(obj): + raise AttributeError, "A frozendict cannot be modified." + _blocked_attribute = property(_blocked_attribute) + + __delitem__ = __setitem__ = clear = _blocked_attribute + pop = popitem = setdefault = update = _blocked_attribute + + def __new__(cls, *args, **kw): + new = dict.__new__(cls) + + args_ = [] + for arg in args: + if isinstance(arg, dict): + arg = copy.copy(arg) + for k, v in arg.iteritems(): + if isinstance(v, frozendict): + arg[k] = v + elif isinstance(v, dict): + arg[k] = frozendict(v) + elif isinstance(v, list): + v_ = list() + for elm in v: + if isinstance(elm, dict): + v_.append( frozendict(elm) ) + else: + v_.append( elm ) + arg[k] = tuple(v_) + args_.append( arg ) + else: + args_.append( arg ) + + dict.__init__(new, *args_, **kw) + return new + + def __init__(self, *args, **kw): + pass + + def __hash__(self): + try: + return self._cached_hash + except AttributeError: + h = self._cached_hash = hash(tuple(sorted(self.iteritems()))) + return h + + def __repr__(self): + return "frozendict(%s)" % dict.__repr__(self) + + +dot_keywords = ['graph', 'subgraph', 'digraph', 'node', 'edge', 'strict'] + +id_re_alpha_nums = re.compile('^[_a-zA-Z][a-zA-Z0-9_,]*$', re.UNICODE) +id_re_alpha_nums_with_ports = re.compile('^[_a-zA-Z][a-zA-Z0-9_,:\"]*[a-zA-Z0-9_,\"]+$', re.UNICODE) +id_re_num = re.compile('^[0-9,]+$', re.UNICODE) +id_re_with_port = re.compile('^([^:]*):([^:]*)$', re.UNICODE) +id_re_dbl_quoted = re.compile('^\".*\"$', re.S|re.UNICODE) +id_re_html = re.compile('^<.*>$', re.S|re.UNICODE) + + +def needs_quotes( s ): + """Checks whether a string is a dot language ID. + + It will check whether the string is solely composed + by the characters allowed in an ID or not. + If the string is one of the reserved keywords it will + need quotes too but the user will need to add them + manually. + """ + + # If the name is a reserved keyword it will need quotes but pydot + # can't tell when it's being used as a keyword or when it's simply + # a name. Hence the user needs to supply the quotes when an element + # would use a reserved keyword as name. This function will return + # false indicating that a keyword string, if provided as-is, won't + # need quotes. + if s in dot_keywords: + return False + + chars = [ord(c) for c in s if ord(c)>0x7f or ord(c)==0] + if chars and not id_re_dbl_quoted.match(s) and not id_re_html.match(s): + return True + + for test_re in [id_re_alpha_nums, id_re_num, id_re_dbl_quoted, id_re_html, id_re_alpha_nums_with_ports]: + if test_re.match(s): + return False + + m = id_re_with_port.match(s) + if m: + return needs_quotes(m.group(1)) or needs_quotes(m.group(2)) + + return True + + +def quote_if_necessary(s): + + if isinstance(s, bool): + if s is True: + return 'True' + return 'False' + + if not isinstance( s, basestring ): + return s + + if not s: + return s + + if needs_quotes(s): + replace = {'"' : r'\"', + "\n" : r'\n', + "\r" : r'\r'} + for (a,b) in replace.items(): + s = s.replace(a, b) + + return '"' + s + '"' + + return s + + + +def graph_from_dot_data(data): + """Load graph as defined by data in DOT format. + + The data is assumed to be in DOT format. It will + be parsed and a Dot class will be returned, + representing the graph. + """ + + return dot_parser.parse_dot_data(data) + + +def graph_from_dot_file(path): + """Load graph as defined by a DOT file. + + The file is assumed to be in DOT format. It will + be loaded, parsed and a Dot class will be returned, + representing the graph. + """ + + fd = file(path, 'rb') + data = fd.read() + fd.close() + + return graph_from_dot_data(data) + + + +def graph_from_edges(edge_list, node_prefix='', directed=False): + """Creates a basic graph out of an edge list. + + The edge list has to be a list of tuples representing + the nodes connected by the edge. + The values can be anything: bool, int, float, str. + + If the graph is undirected by default, it is only + calculated from one of the symmetric halves of the matrix. + """ + + if directed: + graph = Dot(graph_type='digraph') + + else: + graph = Dot(graph_type='graph') + + for edge in edge_list: + + if isinstance(edge[0], str): + src = node_prefix + edge[0] + else: + src = node_prefix + str(edge[0]) + + if isinstance(edge[1], str): + dst = node_prefix + edge[1] + else: + dst = node_prefix + str(edge[1]) + + e = Edge( src, dst ) + graph.add_edge(e) + + return graph + + +def graph_from_adjacency_matrix(matrix, node_prefix= u'', directed=False): + """Creates a basic graph out of an adjacency matrix. + + The matrix has to be a list of rows of values + representing an adjacency matrix. + The values can be anything: bool, int, float, as long + as they can evaluate to True or False. + """ + + node_orig = 1 + + if directed: + graph = Dot(graph_type='digraph') + else: + graph = Dot(graph_type='graph') + + for row in matrix: + if not directed: + skip = matrix.index(row) + r = row[skip:] + else: + skip = 0 + r = row + node_dest = skip+1 + + for e in r: + if e: + graph.add_edge( + Edge( node_prefix + node_orig, + node_prefix + node_dest) ) + node_dest += 1 + node_orig += 1 + + return graph + + + +def graph_from_incidence_matrix(matrix, node_prefix='', directed=False): + """Creates a basic graph out of an incidence matrix. + + The matrix has to be a list of rows of values + representing an incidence matrix. + The values can be anything: bool, int, float, as long + as they can evaluate to True or False. + """ + + node_orig = 1 + + if directed: + graph = Dot(graph_type='digraph') + else: + graph = Dot(graph_type='graph') + + for row in matrix: + nodes = [] + c = 1 + + for node in row: + if node: + nodes.append(c*node) + c += 1 + nodes.sort() + + if len(nodes) == 2: + graph.add_edge( + Edge( node_prefix + abs(nodes[0]), + node_prefix + nodes[1] )) + + if not directed: + graph.set_simplify(True) + + return graph + + + + +def __find_executables(path): + """Used by find_graphviz + + path - single directory as a string + + If any of the executables are found, it will return a dictionary + containing the program names as keys and their paths as values. + + Otherwise returns None + """ + + success = False + progs = {'dot': '', 'twopi': '', 'neato': '', 'circo': '', 'fdp': '', 'sfdp': ''} + + was_quoted = False + path = path.strip() + if path.startswith('"') and path.endswith('"'): + path = path[1:-1] + was_quoted = True + + if os.path.isdir(path) : + + for prg in progs.iterkeys(): + + if progs[prg]: + continue + + if os.path.exists( os.path.join(path, prg) ): + + if was_quoted: + progs[prg] = '"' + os.path.join(path, prg) + '"' + else: + progs[prg] = os.path.join(path, prg) + + success = True + + elif os.path.exists( os.path.join(path, prg + '.exe') ): + + if was_quoted: + progs[prg] = '"' + os.path.join(path, prg + '.exe') + '"' + else: + progs[prg] = os.path.join(path, prg + '.exe') + + success = True + + if success: + + return progs + + else: + + return None + + + +# The multi-platform version of this 'find_graphviz' function was +# contributed by Peter Cock +# +def find_graphviz(): + """Locate Graphviz's executables in the system. + + Tries three methods: + + First: Windows Registry (Windows only) + This requires Mark Hammond's pywin32 is installed. + + Secondly: Search the path + It will look for 'dot', 'twopi' and 'neato' in all the directories + specified in the PATH environment variable. + + Thirdly: Default install location (Windows only) + It will look for 'dot', 'twopi' and 'neato' in the default install + location under the "Program Files" directory. + + It will return a dictionary containing the program names as keys + and their paths as values. + + If this fails, it returns None. + """ + + # Method 1 (Windows only) + # + if os.sys.platform == 'win32': + + HKEY_LOCAL_MACHINE = 0x80000002 + KEY_QUERY_VALUE = 0x0001 + + RegOpenKeyEx = None + RegQueryValueEx = None + RegCloseKey = None + + try: + import win32api, win32con + RegOpenKeyEx = win32api.RegOpenKeyEx + RegQueryValueEx = win32api.RegQueryValueEx + RegCloseKey = win32api.RegCloseKey + + except ImportError: + # Print a messaged suggesting they install these? + # + pass + + try: + import ctypes + + def RegOpenKeyEx(key, subkey, opt, sam): + result = ctypes.c_uint(0) + ctypes.windll.advapi32.RegOpenKeyExA(key, subkey, opt, sam, ctypes.byref(result)) + return result.value + + def RegQueryValueEx( hkey, valuename ): + data_type = ctypes.c_uint(0) + data_len = ctypes.c_uint(1024) + data = ctypes.create_string_buffer( 1024 ) + + res = ctypes.windll.advapi32.RegQueryValueExA(hkey, valuename, 0, + ctypes.byref(data_type), data, ctypes.byref(data_len)) + + return data.value + + RegCloseKey = ctypes.windll.advapi32.RegCloseKey + + except ImportError: + # Print a messaged suggesting they install these? + # + pass + + if RegOpenKeyEx is not None: + + # Get the GraphViz install path from the registry + # + hkey = None + potentialKeys = [ + "SOFTWARE\\ATT\\Graphviz", + "SOFTWARE\\AT&T Research Labs\\Graphviz", + ] + for potentialKey in potentialKeys: + + try: + hkey = RegOpenKeyEx( HKEY_LOCAL_MACHINE, + potentialKey, 0, KEY_QUERY_VALUE ) + + if hkey is not None: + path = RegQueryValueEx( hkey, "InstallPath" ) + RegCloseKey( hkey ) + + # The regitry variable might exist, left by old installations + # but with no value, in those cases we keep searching... + if not path: + continue + + # Now append the "bin" subdirectory: + # + path = os.path.join(path, "bin") + progs = __find_executables(path) + if progs is not None : + #print "Used Windows registry" + return progs + + except Exception, excp: + #raise excp + pass + else: + break + + + + # Method 2 (Linux, Windows etc) + # + if os.environ.has_key('PATH'): + + for path in os.environ['PATH'].split(os.pathsep): + progs = __find_executables(path) + if progs is not None : + #print "Used path" + return progs + + # Method 3 (Windows only) + # + if os.sys.platform == 'win32': + + # Try and work out the equivalent of "C:\Program Files" on this + # machine (might be on drive D:, or in a different language) + # + + if os.environ.has_key('PROGRAMFILES'): + + # Note, we could also use the win32api to get this + # information, but win32api may not be installed. + + path = os.path.join(os.environ['PROGRAMFILES'], 'ATT', 'GraphViz', 'bin') + + else: + + #Just in case, try the default... + path = r"C:\Program Files\att\Graphviz\bin" + + progs = __find_executables(path) + + if progs is not None : + + #print "Used default install location" + return progs + + + for path in ( + '/usr/bin', '/usr/local/bin', + '/opt/local/bin', + '/opt/bin', '/sw/bin', '/usr/share', + '/Applications/Graphviz.app/Contents/MacOS/' ): + + progs = __find_executables(path) + if progs is not None : + #print "Used path" + return progs + + # Failed to find GraphViz + # + return None + + +class Common(object): + """Common information to several classes. + + Should not be directly used, several classes are derived from + this one. + """ + + + def __getstate__(self): + + dict = copy.copy(self.obj_dict) + + return dict + + + def __setstate__(self, state): + + self.obj_dict = state + + + def __get_attribute__(self, attr): + """Look for default attributes for this node""" + + attr_val = self.obj_dict['attributes'].get(attr, None) + + if attr_val is None: + # get the defaults for nodes/edges + + default_node_name = self.obj_dict['type'] + + # The defaults for graphs are set on a node named 'graph' + if default_node_name in ('subgraph', 'digraph', 'cluster'): + default_node_name = 'graph' + + g = self.get_parent_graph() + if g is not None: + defaults = g.get_node( default_node_name ) + else: + return None + + # Multiple defaults could be set by having repeated 'graph [...]' + # 'node [...]', 'edge [...]' statements. In such case, if the + # same attribute is set in different statements, only the first + # will be returned. In order to get all, one would call the + # get_*_defaults() methods and handle those. Or go node by node + # (of the ones specifying defaults) and modify the attributes + # individually. + # + if not isinstance(defaults, (list, tuple)): + defaults = [defaults] + + for default in defaults: + attr_val = default.obj_dict['attributes'].get(attr, None) + if attr_val: + return attr_val + else: + return attr_val + + return None + + + def set_parent_graph(self, parent_graph): + + self.obj_dict['parent_graph'] = parent_graph + + + def get_parent_graph(self): + + return self.obj_dict.get('parent_graph', None) + + + def set(self, name, value): + """Set an attribute value by name. + + Given an attribute 'name' it will set its value to 'value'. + There's always the possibility of using the methods: + + set_'name'(value) + + which are defined for all the existing attributes. + """ + + self.obj_dict['attributes'][name] = value + + + def get(self, name): + """Get an attribute value by name. + + Given an attribute 'name' it will get its value. + There's always the possibility of using the methods: + + get_'name'() + + which are defined for all the existing attributes. + """ + + return self.obj_dict['attributes'].get(name, None) + + + def get_attributes(self): + """""" + + return self.obj_dict['attributes'] + + + def set_sequence(self, seq): + + self.obj_dict['sequence'] = seq + + + def get_sequence(self): + + return self.obj_dict['sequence'] + + + def create_attribute_methods(self, obj_attributes): + + #for attr in self.obj_dict['attributes']: + for attr in obj_attributes: + + # Generate all the Setter methods. + # + self.__setattr__( 'set_'+attr, lambda x, a=attr : self.obj_dict['attributes'].__setitem__(a, x) ) + + # Generate all the Getter methods. + # + self.__setattr__('get_'+attr, lambda a=attr : self.__get_attribute__(a)) + + + +class Error(Exception): + """General error handling class. + """ + def __init__(self, value): + self.value = value + def __str__(self): + return self.value + + +class InvocationException(Exception): + """To indicate that a ploblem occurred while running any of the GraphViz executables. + """ + def __init__(self, value): + self.value = value + def __str__(self): + return self.value + + + +class Node(Common): + """A graph node. + + This class represents a graph's node with all its attributes. + + node(name, attribute=value, ...) + + name: node's name + + All the attributes defined in the Graphviz dot language should + be supported. + """ + + def __init__(self, name = '', obj_dict = None, **attrs): + + # + # Nodes will take attributes of all other types because the defaults + # for any GraphViz object are dealt with as if they were Node definitions + # + + if obj_dict is not None: + + self.obj_dict = obj_dict + + else: + + self.obj_dict = dict() + + # Copy the attributes + # + self.obj_dict[ 'attributes' ] = dict( attrs ) + self.obj_dict[ 'type' ] = 'node' + self.obj_dict[ 'parent_graph' ] = None + self.obj_dict[ 'parent_node_list' ] = None + self.obj_dict[ 'sequence' ] = None + + # Remove the compass point + # + port = None + if isinstance(name, basestring) and not name.startswith('"'): + idx = name.find(':') + if idx > 0 and idx+1 < len(name): + name, port = name[:idx], name[idx:] + + if isinstance(name, (long, int)): + name = str(name) + + self.obj_dict['name'] = quote_if_necessary( name ) + self.obj_dict['port'] = port + + self.create_attribute_methods(NODE_ATTRIBUTES) + + + + def set_name(self, node_name): + """Set the node's name.""" + + self.obj_dict['name'] = node_name + + + def get_name(self): + """Get the node's name.""" + + return self.obj_dict['name'] + + + def get_port(self): + """Get the node's port.""" + + return self.obj_dict['port'] + + + def add_style(self, style): + + styles = self.obj_dict['attributes'].get('style', None) + if not styles and style: + styles = [ style ] + else: + styles = styles.split(',') + styles.append( style ) + + self.obj_dict['attributes']['style'] = ','.join( styles ) + + + def to_string(self): + """Returns a string representation of the node in dot language. + """ + + + # RMF: special case defaults for node, edge and graph properties. + # + node = quote_if_necessary(self.obj_dict['name']) + + node_attr = list() + + for attr, value in self.obj_dict['attributes'].iteritems(): + if value is not None: + node_attr.append( '%s=%s' % (attr, quote_if_necessary(value) ) ) + else: + node_attr.append( attr ) + + + # No point in having nodes setting any defaults if the don't set + # any attributes... + # + if node in ('graph', 'node', 'edge') and len(node_attr) == 0: + return '' + + node_attr = ', '.join(node_attr) + + if node_attr: + node += ' [' + node_attr + ']' + + return node + ';' + + + +class Edge(Common): + """A graph edge. + + This class represents a graph's edge with all its attributes. + + edge(src, dst, attribute=value, ...) + + src: source node's name + dst: destination node's name + + All the attributes defined in the Graphviz dot language should + be supported. + + Attributes can be set through the dynamically generated methods: + + set_[attribute name], i.e. set_label, set_fontname + + or directly by using the instance's special dictionary: + + Edge.obj_dict['attributes'][attribute name], i.e. + + edge_instance.obj_dict['attributes']['label'] + edge_instance.obj_dict['attributes']['fontname'] + + """ + + + + def __init__(self, src='', dst='', obj_dict=None, **attrs): + + if isinstance(src, (list, tuple)) and dst == '': + src, dst = src + + if obj_dict is not None: + + self.obj_dict = obj_dict + + else: + + self.obj_dict = dict() + + # Copy the attributes + # + self.obj_dict[ 'attributes' ] = dict( attrs ) + self.obj_dict[ 'type' ] = 'edge' + self.obj_dict[ 'parent_graph' ] = None + self.obj_dict[ 'parent_edge_list' ] = None + self.obj_dict[ 'sequence' ] = None + + if isinstance(src, Node): + src = src.get_name() + + if isinstance(dst, Node): + dst = dst.get_name() + + points = ( quote_if_necessary( src) , quote_if_necessary( dst) ) + + self.obj_dict['points'] = points + + self.create_attribute_methods(EDGE_ATTRIBUTES) + + + def get_source(self): + """Get the edges source node name.""" + + return self.obj_dict['points'][0] + + + def get_destination(self): + """Get the edge's destination node name.""" + + return self.obj_dict['points'][1] + + + def __hash__(self): + + return hash( hash(self.get_source()) + hash(self.get_destination()) ) + + + def __eq__(self, edge): + """Compare two edges. + + If the parent graph is directed, arcs linking + node A to B are considered equal and A->B != B->A + + If the parent graph is undirected, any edge + connecting two nodes is equal to any other + edge connecting the same nodes, A->B == B->A + """ + + if not isinstance(edge, Edge): + raise Error, "Can't compare and edge to a non-edge object." + + if self.get_parent_graph().get_top_graph_type() == 'graph': + + # If the graph is undirected, the edge has neither + # source nor destination. + # + if ( ( self.get_source() == edge.get_source() and self.get_destination() == edge.get_destination() ) or + ( edge.get_source() == self.get_destination() and edge.get_destination() == self.get_source() ) ): + return True + + else: + + if self.get_source()==edge.get_source() and self.get_destination()==edge.get_destination() : + return True + + return False + + + + def parse_node_ref(self, node_str): + + if not isinstance(node_str, str): + return node_str + + if node_str.startswith('"') and node_str.endswith('"'): + + return node_str + + node_port_idx = node_str.rfind(':') + + if node_port_idx>0 and node_str[0]=='"' and node_str[node_port_idx-1]=='"': + + return node_str + + if node_port_idx>0: + + a = node_str[:node_port_idx] + b = node_str[node_port_idx+1:] + + node = quote_if_necessary(a) + + node += ':'+quote_if_necessary(b) + + return node + + return node_str + + + def to_string(self): + """Returns a string representation of the edge in dot language. + """ + + src = self.parse_node_ref( self.get_source() ) + dst = self.parse_node_ref( self.get_destination() ) + + if isinstance(src, frozendict): + edge = [ Subgraph(obj_dict=src).to_string() ] + elif isinstance(src, (int, long)): + edge = [ str(src) ] + else: + edge = [ src ] + + if (self.get_parent_graph() and + self.get_parent_graph().get_top_graph_type() and + self.get_parent_graph().get_top_graph_type() == 'digraph' ): + + edge.append( '->' ) + + else: + edge.append( '--' ) + + if isinstance(dst, frozendict): + edge.append( Subgraph(obj_dict=dst).to_string() ) + elif isinstance(dst, (int, long)): + edge.append( str(dst) ) + else: + edge.append( dst ) + + + edge_attr = list() + + for attr, value in self.obj_dict['attributes'].iteritems(): + + if value is not None: + edge_attr.append( '%s=%s' % (attr, quote_if_necessary(value) ) ) + else: + edge_attr.append( attr ) + + edge_attr = ', '.join(edge_attr) + + if edge_attr: + edge.append( ' [' + edge_attr + ']' ) + + return ' '.join(edge) + ';' + + + + + +class Graph(Common): + """Class representing a graph in Graphviz's dot language. + + This class implements the methods to work on a representation + of a graph in Graphviz's dot language. + + graph( graph_name='G', graph_type='digraph', + strict=False, suppress_disconnected=False, attribute=value, ...) + + graph_name: + the graph's name + graph_type: + can be 'graph' or 'digraph' + suppress_disconnected: + defaults to False, which will remove from the + graph any disconnected nodes. + simplify: + if True it will avoid displaying equal edges, i.e. + only one edge between two nodes. removing the + duplicated ones. + + All the attributes defined in the Graphviz dot language should + be supported. + + Attributes can be set through the dynamically generated methods: + + set_[attribute name], i.e. set_size, set_fontname + + or using the instance's attributes: + + Graph.obj_dict['attributes'][attribute name], i.e. + + graph_instance.obj_dict['attributes']['label'] + graph_instance.obj_dict['attributes']['fontname'] + """ + + + def __init__(self, graph_name='G', obj_dict=None, graph_type='digraph', strict=False, + suppress_disconnected=False, simplify=False, **attrs): + + if obj_dict is not None: + self.obj_dict = obj_dict + + else: + + self.obj_dict = dict() + + self.obj_dict['attributes'] = dict(attrs) + + if graph_type not in ['graph', 'digraph']: + raise Error, 'Invalid type "%s". Accepted graph types are: graph, digraph, subgraph' % graph_type + + + self.obj_dict['name'] = quote_if_necessary(graph_name) + self.obj_dict['type'] = graph_type + + self.obj_dict['strict'] = strict + self.obj_dict['suppress_disconnected'] = suppress_disconnected + self.obj_dict['simplify'] = simplify + + self.obj_dict['current_child_sequence'] = 1 + self.obj_dict['nodes'] = dict() + self.obj_dict['edges'] = dict() + self.obj_dict['subgraphs'] = dict() + + self.set_parent_graph(self) + + + self.create_attribute_methods(GRAPH_ATTRIBUTES) + + + def get_graph_type(self): + + return self.obj_dict['type'] + + + def get_top_graph_type(self): + + parent = self + while True: + parent_ = parent.get_parent_graph() + if parent_ == parent: + break + parent = parent_ + + return parent.obj_dict['type'] + + + def set_graph_defaults(self, **attrs): + + self.add_node( Node('graph', **attrs) ) + + + def get_graph_defaults(self, **attrs): + + graph_nodes = self.get_node('graph') + + if isinstance( graph_nodes, (list, tuple)): + return [ node.get_attributes() for node in graph_nodes ] + + return graph_nodes.get_attributes() + + + + def set_node_defaults(self, **attrs): + + self.add_node( Node('node', **attrs) ) + + + def get_node_defaults(self, **attrs): + + + graph_nodes = self.get_node('node') + + if isinstance( graph_nodes, (list, tuple)): + return [ node.get_attributes() for node in graph_nodes ] + + return graph_nodes.get_attributes() + + + def set_edge_defaults(self, **attrs): + + self.add_node( Node('edge', **attrs) ) + + + + def get_edge_defaults(self, **attrs): + + graph_nodes = self.get_node('edge') + + if isinstance( graph_nodes, (list, tuple)): + return [ node.get_attributes() for node in graph_nodes ] + + return graph_nodes.get_attributes() + + + + def set_simplify(self, simplify): + """Set whether to simplify or not. + + If True it will avoid displaying equal edges, i.e. + only one edge between two nodes. removing the + duplicated ones. + """ + + self.obj_dict['simplify'] = simplify + + + + def get_simplify(self): + """Get whether to simplify or not. + + Refer to set_simplify for more information. + """ + + return self.obj_dict['simplify'] + + + def set_type(self, graph_type): + """Set the graph's type, 'graph' or 'digraph'.""" + + self.obj_dict['type'] = graph_type + + + + def get_type(self): + """Get the graph's type, 'graph' or 'digraph'.""" + + return self.obj_dict['type'] + + + + def set_name(self, graph_name): + """Set the graph's name.""" + + self.obj_dict['name'] = graph_name + + + + def get_name(self): + """Get the graph's name.""" + + return self.obj_dict['name'] + + + + def set_strict(self, val): + """Set graph to 'strict' mode. + + This option is only valid for top level graphs. + """ + + self.obj_dict['strict'] = val + + + + def get_strict(self, val): + """Get graph's 'strict' mode (True, False). + + This option is only valid for top level graphs. + """ + + return self.obj_dict['strict'] + + + + def set_suppress_disconnected(self, val): + """Suppress disconnected nodes in the output graph. + + This option will skip nodes in the graph with no incoming or outgoing + edges. This option works also for subgraphs and has effect only in the + current graph/subgraph. + """ + + self.obj_dict['suppress_disconnected'] = val + + + + def get_suppress_disconnected(self, val): + """Get if suppress disconnected is set. + + Refer to set_suppress_disconnected for more information. + """ + + return self.obj_dict['suppress_disconnected'] + + + def get_next_sequence_number(self): + + seq = self.obj_dict['current_child_sequence'] + + self.obj_dict['current_child_sequence'] += 1 + + return seq + + + + def add_node(self, graph_node): + """Adds a node object to the graph. + + It takes a node object as its only argument and returns + None. + """ + + if not isinstance(graph_node, Node): + raise TypeError('add_node() received a non node class object: ' + str(graph_node)) + + + node = self.get_node(graph_node.get_name()) + + if not node: + + self.obj_dict['nodes'][graph_node.get_name()] = [ graph_node.obj_dict ] + + #self.node_dict[graph_node.get_name()] = graph_node.attributes + graph_node.set_parent_graph(self.get_parent_graph()) + + else: + + self.obj_dict['nodes'][graph_node.get_name()].append( graph_node.obj_dict ) + + graph_node.set_sequence(self.get_next_sequence_number()) + + + + def del_node(self, name, index=None): + """Delete a node from the graph. + + Given a node's name all node(s) with that same name + will be deleted if 'index' is not specified or set + to None. + If there are several nodes with that same name and + 'index' is given, only the node in that position + will be deleted. + + 'index' should be an integer specifying the position + of the node to delete. If index is larger than the + number of nodes with that name, no action is taken. + + If nodes are deleted it returns True. If no action + is taken it returns False. + """ + + if isinstance(name, Node): + name = name.get_name() + + if self.obj_dict['nodes'].has_key(name): + + if index is not None and index < len(self.obj_dict['nodes'][name]): + del self.obj_dict['nodes'][name][index] + return True + else: + del self.obj_dict['nodes'][name] + return True + + return False + + + def get_node(self, name): + """Retrieve a node from the graph. + + Given a node's name the corresponding Node + instance will be returned. + + If one or more nodes exist with that name a list of + Node instances is returned. + An empty list is returned otherwise. + """ + + match = list() + + if self.obj_dict['nodes'].has_key(name): + + match.extend( [ Node( obj_dict = obj_dict ) for obj_dict in self.obj_dict['nodes'][name] ]) + + return match + + + def get_nodes(self): + """Get the list of Node instances.""" + + return self.get_node_list() + + + def get_node_list(self): + """Get the list of Node instances. + + This method returns the list of Node instances + composing the graph. + """ + + node_objs = list() + + for node, obj_dict_list in self.obj_dict['nodes'].iteritems(): + node_objs.extend( [ Node( obj_dict = obj_d ) for obj_d in obj_dict_list ] ) + + return node_objs + + + + def add_edge(self, graph_edge): + """Adds an edge object to the graph. + + It takes a edge object as its only argument and returns + None. + """ + + if not isinstance(graph_edge, Edge): + raise TypeError('add_edge() received a non edge class object: ' + str(graph_edge)) + + edge_points = ( graph_edge.get_source(), graph_edge.get_destination() ) + + if self.obj_dict['edges'].has_key(edge_points): + + edge_list = self.obj_dict['edges'][edge_points] + edge_list.append(graph_edge.obj_dict) + + else: + + self.obj_dict['edges'][edge_points] = [ graph_edge.obj_dict ] + + + graph_edge.set_sequence( self.get_next_sequence_number() ) + + graph_edge.set_parent_graph( self.get_parent_graph() ) + + + + def del_edge(self, src_or_list, dst=None, index=None): + """Delete an edge from the graph. + + Given an edge's (source, destination) node names all + matching edges(s) will be deleted if 'index' is not + specified or set to None. + If there are several matching edges and 'index' is + given, only the edge in that position will be deleted. + + 'index' should be an integer specifying the position + of the edge to delete. If index is larger than the + number of matching edges, no action is taken. + + If edges are deleted it returns True. If no action + is taken it returns False. + """ + + if isinstance( src_or_list, (list, tuple)): + if dst is not None and isinstance(dst, (int, long)): + index = dst + src, dst = src_or_list + else: + src, dst = src_or_list, dst + + if isinstance(src, Node): + src = src.get_name() + + if isinstance(dst, Node): + dst = dst.get_name() + + if self.obj_dict['edges'].has_key( (src, dst) ): + + if index is not None and index < len(self.obj_dict['edges'][(src, dst)]): + del self.obj_dict['edges'][(src, dst)][index] + return True + else: + del self.obj_dict['edges'][(src, dst)] + return True + + return False + + + def get_edge(self, src_or_list, dst=None): + """Retrieved an edge from the graph. + + Given an edge's source and destination the corresponding + Edge instance(s) will be returned. + + If one or more edges exist with that source and destination + a list of Edge instances is returned. + An empty list is returned otherwise. + """ + + if isinstance( src_or_list, (list, tuple)) and dst is None: + edge_points = tuple(src_or_list) + edge_points_reverse = (edge_points[1], edge_points[0]) + else: + edge_points = (src_or_list, dst) + edge_points_reverse = (dst, src_or_list) + + match = list() + + if self.obj_dict['edges'].has_key( edge_points ) or ( + self.get_top_graph_type() == 'graph' and self.obj_dict['edges'].has_key( edge_points_reverse )): + + edges_obj_dict = self.obj_dict['edges'].get( + edge_points, + self.obj_dict['edges'].get( edge_points_reverse, None )) + + for edge_obj_dict in edges_obj_dict: + match.append( Edge( edge_points[0], edge_points[1], obj_dict = edge_obj_dict ) ) + + return match + + + def get_edges(self): + return self.get_edge_list() + + + def get_edge_list(self): + """Get the list of Edge instances. + + This method returns the list of Edge instances + composing the graph. + """ + + edge_objs = list() + + for edge, obj_dict_list in self.obj_dict['edges'].iteritems(): + edge_objs.extend( [ Edge( obj_dict = obj_d ) for obj_d in obj_dict_list ] ) + + return edge_objs + + + + def add_subgraph(self, sgraph): + """Adds an subgraph object to the graph. + + It takes a subgraph object as its only argument and returns + None. + """ + + if not isinstance(sgraph, Subgraph) and not isinstance(sgraph, Cluster): + raise TypeError('add_subgraph() received a non subgraph class object:' + str(sgraph)) + + if self.obj_dict['subgraphs'].has_key(sgraph.get_name()): + + sgraph_list = self.obj_dict['subgraphs'][ sgraph.get_name() ] + sgraph_list.append( sgraph.obj_dict ) + + else: + self.obj_dict['subgraphs'][ sgraph.get_name() ] = [ sgraph.obj_dict ] + + sgraph.set_sequence( self.get_next_sequence_number() ) + + sgraph.set_parent_graph( self.get_parent_graph() ) + + + + + def get_subgraph(self, name): + """Retrieved a subgraph from the graph. + + Given a subgraph's name the corresponding + Subgraph instance will be returned. + + If one or more subgraphs exist with the same name, a list of + Subgraph instances is returned. + An empty list is returned otherwise. + """ + + match = list() + + if self.obj_dict['subgraphs'].has_key( name ): + + sgraphs_obj_dict = self.obj_dict['subgraphs'].get( name ) + + for obj_dict_list in sgraphs_obj_dict: + #match.extend( Subgraph( obj_dict = obj_d ) for obj_d in obj_dict_list ) + match.append( Subgraph( obj_dict = obj_dict_list ) ) + + return match + + + def get_subgraphs(self): + + return self.get_subgraph_list() + + + def get_subgraph_list(self): + """Get the list of Subgraph instances. + + This method returns the list of Subgraph instances + in the graph. + """ + + sgraph_objs = list() + + for sgraph, obj_dict_list in self.obj_dict['subgraphs'].iteritems(): + sgraph_objs.extend( [ Subgraph( obj_dict = obj_d ) for obj_d in obj_dict_list ] ) + + return sgraph_objs + + + + def set_parent_graph(self, parent_graph): + + self.obj_dict['parent_graph'] = parent_graph + + for obj_list in self.obj_dict['nodes'].itervalues(): + for obj in obj_list: + obj['parent_graph'] = parent_graph + + for obj_list in self.obj_dict['edges'].itervalues(): + for obj in obj_list: + obj['parent_graph'] = parent_graph + + for obj_list in self.obj_dict['subgraphs'].itervalues(): + for obj in obj_list: + Graph(obj_dict=obj).set_parent_graph(parent_graph) + + + + def to_string(self): + """Returns a string representation of the graph in dot language. + + It will return the graph and all its subelements in string from. + """ + + + graph = list() + + if self.obj_dict.get('strict', None) is not None: + + if self==self.get_parent_graph() and self.obj_dict['strict']: + + graph.append('strict ') + + if self.obj_dict['name'] == '': + if 'show_keyword' in self.obj_dict and self.obj_dict['show_keyword']: + graph.append( 'subgraph {\n' ) + else: + graph.append( '{\n' ) + else: + graph.append( '%s %s {\n' % (self.obj_dict['type'], self.obj_dict['name']) ) + + + for attr in self.obj_dict['attributes'].iterkeys(): + + if self.obj_dict['attributes'].get(attr, None) is not None: + + val = self.obj_dict['attributes'].get(attr) + if val is not None: + graph.append( '%s=%s' % (attr, quote_if_necessary(val)) ) + else: + graph.append( attr ) + + graph.append( ';\n' ) + + + edges_done = set() + + edge_obj_dicts = list() + for e in self.obj_dict['edges'].itervalues(): + edge_obj_dicts.extend(e) + + if edge_obj_dicts: + edge_src_set, edge_dst_set = zip( *[obj['points'] for obj in edge_obj_dicts] ) + edge_src_set, edge_dst_set = set(edge_src_set), set(edge_dst_set) + else: + edge_src_set, edge_dst_set = set(), set() + + node_obj_dicts = list() + for e in self.obj_dict['nodes'].itervalues(): + node_obj_dicts.extend(e) + + sgraph_obj_dicts = list() + for sg in self.obj_dict['subgraphs'].itervalues(): + sgraph_obj_dicts.extend(sg) + + + obj_list = [ (obj['sequence'], obj) for obj in (edge_obj_dicts + node_obj_dicts + sgraph_obj_dicts) ] + obj_list.sort() + + for idx, obj in obj_list: + + if obj['type'] == 'node': + + node = Node(obj_dict=obj) + + if self.obj_dict.get('suppress_disconnected', False): + + if (node.get_name() not in edge_src_set and + node.get_name() not in edge_dst_set): + + continue + + graph.append( node.to_string()+'\n' ) + + elif obj['type'] == 'edge': + + edge = Edge(obj_dict=obj) + + if self.obj_dict.get('simplify', False) and edge in edges_done: + continue + + graph.append( edge.to_string() + '\n' ) + edges_done.add(edge) + + else: + + sgraph = Subgraph(obj_dict=obj) + + graph.append( sgraph.to_string()+'\n' ) + + graph.append( '}\n' ) + + return ''.join(graph) + + + +class Subgraph(Graph): + + """Class representing a subgraph in Graphviz's dot language. + + This class implements the methods to work on a representation + of a subgraph in Graphviz's dot language. + + subgraph(graph_name='subG', suppress_disconnected=False, attribute=value, ...) + + graph_name: + the subgraph's name + suppress_disconnected: + defaults to false, which will remove from the + subgraph any disconnected nodes. + All the attributes defined in the Graphviz dot language should + be supported. + + Attributes can be set through the dynamically generated methods: + + set_[attribute name], i.e. set_size, set_fontname + + or using the instance's attributes: + + Subgraph.obj_dict['attributes'][attribute name], i.e. + + subgraph_instance.obj_dict['attributes']['label'] + subgraph_instance.obj_dict['attributes']['fontname'] + """ + + + # RMF: subgraph should have all the attributes of graph so it can be passed + # as a graph to all methods + # + def __init__(self, graph_name='', obj_dict=None, suppress_disconnected=False, + simplify=False, **attrs): + + + Graph.__init__(self, graph_name=graph_name, obj_dict=obj_dict, + suppress_disconnected=suppress_disconnected, simplify=simplify, **attrs) + + if obj_dict is None: + + self.obj_dict['type'] = 'subgraph' + + + + +class Cluster(Graph): + + """Class representing a cluster in Graphviz's dot language. + + This class implements the methods to work on a representation + of a cluster in Graphviz's dot language. + + cluster(graph_name='subG', suppress_disconnected=False, attribute=value, ...) + + graph_name: + the cluster's name (the string 'cluster' will be always prepended) + suppress_disconnected: + defaults to false, which will remove from the + cluster any disconnected nodes. + All the attributes defined in the Graphviz dot language should + be supported. + + Attributes can be set through the dynamically generated methods: + + set_[attribute name], i.e. set_color, set_fontname + + or using the instance's attributes: + + Cluster.obj_dict['attributes'][attribute name], i.e. + + cluster_instance.obj_dict['attributes']['label'] + cluster_instance.obj_dict['attributes']['fontname'] + """ + + + def __init__(self, graph_name='subG', obj_dict=None, suppress_disconnected=False, + simplify=False, **attrs): + + Graph.__init__(self, graph_name=graph_name, obj_dict=obj_dict, + suppress_disconnected=suppress_disconnected, simplify=simplify, **attrs) + + if obj_dict is None: + + self.obj_dict['type'] = 'subgraph' + self.obj_dict['name'] = 'cluster_'+graph_name + + self.create_attribute_methods(CLUSTER_ATTRIBUTES) + + + + + + +class Dot(Graph): + """A container for handling a dot language file. + + This class implements methods to write and process + a dot language file. It is a derived class of + the base class 'Graph'. + """ + + + + def __init__(self, *argsl, **argsd): + Graph.__init__(self, *argsl, **argsd) + + self.shape_files = list() + + self.progs = None + + self.formats = ['canon', 'cmap', 'cmapx', 'cmapx_np', 'dia', 'dot', + 'fig', 'gd', 'gd2', 'gif', 'hpgl', 'imap', 'imap_np', 'ismap', + 'jpe', 'jpeg', 'jpg', 'mif', 'mp', 'pcl', 'pdf', 'pic', 'plain', + 'plain-ext', 'png', 'ps', 'ps2', 'svg', 'svgz', 'vml', 'vmlz', + 'vrml', 'vtx', 'wbmp', 'xdot', 'xlib' ] + + self.prog = 'dot' + + # Automatically creates all the methods enabling the creation + # of output in any of the supported formats. + for frmt in self.formats: + self.__setattr__( + 'create_'+frmt, + lambda f=frmt, prog=self.prog : self.create(format=f, prog=prog)) + f = self.__dict__['create_'+frmt] + f.__doc__ = '''Refer to the docstring accompanying the 'create' method for more information.''' + + for frmt in self.formats+['raw']: + self.__setattr__( + 'write_'+frmt, + lambda path, f=frmt, prog=self.prog : self.write(path, format=f, prog=prog)) + + f = self.__dict__['write_'+frmt] + f.__doc__ = '''Refer to the docstring accompanying the 'write' method for more information.''' + + + + def __getstate__(self): + + dict = copy.copy(self.obj_dict) + + return dict + + def __setstate__(self, state): + + self.obj_dict = state + + + def set_shape_files(self, file_paths): + """Add the paths of the required image files. + + If the graph needs graphic objects to be used as shapes or otherwise + those need to be in the same folder as the graph is going to be rendered + from. Alternatively the absolute path to the files can be specified when + including the graphics in the graph. + + The files in the location pointed to by the path(s) specified as arguments + to this method will be copied to the same temporary location where the + graph is going to be rendered. + """ + + if isinstance( file_paths, basestring ): + self.shape_files.append( file_paths ) + + if isinstance( file_paths, (list, tuple) ): + self.shape_files.extend( file_paths ) + + + def set_prog(self, prog): + """Sets the default program. + + Sets the default program in charge of processing + the dot file into a graph. + """ + self.prog = prog + + + def set_graphviz_executables(self, paths): + """This method allows to manually specify the location of the GraphViz executables. + + The argument to this method should be a dictionary where the keys are as follows: + + {'dot': '', 'twopi': '', 'neato': '', 'circo': '', 'fdp': ''} + + and the values are the paths to the corresponding executable, including the name + of the executable itself. + """ + + self.progs = paths + + + def write(self, path, prog=None, format='raw'): + """Writes a graph to a file. + + Given a filename 'path' it will open/create and truncate + such file and write on it a representation of the graph + defined by the dot object and in the format specified by + 'format'. + The format 'raw' is used to dump the string representation + of the Dot object, without further processing. + The output can be processed by any of graphviz tools, defined + in 'prog', which defaults to 'dot' + Returns True or False according to the success of the write + operation. + + There's also the preferred possibility of using: + + write_'format'(path, prog='program') + + which are automatically defined for all the supported formats. + [write_ps(), write_gif(), write_dia(), ...] + """ + + if prog is None: + prog = self.prog + + dot_fd = file(path, "w+b") + if format == 'raw': + data = self.to_string() + if isinstance(data, basestring): + if not isinstance(data, unicode): + try: + data = unicode(data, 'utf-8') + except: + pass + + try: + data = data.encode('utf-8') + except: + pass + dot_fd.write(data) + else: + dot_fd.write(self.create(prog, format)) + dot_fd.close() + + return True + + + + def create(self, prog=None, format='ps'): + """Creates and returns a Postscript representation of the graph. + + create will write the graph to a temporary dot file and process + it with the program given by 'prog' (which defaults to 'twopi'), + reading the Postscript output and returning it as a string is the + operation is successful. + On failure None is returned. + + There's also the preferred possibility of using: + + create_'format'(prog='program') + + which are automatically defined for all the supported formats. + [create_ps(), create_gif(), create_dia(), ...] + + If 'prog' is a list instead of a string the fist item is expected + to be the program name, followed by any optional command-line + arguments for it: + + [ 'twopi', '-Tdot', '-s10' ] + """ + + if prog is None: + prog = self.prog + + if isinstance(prog, (list, tuple)): + prog, args = prog[0], prog[1:] + else: + args = [] + + if self.progs is None: + self.progs = find_graphviz() + if self.progs is None: + raise InvocationException( + 'GraphViz\'s executables not found' ) + + if not self.progs.has_key(prog): + raise InvocationException( + 'GraphViz\'s executable "%s" not found' % prog ) + + if not os.path.exists( self.progs[prog] ) or not os.path.isfile( self.progs[prog] ): + raise InvocationException( + 'GraphViz\'s executable "%s" is not a file or doesn\'t exist' % self.progs[prog] ) + + + tmp_fd, tmp_name = tempfile.mkstemp() + os.close(tmp_fd) + self.write(tmp_name) + tmp_dir = os.path.dirname(tmp_name ) + + # For each of the image files... + # + for img in self.shape_files: + + # Get its data + # + f = file(img, 'rb') + f_data = f.read() + f.close() + + # And copy it under a file with the same name in the temporary directory + # + f = file( os.path.join( tmp_dir, os.path.basename(img) ), 'wb' ) + f.write(f_data) + f.close() + + cmdline = [self.progs[prog], '-T'+format, tmp_name] + args + + p = subprocess.Popen( + cmdline, + cwd=tmp_dir, + stderr=subprocess.PIPE, stdout=subprocess.PIPE) + + stderr = p.stderr + stdout = p.stdout + + stdout_output = list() + while True: + data = stdout.read() + if not data: + break + stdout_output.append(data) + stdout.close() + + stdout_output = ''.join(stdout_output) + + if not stderr.closed: + stderr_output = list() + while True: + data = stderr.read() + if not data: + break + stderr_output.append(data) + stderr.close() + + if stderr_output: + stderr_output = ''.join(stderr_output) + + #pid, status = os.waitpid(p.pid, 0) + status = p.wait() + + if status != 0 : + raise InvocationException( + 'Program terminated with status: %d. stderr follows: %s' % ( + status, stderr_output) ) + elif stderr_output: + print stderr_output + + # For each of the image files... + # + for img in self.shape_files: + + # remove it + # + os.unlink( os.path.join( tmp_dir, os.path.basename(img) ) ) + + os.unlink(tmp_name) + + return stdout_output diff --git a/Assignment 3/packages/treeViewer.py b/Assignment 3/packages/treeViewer.py new file mode 100755 index 0000000..eadcc28 --- /dev/null +++ b/Assignment 3/packages/treeViewer.py @@ -0,0 +1,17 @@ +from sklearn.externals.six import StringIO +from sklearn import tree +import pydot + +FNAMES = ['Fixed acidity', 'Volatile acidity', 'Citric acid', 'Residual sugar', 'Chlorides', 'Free sulfur dioxide', 'Total sulfur dioxide', 'Density', 'pH', 'Sulphates', 'Alcohol'] + +def view(classifier): + """ Renders a graph representation of classifier, and + saves it to "MyTree.pdf" in the same folder + as the executing script. + """ + tree_dot = StringIO() + tree.export_graphviz(classifier, out_file=tree_dot, feature_names=FNAMES) + graph = pydot.graph_from_dot_data(tree_dot.getvalue()) + print "A" + graph.write_pdf("MyTree.pdf") + print "B" diff --git a/Assignment 3/packages/xlrd/__init__.py b/Assignment 3/packages/xlrd/__init__.py new file mode 100644 index 0000000..5b9274e --- /dev/null +++ b/Assignment 3/packages/xlrd/__init__.py @@ -0,0 +1,467 @@ +from os import path + +from .info import __VERSION__ + +#

Copyright (c) 2005-2012 Stephen John Machin, Lingfo Pty Ltd

+#

This module is part of the xlrd package, which is released under a +# BSD-style licence.

+ +from . import licences + +## +#

A Python module for extracting data from MS Excel (TM) spreadsheet files. +#

+# Version 0.7.4 -- April 2012 +#

+# +#

General information

+# +#

Acknowledgements

+# +#

+# Development of this module would not have been possible without the document +# "OpenOffice.org's Documentation of the Microsoft Excel File Format" +# ("OOo docs" for short). +# The latest version is available from OpenOffice.org in +# PDF format +# and +# ODT format. +# Small portions of the OOo docs are reproduced in this +# document. A study of the OOo docs is recommended for those who wish a +# deeper understanding of the Excel file layout than the xlrd docs can provide. +#

+# +#

Backporting to Python 2.1 was partially funded by +# +# Journyx - provider of timesheet and project accounting solutions. +# +#

+# +#

Provision of formatting information in version 0.6.1 was funded by +# +# Simplistix Ltd. +# +#

+# +#

Unicode

+# +#

This module presents all text strings as Python unicode objects. +# From Excel 97 onwards, text in Excel spreadsheets has been stored as Unicode. +# Older files (Excel 95 and earlier) don't keep strings in Unicode; +# a CODEPAGE record provides a codepage number (for example, 1252) which is +# used by xlrd to derive the encoding (for same example: "cp1252") which is +# used to translate to Unicode.

+# +#

If the CODEPAGE record is missing (possible if the file was created +# by third-party software), xlrd will assume that the encoding is ascii, and keep going. +# If the actual encoding is not ascii, a UnicodeDecodeError exception will be raised and +# you will need to determine the encoding yourself, and tell xlrd: +# 

+#     book = xlrd.open_workbook(..., encoding_override="cp1252")
+#

+#

If the CODEPAGE record exists but is wrong (for example, the codepage +# number is 1251, but the strings are actually encoded in koi8_r), +# it can be overridden using the same mechanism. +# The supplied runxlrd.py has a corresponding command-line argument, which +# may be used for experimentation: +# 

+#     runxlrd.py -e koi8_r 3rows myfile.xls
+#

+#

The first place to look for an encoding ("codec name") is +# +# the Python documentation. +#

+# +# +#

Dates in Excel spreadsheets

+# +#

In reality, there are no such things. What you have are floating point +# numbers and pious hope. +# There are several problems with Excel dates:

+# +#

(1) Dates are not stored as a separate data type; they are stored as +# floating point numbers and you have to rely on +# (a) the "number format" applied to them in Excel and/or +# (b) knowing which cells are supposed to have dates in them. +# This module helps with (a) by inspecting the +# format that has been applied to each number cell; +# if it appears to be a date format, the cell +# is classified as a date rather than a number. Feedback on this feature, +# especially from non-English-speaking locales, would be appreciated.

+# +#

(2) Excel for Windows stores dates by default as the number of +# days (or fraction thereof) since 1899-12-31T00:00:00. Excel for +# Macintosh uses a default start date of 1904-01-01T00:00:00. The date +# system can be changed in Excel on a per-workbook basis (for example: +# Tools -> Options -> Calculation, tick the "1904 date system" box). +# This is of course a bad idea if there are already dates in the +# workbook. There is no good reason to change it even if there are no +# dates in the workbook. Which date system is in use is recorded in the +# workbook. A workbook transported from Windows to Macintosh (or vice +# versa) will work correctly with the host Excel. When using this +# module's xldate_as_tuple function to convert numbers from a workbook, +# you must use the datemode attribute of the Book object. If you guess, +# or make a judgement depending on where you believe the workbook was +# created, you run the risk of being 1462 days out of kilter.

+# +#

Reference: +# http://support.microsoft.com/default.aspx?scid=KB;EN-US;q180162

+# +# +#

(3) The Excel implementation of the Windows-default 1900-based date system works on the +# incorrect premise that 1900 was a leap year. It interprets the number 60 as meaning 1900-02-29, +# which is not a valid date. Consequently any number less than 61 is ambiguous. Example: is 59 the +# result of 1900-02-28 entered directly, or is it 1900-03-01 minus 2 days? The OpenOffice.org Calc +# program "corrects" the Microsoft problem; entering 1900-02-27 causes the number 59 to be stored. +# Save as an XLS file, then open the file with Excel -- you'll see 1900-02-28 displayed.

+# +#

Reference: http://support.microsoft.com/default.aspx?scid=kb;en-us;214326

+# +#

(4) The Macintosh-default 1904-based date system counts 1904-01-02 as day 1 and 1904-01-01 as day zero. +# Thus any number such that (0.0 <= number < 1.0) is ambiguous. Is 0.625 a time of day (15:00:00), +# independent of the calendar, +# or should it be interpreted as an instant on a particular day (1904-01-01T15:00:00)? +# The xldate_* functions in this module +# take the view that such a number is a calendar-independent time of day (like Python's datetime.time type) for both +# date systems. This is consistent with more recent Microsoft documentation +# (for example, the help file for Excel 2002 which says that the first day +# in the 1904 date system is 1904-01-02). +# +#

(5) Usage of the Excel DATE() function may leave strange dates in a spreadsheet. Quoting the help file, +# in respect of the 1900 date system: "If year is between 0 (zero) and 1899 (inclusive), +# Excel adds that value to 1900 to calculate the year. For example, DATE(108,1,2) returns January 2, 2008 (1900+108)." +# This gimmick, semi-defensible only for arguments up to 99 and only in the pre-Y2K-awareness era, +# means that DATE(1899, 12, 31) is interpreted as 3799-12-31.

+# +#

For further information, please refer to the documentation for the xldate_* functions.

+# +#

Named references, constants, formulas, and macros

+# +#

+# A name is used to refer to a cell, a group of cells, a constant +# value, a formula, or a macro. Usually the scope of a name is global +# across the whole workbook. However it can be local to a worksheet. +# For example, if the sales figures are in different cells in +# different sheets, the user may define the name "Sales" in each +# sheet. There are built-in names, like "Print_Area" and +# "Print_Titles"; these two are naturally local to a sheet. +#

+# To inspect the names with a user interface like MS Excel, OOo Calc, +# or Gnumeric, click on Insert/Names/Define. This will show the global +# names, plus those local to the currently selected sheet. +#

+# A Book object provides two dictionaries (name_map and +# name_and_scope_map) and a list (name_obj_list) which allow various +# ways of accessing the Name objects. There is one Name object for +# each NAME record found in the workbook. Name objects have many +# attributes, several of which are relevant only when obj.macro is 1. +#

+# In the examples directory you will find namesdemo.xls which +# showcases the many different ways that names can be used, and +# xlrdnamesAPIdemo.py which offers 3 different queries for inspecting +# the names in your files, and shows how to extract whatever a name is +# referring to. There is currently one "convenience method", +# Name.cell(), which extracts the value in the case where the name +# refers to a single cell. More convenience methods are planned. The +# source code for Name.cell (in __init__.py) is an extra source of +# information on how the Name attributes hang together. +#

+# +#

Name information is not extracted from files older than +# Excel 5.0 (Book.biff_version < 50)

+# +#

Formatting

+# +#

Introduction

+# +#

This collection of features, new in xlrd version 0.6.1, is intended +# to provide the information needed to (1) display/render spreadsheet contents +# (say) on a screen or in a PDF file, and (2) copy spreadsheet data to another +# file without losing the ability to display/render it.

+# +#

The Palette; Colour Indexes

+# +#

A colour is represented in Excel as a (red, green, blue) ("RGB") tuple +# with each component in range(256). However it is not possible to access an +# unlimited number of colours; each spreadsheet is limited to a palette of 64 different +# colours (24 in Excel 3.0 and 4.0, 8 in Excel 2.0). Colours are referenced by an index +# ("colour index") into this palette. +# +# Colour indexes 0 to 7 represent 8 fixed built-in colours: black, white, red, green, blue, +# yellow, magenta, and cyan.

+# +# The remaining colours in the palette (8 to 63 in Excel 5.0 and later) +# can be changed by the user. In the Excel 2003 UI, Tools/Options/Color presents a palette +# of 7 rows of 8 colours. The last two rows are reserved for use in charts.
+# The correspondence between this grid and the assigned +# colour indexes is NOT left-to-right top-to-bottom.
+# Indexes 8 to 15 correspond to changeable +# parallels of the 8 fixed colours -- for example, index 7 is forever cyan; +# index 15 starts off being cyan but can be changed by the user.
+# +# The default colour for each index depends on the file version; tables of the defaults +# are available in the source code. If the user changes one or more colours, +# a PALETTE record appears in the XLS file -- it gives the RGB values for *all* changeable +# indexes.
+# Note that colours can be used in "number formats": "[CYAN]...." and "[COLOR8]...." refer +# to colour index 7; "[COLOR16]...." will produce cyan +# unless the user changes colour index 15 to something else.
+# +#

In addition, there are several "magic" colour indexes used by Excel:
+# 0x18 (BIFF3-BIFF4), 0x40 (BIFF5-BIFF8): System window text colour for border lines +# (used in XF, CF, and WINDOW2 records)
+# 0x19 (BIFF3-BIFF4), 0x41 (BIFF5-BIFF8): System window background colour for pattern background +# (used in XF and CF records )
+# 0x43: System face colour (dialogue background colour)
+# 0x4D: System window text colour for chart border lines
+# 0x4E: System window background colour for chart areas
+# 0x4F: Automatic colour for chart border lines (seems to be always Black)
+# 0x50: System ToolTip background colour (used in note objects)
+# 0x51: System ToolTip text colour (used in note objects)
+# 0x7FFF: System window text colour for fonts (used in FONT and CF records)
+# Note 0x7FFF appears to be the *default* colour index. It appears quite often in FONT +# records.
+# +#

Default Formatting

+# +# Default formatting is applied to all empty cells (those not described by a cell record). +# Firstly row default information (ROW record, Rowinfo class) is used if available. +# Failing that, column default information (COLINFO record, Colinfo class) is used if available. +# As a last resort the worksheet/workbook default cell format will be used; this +# should always be present in an Excel file, +# described by the XF record with the fixed index 15 (0-based). By default, it uses the +# worksheet/workbook default cell style, described by the very first XF record (index 0). +# +#

Formatting features not included in xlrd version 0.6.1

+# +# +#

Loading worksheets on demand

+# +#

This feature, new in version 0.7.1, is governed by the on_demand argument +# to the open_workbook() function and allows saving memory and time by loading +# only those sheets that the caller is interested in, and releasing sheets +# when no longer required.

+# +#

on_demand=False (default): No change. open_workbook() loads global data +# and all sheets, releases resources no longer required (principally the +# str or mmap object containing the Workbook stream), and returns.

+# +#

on_demand=True and BIFF version < 5.0: A warning message is emitted, +# on_demand is recorded as False, and the old process is followed.

+# +#

on_demand=True and BIFF version >= 5.0: open_workbook() loads global +# data and returns without releasing resources. At this stage, the only +# information available about sheets is Book.nsheets and Book.sheet_names().

+# +#

Book.sheet_by_name() and Book.sheet_by_index() will load the requested +# sheet if it is not already loaded.

+# +#

Book.sheets() will load all/any unloaded sheets.

+# +#

The caller may save memory by calling +# Book.unload_sheet(sheet_name_or_index) when finished with the sheet. +# This applies irrespective of the state of on_demand.

+# +#

The caller may re-load an unloaded sheet by calling Book.sheet_by_xxxx() +# -- except if those required resources have been released (which will +# have happened automatically when on_demand is false). This is the only +# case where an exception will be raised.

+# +#

The caller may query the state of a sheet: +# Book.sheet_loaded(sheet_name_or_index) -> a bool

+# +#

Book.release_resources() may used to save memory and close +# any memory-mapped file before proceding to examine already-loaded +# sheets. Once resources are released, no further sheets can be loaded.

+# +#

When using on-demand, it is advisable to ensure that +# Book.release_resources() is always called even if an exception +# is raised in your own code; otherwise if the input file has been +# memory-mapped, the mmap.mmap object will not be closed and you will +# not be able to access the physical file until your Python process +# terminates. This can be done by calling Book.release_resources() +# explicitly in the finally suite of a try/finally block. +# New in xlrd 0.7.2: the Book object is a "context manager", so if +# using Python 2.5 or later, you can wrap your code in a "with" +# statement.

+## + +import sys, zipfile, pprint +from . import timemachine +from .biffh import ( + XLRDError, + biff_text_from_num, + error_text_from_code, + XL_CELL_BLANK, + XL_CELL_TEXT, + XL_CELL_BOOLEAN, + XL_CELL_ERROR, + XL_CELL_EMPTY, + XL_CELL_DATE, + XL_CELL_NUMBER + ) +from .formula import * # is constrained by __all__ +from .book import Book, colname #### TODO #### formula also has `colname` (restricted to 256 cols) +from .sheet import empty_cell +from .xldate import XLDateError, xldate_as_tuple + +if sys.version.startswith("IronPython"): + # print >> sys.stderr, "...importing encodings" + import encodings + +try: + import mmap + MMAP_AVAILABLE = 1 +except ImportError: + MMAP_AVAILABLE = 0 +USE_MMAP = MMAP_AVAILABLE + +## +# +# Open a spreadsheet file for data extraction. +# +# @param filename The path to the spreadsheet file to be opened. +# +# @param logfile An open file to which messages and diagnostics are written. +# +# @param verbosity Increases the volume of trace material written to the logfile. +# +# @param use_mmap Whether to use the mmap module is determined heuristically. +# Use this arg to override the result. Current heuristic: mmap is used if it exists. +# +# @param file_contents ... as a string or an mmap.mmap object or some other behave-alike object. +# If file_contents is supplied, filename will not be used, except (possibly) in messages. +# +# @param encoding_override Used to overcome missing or bad codepage information +# in older-version files. Refer to discussion in the Unicode section above. +#
-- New in version 0.6.0 +# +# @param formatting_info Governs provision of a reference to an XF (eXtended Format) object +# for each cell in the worksheet. +#
Default is False. This is backwards compatible and saves memory. +# "Blank" cells (those with their own formatting information but no data) are treated as empty +# (by ignoring the file's BLANK and MULBLANK records). +# It cuts off any bottom "margin" of rows of empty (and blank) cells and +# any right "margin" of columns of empty (and blank) cells. +# Only cell_value and cell_type are available. +#
True provides all cells, including empty and blank cells. +# XF information is available for each cell. +#
-- New in version 0.6.1 +# +# @param on_demand Governs whether sheets are all loaded initially or when demanded +# by the caller. Please refer back to the section "Loading worksheets on demand" for details. +#
-- New in version 0.7.1 +# +# @param ragged_rows False (the default) means all rows are padded out with empty cells so that all +# rows have the same size (Sheet.ncols). True means that there are no empty cells at the ends of rows. +# This can result in substantial memory savings if rows are of widely varying sizes. See also the +# Sheet.row_len() method. +#
-- New in version 0.7.2 +# +# @return An instance of the Book class. + +def open_workbook(filename=None, + logfile=sys.stdout, + verbosity=0, + use_mmap=USE_MMAP, + file_contents=None, + encoding_override=None, + formatting_info=False, + on_demand=False, + ragged_rows=False, + ): + peeksz = 4 + if file_contents: + peek = file_contents[:peeksz] + else: + f = open(filename, "rb") + peek = f.read(peeksz) + f.close() + if peek == b"PK\x03\x04": # a ZIP file + if file_contents: + zf = zipfile.ZipFile(timemachine.BYTES_IO(file_contents)) + else: + zf = zipfile.ZipFile(filename) + + # Workaround for some third party files that use forward slashes and + # lower case names. We map the expected name in lowercase to the + # actual filename in the zip container. + component_names = dict([(name.replace('\\', '/').lower(), name) + for name in zf.namelist()]) + + if verbosity: + logfile.write('ZIP component_names:\n') + pprint.pprint(component_names, logfile) + if 'xl/workbook.xml' in component_names: + from . import xlsx + bk = xlsx.open_workbook_2007_xml( + zf, + component_names, + logfile=logfile, + verbosity=verbosity, + use_mmap=use_mmap, + formatting_info=formatting_info, + on_demand=on_demand, + ragged_rows=ragged_rows, + ) + return bk + if 'xl/workbook.bin' in component_names: + raise XLRDError('Excel 2007 xlsb file; not supported') + if 'content.xml' in component_names: + raise XLRDError('Openoffice.org ODS file; not supported') + raise XLRDError('ZIP file contents not a known type of workbook') + + from . import book + bk = book.open_workbook_xls( + filename=filename, + logfile=logfile, + verbosity=verbosity, + use_mmap=use_mmap, + file_contents=file_contents, + encoding_override=encoding_override, + formatting_info=formatting_info, + on_demand=on_demand, + ragged_rows=ragged_rows, + ) + return bk + +## +# For debugging: dump an XLS file's BIFF records in char & hex. +# @param filename The path to the file to be dumped. +# @param outfile An open file, to which the dump is written. +# @param unnumbered If true, omit offsets (for meaningful diffs). + +def dump(filename, outfile=sys.stdout, unnumbered=False): + from .biffh import biff_dump + bk = Book() + bk.biff2_8_load(filename=filename, logfile=outfile, ) + biff_dump(bk.mem, bk.base, bk.stream_len, 0, outfile, unnumbered) + +## +# For debugging and analysis: summarise the file's BIFF records. +# I.e. produce a sorted file of (record_name, count). +# @param filename The path to the file to be summarised. +# @param outfile An open file, to which the summary is written. + +def count_records(filename, outfile=sys.stdout): + from .biffh import biff_count_records + bk = Book() + bk.biff2_8_load(filename=filename, logfile=outfile, ) + biff_count_records(bk.mem, bk.base, bk.stream_len, outfile) diff --git a/Assignment 3/packages/xlrd/biffh.py b/Assignment 3/packages/xlrd/biffh.py new file mode 100644 index 0000000..f3a6d4d --- /dev/null +++ b/Assignment 3/packages/xlrd/biffh.py @@ -0,0 +1,663 @@ +# -*- coding: cp1252 -*- + +## +# Support module for the xlrd package. +# +#

Portions copyright © 2005-2010 Stephen John Machin, Lingfo Pty Ltd

+#

This module is part of the xlrd package, which is released under a BSD-style licence.

+## + +# 2010-03-01 SJM Reading SCL record +# 2010-03-01 SJM Added more record IDs for biff_dump & biff_count +# 2008-02-10 SJM BIFF2 BLANK record +# 2008-02-08 SJM Preparation for Excel 2.0 support +# 2008-02-02 SJM Added suffixes (_B2, _B2_ONLY, etc) on record names for biff_dump & biff_count +# 2007-12-04 SJM Added support for Excel 2.x (BIFF2) files. +# 2007-09-08 SJM Avoid crash when zero-length Unicode string missing options byte. +# 2007-04-22 SJM Remove experimental "trimming" facility. + +from __future__ import print_function + +DEBUG = 0 + +from struct import unpack +import sys +from .timemachine import * + +class XLRDError(Exception): + pass + +## +# Parent of almost all other classes in the package. Defines a common "dump" method +# for debugging. + +class BaseObject(object): + + _repr_these = [] + + ## + # @param f open file object, to which the dump is written + # @param header text to write before the dump + # @param footer text to write after the dump + # @param indent number of leading spaces (for recursive calls) + + def dump(self, f=None, header=None, footer=None, indent=0): + if f is None: + f = sys.stderr + if hasattr(self, "__slots__"): + alist = [] + for attr in self.__slots__: + alist.append((attr, getattr(self, attr))) + else: + alist = self.__dict__.items() + alist = sorted(alist) + pad = " " * indent + if header is not None: print(header, file=f) + list_type = type([]) + dict_type = type({}) + for attr, value in alist: + if getattr(value, 'dump', None) and attr != 'book': + value.dump(f, + header="%s%s (%s object):" % (pad, attr, value.__class__.__name__), + indent=indent+4) + elif attr not in self._repr_these and ( + isinstance(value, list_type) or isinstance(value, dict_type) + ): + print("%s%s: %s, len = %d" % (pad, attr, type(value), len(value)), file=f) + else: + fprintf(f, "%s%s: %r\n", pad, attr, value) + if footer is not None: print(footer, file=f) + +FUN, FDT, FNU, FGE, FTX = range(5) # unknown, date, number, general, text +DATEFORMAT = FDT +NUMBERFORMAT = FNU + +( + XL_CELL_EMPTY, + XL_CELL_TEXT, + XL_CELL_NUMBER, + XL_CELL_DATE, + XL_CELL_BOOLEAN, + XL_CELL_ERROR, + XL_CELL_BLANK, # for use in debugging, gathering stats, etc +) = range(7) + +biff_text_from_num = { + 0: "(not BIFF)", + 20: "2.0", + 21: "2.1", + 30: "3", + 40: "4S", + 45: "4W", + 50: "5", + 70: "7", + 80: "8", + 85: "8X", + } + +## +#

This dictionary can be used to produce a text version of the internal codes +# that Excel uses for error cells. Here are its contents: +# 

+# 0x00: '#NULL!',  # Intersection of two cell ranges is empty
+# 0x07: '#DIV/0!', # Division by zero
+# 0x0F: '#VALUE!', # Wrong type of operand
+# 0x17: '#REF!',   # Illegal or deleted cell reference
+# 0x1D: '#NAME?',  # Wrong function or range name
+# 0x24: '#NUM!',   # Value range overflow
+# 0x2A: '#N/A',    # Argument or function not available
+#

+ +error_text_from_code = { + 0x00: '#NULL!', # Intersection of two cell ranges is empty + 0x07: '#DIV/0!', # Division by zero + 0x0F: '#VALUE!', # Wrong type of operand + 0x17: '#REF!', # Illegal or deleted cell reference + 0x1D: '#NAME?', # Wrong function or range name + 0x24: '#NUM!', # Value range overflow + 0x2A: '#N/A', # Argument or function not available +} + +BIFF_FIRST_UNICODE = 80 + +XL_WORKBOOK_GLOBALS = WBKBLOBAL = 0x5 +XL_WORKBOOK_GLOBALS_4W = 0x100 +XL_WORKSHEET = WRKSHEET = 0x10 + +XL_BOUNDSHEET_WORKSHEET = 0x00 +XL_BOUNDSHEET_CHART = 0x02 +XL_BOUNDSHEET_VB_MODULE = 0x06 + +# XL_RK2 = 0x7e +XL_ARRAY = 0x0221 +XL_ARRAY2 = 0x0021 +XL_BLANK = 0x0201 +XL_BLANK_B2 = 0x01 +XL_BOF = 0x809 +XL_BOOLERR = 0x205 +XL_BOOLERR_B2 = 0x5 +XL_BOUNDSHEET = 0x85 +XL_BUILTINFMTCOUNT = 0x56 +XL_CF = 0x01B1 +XL_CODEPAGE = 0x42 +XL_COLINFO = 0x7D +XL_COLUMNDEFAULT = 0x20 # BIFF2 only +XL_COLWIDTH = 0x24 # BIFF2 only +XL_CONDFMT = 0x01B0 +XL_CONTINUE = 0x3c +XL_COUNTRY = 0x8C +XL_DATEMODE = 0x22 +XL_DEFAULTROWHEIGHT = 0x0225 +XL_DEFCOLWIDTH = 0x55 +XL_DIMENSION = 0x200 +XL_DIMENSION2 = 0x0 +XL_EFONT = 0x45 +XL_EOF = 0x0a +XL_EXTERNNAME = 0x23 +XL_EXTERNSHEET = 0x17 +XL_EXTSST = 0xff +XL_FEAT11 = 0x872 +XL_FILEPASS = 0x2f +XL_FONT = 0x31 +XL_FONT_B3B4 = 0x231 +XL_FORMAT = 0x41e +XL_FORMAT2 = 0x1E # BIFF2, BIFF3 +XL_FORMULA = 0x6 +XL_FORMULA3 = 0x206 +XL_FORMULA4 = 0x406 +XL_GCW = 0xab +XL_HLINK = 0x01B8 +XL_QUICKTIP = 0x0800 +XL_HORIZONTALPAGEBREAKS = 0x1b +XL_INDEX = 0x20b +XL_INTEGER = 0x2 # BIFF2 only +XL_IXFE = 0x44 # BIFF2 only +XL_LABEL = 0x204 +XL_LABEL_B2 = 0x04 +XL_LABELRANGES = 0x15f +XL_LABELSST = 0xfd +XL_LEFTMARGIN = 0x26 +XL_TOPMARGIN = 0x28 +XL_RIGHTMARGIN = 0x27 +XL_BOTTOMMARGIN = 0x29 +XL_HEADER = 0x14 +XL_FOOTER = 0x15 +XL_HCENTER = 0x83 +XL_VCENTER = 0x84 +XL_MERGEDCELLS = 0xE5 +XL_MSO_DRAWING = 0x00EC +XL_MSO_DRAWING_GROUP = 0x00EB +XL_MSO_DRAWING_SELECTION = 0x00ED +XL_MULRK = 0xbd +XL_MULBLANK = 0xbe +XL_NAME = 0x18 +XL_NOTE = 0x1c +XL_NUMBER = 0x203 +XL_NUMBER_B2 = 0x3 +XL_OBJ = 0x5D +XL_PAGESETUP = 0xA1 +XL_PALETTE = 0x92 +XL_PANE = 0x41 +XL_PRINTGRIDLINES = 0x2B +XL_PRINTHEADERS = 0x2A +XL_RK = 0x27e +XL_ROW = 0x208 +XL_ROW_B2 = 0x08 +XL_RSTRING = 0xd6 +XL_SCL = 0x00A0 +XL_SHEETHDR = 0x8F # BIFF4W only +XL_SHEETPR = 0x81 +XL_SHEETSOFFSET = 0x8E # BIFF4W only +XL_SHRFMLA = 0x04bc +XL_SST = 0xfc +XL_STANDARDWIDTH = 0x99 +XL_STRING = 0x207 +XL_STRING_B2 = 0x7 +XL_STYLE = 0x293 +XL_SUPBOOK = 0x1AE # aka EXTERNALBOOK in OOo docs +XL_TABLEOP = 0x236 +XL_TABLEOP2 = 0x37 +XL_TABLEOP_B2 = 0x36 +XL_TXO = 0x1b6 +XL_UNCALCED = 0x5e +XL_UNKNOWN = 0xffff +XL_VERTICALPAGEBREAKS = 0x1a +XL_WINDOW2 = 0x023E +XL_WINDOW2_B2 = 0x003E +XL_WRITEACCESS = 0x5C +XL_WSBOOL = XL_SHEETPR +XL_XF = 0xe0 +XL_XF2 = 0x0043 # BIFF2 version of XF record +XL_XF3 = 0x0243 # BIFF3 version of XF record +XL_XF4 = 0x0443 # BIFF4 version of XF record + +boflen = {0x0809: 8, 0x0409: 6, 0x0209: 6, 0x0009: 4} +bofcodes = (0x0809, 0x0409, 0x0209, 0x0009) + +XL_FORMULA_OPCODES = (0x0006, 0x0406, 0x0206) + +_cell_opcode_list = [ + XL_BOOLERR, + XL_FORMULA, + XL_FORMULA3, + XL_FORMULA4, + XL_LABEL, + XL_LABELSST, + XL_MULRK, + XL_NUMBER, + XL_RK, + XL_RSTRING, + ] +_cell_opcode_dict = {} +for _cell_opcode in _cell_opcode_list: + _cell_opcode_dict[_cell_opcode] = 1 + +def is_cell_opcode(c): + return c in _cell_opcode_dict + +def upkbits(tgt_obj, src, manifest, local_setattr=setattr): + for n, mask, attr in manifest: + local_setattr(tgt_obj, attr, (src & mask) >> n) + +def upkbitsL(tgt_obj, src, manifest, local_setattr=setattr, local_int=int): + for n, mask, attr in manifest: + local_setattr(tgt_obj, attr, local_int((src & mask) >> n)) + +def unpack_string(data, pos, encoding, lenlen=1): + nchars = unpack('<' + 'BH'[lenlen-1], data[pos:pos+lenlen])[0] + pos += lenlen + return unicode(data[pos:pos+nchars], encoding) + +def unpack_string_update_pos(data, pos, encoding, lenlen=1, known_len=None): + if known_len is not None: + # On a NAME record, the length byte is detached from the front of the string. + nchars = known_len + else: + nchars = unpack('<' + 'BH'[lenlen-1], data[pos:pos+lenlen])[0] + pos += lenlen + newpos = pos + nchars + return (unicode(data[pos:newpos], encoding), newpos) + +def unpack_unicode(data, pos, lenlen=2): + "Return unicode_strg" + nchars = unpack('<' + 'BH'[lenlen-1], data[pos:pos+lenlen])[0] + if not nchars: + # Ambiguous whether 0-length string should have an "options" byte. + # Avoid crash if missing. + return UNICODE_LITERAL("") + pos += lenlen + options = BYTES_ORD(data[pos]) + pos += 1 + # phonetic = options & 0x04 + # richtext = options & 0x08 + if options & 0x08: + # rt = unpack(' endpos=%d pos=%d endsub=%d substrg=%r\n', + ofs, dlen, base, endpos, pos, endsub, substrg) + break + hexd = ''.join(["%02x " % BYTES_ORD(c) for c in substrg]) + + chard = '' + for c in substrg: + c = chr(BYTES_ORD(c)) + if c == '\0': + c = '~' + elif not (' ' <= c <= '~'): + c = '?' + chard += c + if numbered: + num_prefix = "%5d: " % (base+pos-ofs) + + fprintf(fout, "%s %-48s %s\n", num_prefix, hexd, chard) + pos = endsub + +def biff_dump(mem, stream_offset, stream_len, base=0, fout=sys.stdout, unnumbered=False): + pos = stream_offset + stream_end = stream_offset + stream_len + adj = base - stream_offset + dummies = 0 + numbered = not unnumbered + num_prefix = '' + while stream_end - pos >= 4: + rc, length = unpack('') + if numbered: + num_prefix = "%5d: " % (adj + pos) + fprintf(fout, "%s%04x %s len = %04x (%d)\n", num_prefix, rc, recname, length, length) + pos += 4 + hex_char_dump(mem, pos, length, adj+pos, fout, unnumbered) + pos += length + if dummies: + if numbered: + num_prefix = "%5d: " % (adj + savpos) + fprintf(fout, "%s---- %d zero bytes skipped ----\n", num_prefix, dummies) + if pos < stream_end: + if numbered: + num_prefix = "%5d: " % (adj + pos) + fprintf(fout, "%s---- Misc bytes at end ----\n", num_prefix) + hex_char_dump(mem, pos, stream_end-pos, adj + pos, fout, unnumbered) + elif pos > stream_end: + fprintf(fout, "Last dumped record has length (%d) that is too large\n", length) + +def biff_count_records(mem, stream_offset, stream_len, fout=sys.stdout): + pos = stream_offset + stream_end = stream_offset + stream_len + tally = {} + while stream_end - pos >= 4: + rc, length = unpack('> sys.stderr, "...importing encodings" + import encodings + +empty_cell = sheet.empty_cell # for exposure to the world ... + +DEBUG = 0 + +USE_FANCY_CD = 1 + +TOGGLE_GC = 0 +import gc +# gc.set_debug(gc.DEBUG_STATS) + +try: + import mmap + MMAP_AVAILABLE = 1 +except ImportError: + MMAP_AVAILABLE = 0 +USE_MMAP = MMAP_AVAILABLE + +MY_EOF = 0xF00BAAA # not a 16-bit number + +SUPBOOK_UNK, SUPBOOK_INTERNAL, SUPBOOK_EXTERNAL, SUPBOOK_ADDIN, SUPBOOK_DDEOLE = range(5) + +SUPPORTED_VERSIONS = (80, 70, 50, 45, 40, 30, 21, 20) + +_code_from_builtin_name = { + "Consolidate_Area": "\x00", + "Auto_Open": "\x01", + "Auto_Close": "\x02", + "Extract": "\x03", + "Database": "\x04", + "Criteria": "\x05", + "Print_Area": "\x06", + "Print_Titles": "\x07", + "Recorder": "\x08", + "Data_Form": "\x09", + "Auto_Activate": "\x0A", + "Auto_Deactivate": "\x0B", + "Sheet_Title": "\x0C", + "_FilterDatabase": "\x0D", + } +builtin_name_from_code = {} +code_from_builtin_name = {} +for _bin, _bic in _code_from_builtin_name.items(): + _bin = UNICODE_LITERAL(_bin) + _bic = UNICODE_LITERAL(_bic) + code_from_builtin_name[_bin] = _bic + builtin_name_from_code[_bic] = _bin +del _bin, _bic, _code_from_builtin_name + +def open_workbook_xls(filename=None, + logfile=sys.stdout, verbosity=0, use_mmap=USE_MMAP, + file_contents=None, + encoding_override=None, + formatting_info=False, on_demand=False, ragged_rows=False, + ): + t0 = time.clock() + if TOGGLE_GC: + orig_gc_enabled = gc.isenabled() + if orig_gc_enabled: + gc.disable() + bk = Book() + try: + bk.biff2_8_load( + filename=filename, file_contents=file_contents, + logfile=logfile, verbosity=verbosity, use_mmap=use_mmap, + encoding_override=encoding_override, + formatting_info=formatting_info, + on_demand=on_demand, + ragged_rows=ragged_rows, + ) + t1 = time.clock() + bk.load_time_stage_1 = t1 - t0 + biff_version = bk.getbof(XL_WORKBOOK_GLOBALS) + if not biff_version: + raise XLRDError("Can't determine file's BIFF version") + if biff_version not in SUPPORTED_VERSIONS: + raise XLRDError( + "BIFF version %s is not supported" + % biff_text_from_num[biff_version] + ) + bk.biff_version = biff_version + if biff_version <= 40: + # no workbook globals, only 1 worksheet + if on_demand: + fprintf(bk.logfile, + "*** WARNING: on_demand is not supported for this Excel version.\n" + "*** Setting on_demand to False.\n") + bk.on_demand = on_demand = False + bk.fake_globals_get_sheet() + elif biff_version == 45: + # worksheet(s) embedded in global stream + bk.parse_globals() + if on_demand: + fprintf(bk.logfile, "*** WARNING: on_demand is not supported for this Excel version.\n" + "*** Setting on_demand to False.\n") + bk.on_demand = on_demand = False + else: + bk.parse_globals() + bk._sheet_list = [None for sh in bk._sheet_names] + if not on_demand: + bk.get_sheets() + bk.nsheets = len(bk._sheet_list) + if biff_version == 45 and bk.nsheets > 1: + fprintf(bk.logfile, + "*** WARNING: Excel 4.0 workbook (.XLW) file contains %d worksheets.\n" + "*** Book-level data will be that of the last worksheet.\n", + bk.nsheets + ) + if TOGGLE_GC: + if orig_gc_enabled: + gc.enable() + t2 = time.clock() + bk.load_time_stage_2 = t2 - t1 + except: + bk.release_resources() + raise + # normal exit + if not on_demand: + bk.release_resources() + return bk + +## +# For debugging: dump the file's BIFF records in char & hex. +# @param filename The path to the file to be dumped. +# @param outfile An open file, to which the dump is written. +# @param unnumbered If true, omit offsets (for meaningful diffs). + +def dump(filename, outfile=sys.stdout, unnumbered=False): + bk = Book() + bk.biff2_8_load(filename=filename, logfile=outfile, ) + biff_dump(bk.mem, bk.base, bk.stream_len, 0, outfile, unnumbered) + +## +# For debugging and analysis: summarise the file's BIFF records. +# I.e. produce a sorted file of (record_name, count). +# @param filename The path to the file to be summarised. +# @param outfile An open file, to which the summary is written. + +def count_records(filename, outfile=sys.stdout): + bk = Book() + bk.biff2_8_load(filename=filename, logfile=outfile, ) + biff_count_records(bk.mem, bk.base, bk.stream_len, outfile) + +## +# Information relating to a named reference, formula, macro, etc. +#
-- New in version 0.6.0 +#
-- Name information is not extracted from files older than +# Excel 5.0 (Book.biff_version < 50) + +class Name(BaseObject): + + _repr_these = ['stack'] + book = None # parent + + ## + # 0 = Visible; 1 = Hidden + hidden = 0 + + ## + # 0 = Command macro; 1 = Function macro. Relevant only if macro == 1 + func = 0 + + ## + # 0 = Sheet macro; 1 = VisualBasic macro. Relevant only if macro == 1 + vbasic = 0 + + ## + # 0 = Standard name; 1 = Macro name + macro = 0 + + ## + # 0 = Simple formula; 1 = Complex formula (array formula or user defined)
+ # No examples have been sighted. + complex = 0 + + ## + # 0 = User-defined name; 1 = Built-in name + # (common examples: Print_Area, Print_Titles; see OOo docs for full list) + builtin = 0 + + ## + # Function group. Relevant only if macro == 1; see OOo docs for values. + funcgroup = 0 + + ## + # 0 = Formula definition; 1 = Binary data
No examples have been sighted. + binary = 0 + + ## + # The index of this object in book.name_obj_list + name_index = 0 + + ## + # A Unicode string. If builtin, decoded as per OOo docs. + name = UNICODE_LITERAL("") + + ## + # An 8-bit string. + raw_formula = b'' + + ## + # -1: The name is global (visible in all calculation sheets).
+ # -2: The name belongs to a macro sheet or VBA sheet.
+ # -3: The name is invalid.
+ # 0 <= scope < book.nsheets: The name is local to the sheet whose index is scope. + scope = -1 + + ## + # The result of evaluating the formula, if any. + # If no formula, or evaluation of the formula encountered problems, + # the result is None. Otherwise the result is a single instance of the + # Operand class. + # + result = None + + ## + # This is a convenience method for the frequent use case where the name + # refers to a single cell. + # @return An instance of the Cell class. + # @throws XLRDError The name is not a constant absolute reference + # to a single cell. + def cell(self): + res = self.result + if res: + # result should be an instance of the Operand class + kind = res.kind + value = res.value + if kind == oREF and len(value) == 1: + ref3d = value[0] + if (0 <= ref3d.shtxlo == ref3d.shtxhi - 1 + and ref3d.rowxlo == ref3d.rowxhi - 1 + and ref3d.colxlo == ref3d.colxhi - 1): + sh = self.book.sheet_by_index(ref3d.shtxlo) + return sh.cell(ref3d.rowxlo, ref3d.colxlo) + self.dump(self.book.logfile, + header="=== Dump of Name object ===", + footer="======= End of dump =======", + ) + raise XLRDError("Not a constant absolute reference to a single cell") + + ## + # This is a convenience method for the use case where the name + # refers to one rectangular area in one worksheet. + # @param clipped If true (the default), the returned rectangle is clipped + # to fit in (0, sheet.nrows, 0, sheet.ncols) -- it is guaranteed that + # 0 <= rowxlo <= rowxhi <= sheet.nrows and that the number of usable rows + # in the area (which may be zero) is rowxhi - rowxlo; likewise for columns. + # @return a tuple (sheet_object, rowxlo, rowxhi, colxlo, colxhi). + # @throws XLRDError The name is not a constant absolute reference + # to a single area in a single sheet. + def area2d(self, clipped=True): + res = self.result + if res: + # result should be an instance of the Operand class + kind = res.kind + value = res.value + if kind == oREF and len(value) == 1: # only 1 reference + ref3d = value[0] + if 0 <= ref3d.shtxlo == ref3d.shtxhi - 1: # only 1 usable sheet + sh = self.book.sheet_by_index(ref3d.shtxlo) + if not clipped: + return sh, ref3d.rowxlo, ref3d.rowxhi, ref3d.colxlo, ref3d.colxhi + rowxlo = min(ref3d.rowxlo, sh.nrows) + rowxhi = max(rowxlo, min(ref3d.rowxhi, sh.nrows)) + colxlo = min(ref3d.colxlo, sh.ncols) + colxhi = max(colxlo, min(ref3d.colxhi, sh.ncols)) + assert 0 <= rowxlo <= rowxhi <= sh.nrows + assert 0 <= colxlo <= colxhi <= sh.ncols + return sh, rowxlo, rowxhi, colxlo, colxhi + self.dump(self.book.logfile, + header="=== Dump of Name object ===", + footer="======= End of dump =======", + ) + raise XLRDError("Not a constant absolute reference to a single area in a single sheet") + +## +# Contents of a "workbook". +#

WARNING: You don't call this class yourself. You use the Book object that +# was returned when you called xlrd.open_workbook("myfile.xls").

+ +class Book(BaseObject): + + ## + # The number of worksheets present in the workbook file. + # This information is available even when no sheets have yet been loaded. + nsheets = 0 + + ## + # Which date system was in force when this file was last saved.
+ # 0 => 1900 system (the Excel for Windows default).
+ # 1 => 1904 system (the Excel for Macintosh default).
+ datemode = 0 # In case it's not specified in the file. + + ## + # Version of BIFF (Binary Interchange File Format) used to create the file. + # Latest is 8.0 (represented here as 80), introduced with Excel 97. + # Earliest supported by this module: 2.0 (represented as 20). + biff_version = 0 + + ## + # List containing a Name object for each NAME record in the workbook. + #
-- New in version 0.6.0 + name_obj_list = [] + + ## + # An integer denoting the character set used for strings in this file. + # For BIFF 8 and later, this will be 1200, meaning Unicode; more precisely, UTF_16_LE. + # For earlier versions, this is used to derive the appropriate Python encoding + # to be used to convert to Unicode. + # Examples: 1252 -> 'cp1252', 10000 -> 'mac_roman' + codepage = None + + ## + # The encoding that was derived from the codepage. + encoding = None + + ## + # A tuple containing the (telephone system) country code for:
+ # [0]: the user-interface setting when the file was created.
+ # [1]: the regional settings.
+ # Example: (1, 61) meaning (USA, Australia). + # This information may give a clue to the correct encoding for an unknown codepage. + # For a long list of observed values, refer to the OpenOffice.org documentation for + # the COUNTRY record. + countries = (0, 0) + + ## + # What (if anything) is recorded as the name of the last user to save the file. + user_name = UNICODE_LITERAL('') + + ## + # A list of Font class instances, each corresponding to a FONT record. + #
-- New in version 0.6.1 + font_list = [] + + ## + # A list of XF class instances, each corresponding to an XF record. + #
-- New in version 0.6.1 + xf_list = [] + + ## + # A list of Format objects, each corresponding to a FORMAT record, in + # the order that they appear in the input file. + # It does not contain builtin formats. + # If you are creating an output file using (for example) pyExcelerator, + # use this list. + # The collection to be used for all visual rendering purposes is format_map. + #
-- New in version 0.6.1 + format_list = [] + + ## + # The mapping from XF.format_key to Format object. + #
-- New in version 0.6.1 + format_map = {} + + ## + # This provides access via name to the extended format information for + # both built-in styles and user-defined styles.
+ # It maps name to (built_in, xf_index), where:
+ # name is either the name of a user-defined style, + # or the name of one of the built-in styles. Known built-in names are + # Normal, RowLevel_1 to RowLevel_7, + # ColLevel_1 to ColLevel_7, Comma, Currency, Percent, "Comma [0]", + # "Currency [0]", Hyperlink, and "Followed Hyperlink".
+ # built_in 1 = built-in style, 0 = user-defined
+ # xf_index is an index into Book.xf_list.
+ # References: OOo docs s6.99 (STYLE record); Excel UI Format/Style + #
-- New in version 0.6.1; since 0.7.4, extracted only if + # open_workbook(..., formatting_info=True) + style_name_map = {} + + ## + # This provides definitions for colour indexes. Please refer to the + # above section "The Palette; Colour Indexes" for an explanation + # of how colours are represented in Excel.
+ # Colour indexes into the palette map into (red, green, blue) tuples. + # "Magic" indexes e.g. 0x7FFF map to None. + # colour_map is what you need if you want to render cells on screen or in a PDF + # file. If you are writing an output XLS file, use palette_record. + #
-- New in version 0.6.1. Extracted only if open_workbook(..., formatting_info=True) + colour_map = {} + + ## + # If the user has changed any of the colours in the standard palette, the XLS + # file will contain a PALETTE record with 56 (16 for Excel 4.0 and earlier) + # RGB values in it, and this list will be e.g. [(r0, b0, g0), ..., (r55, b55, g55)]. + # Otherwise this list will be empty. This is what you need if you are + # writing an output XLS file. If you want to render cells on screen or in a PDF + # file, use colour_map. + #
-- New in version 0.6.1. Extracted only if open_workbook(..., formatting_info=True) + palette_record = [] + + ## + # Time in seconds to extract the XLS image as a contiguous string (or mmap equivalent). + load_time_stage_1 = -1.0 + + ## + # Time in seconds to parse the data from the contiguous string (or mmap equivalent). + load_time_stage_2 = -1.0 + + ## + # @return A list of all sheets in the book. + # All sheets not already loaded will be loaded. + def sheets(self): + for sheetx in xrange(self.nsheets): + if not self._sheet_list[sheetx]: + self.get_sheet(sheetx) + return self._sheet_list[:] + + ## + # @param sheetx Sheet index in range(nsheets) + # @return An object of the Sheet class + def sheet_by_index(self, sheetx): + return self._sheet_list[sheetx] or self.get_sheet(sheetx) + + ## + # @param sheet_name Name of sheet required + # @return An object of the Sheet class + def sheet_by_name(self, sheet_name): + try: + sheetx = self._sheet_names.index(sheet_name) + except ValueError: + raise XLRDError('No sheet named <%r>' % sheet_name) + return self.sheet_by_index(sheetx) + + ## + # @return A list of the names of all the worksheets in the workbook file. + # This information is available even when no sheets have yet been loaded. + def sheet_names(self): + return self._sheet_names[:] + + ## + # @param sheet_name_or_index Name or index of sheet enquired upon + # @return true if sheet is loaded, false otherwise + #
-- New in version 0.7.1 + def sheet_loaded(self, sheet_name_or_index): + if isinstance(sheet_name_or_index, int): + sheetx = sheet_name_or_index + else: + try: + sheetx = self._sheet_names.index(sheet_name_or_index) + except ValueError: + raise XLRDError('No sheet named <%r>' % sheet_name_or_index) + return bool(self._sheet_list[sheetx]) + + ## + # @param sheet_name_or_index Name or index of sheet to be unloaded. + #
-- New in version 0.7.1 + def unload_sheet(self, sheet_name_or_index): + if isinstance(sheet_name_or_index, int): + sheetx = sheet_name_or_index + else: + try: + sheetx = self._sheet_names.index(sheet_name_or_index) + except ValueError: + raise XLRDError('No sheet named <%r>' % sheet_name_or_index) + self._sheet_list[sheetx] = None + + ## + # This method has a dual purpose. You can call it to release + # memory-consuming objects and (possibly) a memory-mapped file + # (mmap.mmap object) when you have finished loading sheets in + # on_demand mode, but still require the Book object to examine the + # loaded sheets. It is also called automatically (a) when open_workbook + # raises an exception and (b) if you are using a "with" statement, when + # the "with" block is exited. Calling this method multiple times on the + # same object has no ill effect. + def release_resources(self): + self._resources_released = 1 + if hasattr(self.mem, "close"): + # must be a mmap.mmap object + self.mem.close() + self.mem = None + if hasattr(self.filestr, "close"): + self.filestr.close() + self.filestr = None + self._sharedstrings = None + self._rich_text_runlist_map = None + + def __enter__(self): + return self + + def __exit__(self, exc_type, exc_value, exc_tb): + self.release_resources() + # return false + + ## + # A mapping from (lower_case_name, scope) to a single Name object. + #
-- New in version 0.6.0 + name_and_scope_map = {} + + ## + # A mapping from lower_case_name to a list of Name objects. The list is + # sorted in scope order. Typically there will be one item (of global scope) + # in the list. + #
-- New in version 0.6.0 + name_map = {} + + def __init__(self): + self._sheet_list = [] + self._sheet_names = [] + self._sheet_visibility = [] # from BOUNDSHEET record + self.nsheets = 0 + self._sh_abs_posn = [] # sheet's absolute position in the stream + self._sharedstrings = [] + self._rich_text_runlist_map = {} + self.raw_user_name = False + self._sheethdr_count = 0 # BIFF 4W only + self.builtinfmtcount = -1 # unknown as yet. BIFF 3, 4S, 4W + self.initialise_format_info() + self._all_sheets_count = 0 # includes macro & VBA sheets + self._supbook_count = 0 + self._supbook_locals_inx = None + self._supbook_addins_inx = None + self._all_sheets_map = [] # maps an all_sheets index to a calc-sheets index (or -1) + self._externsheet_info = [] + self._externsheet_type_b57 = [] + self._extnsht_name_from_num = {} + self._sheet_num_from_name = {} + self._extnsht_count = 0 + self._supbook_types = [] + self._resources_released = 0 + self.addin_func_names = [] + self.name_obj_list = [] + self.colour_map = {} + self.palette_record = [] + self.xf_list = [] + self.style_name_map = {} + self.mem = b'' + self.filestr = b'' + + def biff2_8_load(self, filename=None, file_contents=None, + logfile=sys.stdout, verbosity=0, use_mmap=USE_MMAP, + encoding_override=None, + formatting_info=False, + on_demand=False, + ragged_rows=False, + ): + # DEBUG = 0 + self.logfile = logfile + self.verbosity = verbosity + self.use_mmap = use_mmap and MMAP_AVAILABLE + self.encoding_override = encoding_override + self.formatting_info = formatting_info + self.on_demand = on_demand + self.ragged_rows = ragged_rows + + if not file_contents: + with open(filename, "rb") as f: + f.seek(0, 2) # EOF + size = f.tell() + f.seek(0, 0) # BOF + if size == 0: + raise XLRDError("File size is 0 bytes") + if self.use_mmap: + self.filestr = mmap.mmap(f.fileno(), size, access=mmap.ACCESS_READ) + self.stream_len = size + else: + self.filestr = f.read() + self.stream_len = len(self.filestr) + else: + self.filestr = file_contents + self.stream_len = len(file_contents) + + self.base = 0 + if self.filestr[:8] != compdoc.SIGNATURE: + # got this one at the antique store + self.mem = self.filestr + else: + cd = compdoc.CompDoc(self.filestr, logfile=self.logfile) + if USE_FANCY_CD: + for qname in ['Workbook', 'Book']: + self.mem, self.base, self.stream_len = \ + cd.locate_named_stream(UNICODE_LITERAL(qname)) + if self.mem: break + else: + raise XLRDError("Can't find workbook in OLE2 compound document") + else: + for qname in ['Workbook', 'Book']: + self.mem = cd.get_named_stream(UNICODE_LITERAL(qname)) + if self.mem: break + else: + raise XLRDError("Can't find workbook in OLE2 compound document") + self.stream_len = len(self.mem) + del cd + if self.mem is not self.filestr: + if hasattr(self.filestr, "close"): + self.filestr.close() + self.filestr = b'' + self._position = self.base + if DEBUG: + print("mem: %s, base: %d, len: %d" % (type(self.mem), self.base, self.stream_len), file=self.logfile) + + def initialise_format_info(self): + # needs to be done once per sheet for BIFF 4W :-( + self.format_map = {} + self.format_list = [] + self.xfcount = 0 + self.actualfmtcount = 0 # number of FORMAT records seen so far + self._xf_index_to_xl_type_map = {0: XL_CELL_NUMBER} + self._xf_epilogue_done = 0 + self.xf_list = [] + self.font_list = [] + + def get2bytes(self): + pos = self._position + buff_two = self.mem[pos:pos+2] + lenbuff = len(buff_two) + self._position += lenbuff + if lenbuff < 2: + return MY_EOF + lo, hi = buff_two + return (BYTES_ORD(hi) << 8) | BYTES_ORD(lo) + + def get_record_parts(self): + pos = self._position + mem = self.mem + code, length = unpack('= 2: + fprintf(self.logfile, + "BOUNDSHEET: inx=%d vis=%r sheet_name=%r abs_posn=%d sheet_type=0x%02x\n", + self._all_sheets_count, visibility, sheet_name, abs_posn, sheet_type) + self._all_sheets_count += 1 + if sheet_type != XL_BOUNDSHEET_WORKSHEET: + self._all_sheets_map.append(-1) + descr = { + 1: 'Macro sheet', + 2: 'Chart', + 6: 'Visual Basic module', + }.get(sheet_type, 'UNKNOWN') + + if DEBUG or self.verbosity >= 1: + fprintf(self.logfile, + "NOTE *** Ignoring non-worksheet data named %r (type 0x%02x = %s)\n", + sheet_name, sheet_type, descr) + else: + snum = len(self._sheet_names) + self._all_sheets_map.append(snum) + self._sheet_names.append(sheet_name) + self._sh_abs_posn.append(abs_posn) + self._sheet_visibility.append(visibility) + self._sheet_num_from_name[sheet_name] = snum + + def handle_builtinfmtcount(self, data): + ### N.B. This count appears to be utterly useless. + # DEBUG = 1 + builtinfmtcount = unpack('= 2: + fprintf(self.logfile, "*** No CODEPAGE record; assuming 1200 (utf_16_le)\n") + else: + codepage = self.codepage + if codepage in encoding_from_codepage: + encoding = encoding_from_codepage[codepage] + elif 300 <= codepage <= 1999: + encoding = 'cp' + str(codepage) + else: + encoding = 'unknown_codepage_' + str(codepage) + if DEBUG or (self.verbosity and encoding != self.encoding) : + fprintf(self.logfile, "CODEPAGE: codepage %r -> encoding %r\n", codepage, encoding) + self.encoding = encoding + if self.codepage != 1200: # utf_16_le + # If we don't have a codec that can decode ASCII into Unicode, + # we're well & truly stuffed -- let the punter know ASAP. + try: + _unused = unicode(b'trial', self.encoding) + except BaseException as e: + fprintf(self.logfile, + "ERROR *** codepage %r -> encoding %r -> %s: %s\n", + self.codepage, self.encoding, type(e).__name__.split(".")[-1], e) + raise + if self.raw_user_name: + strg = unpack_string(self.user_name, 0, self.encoding, lenlen=1) + strg = strg.rstrip() + # if DEBUG: + # print "CODEPAGE: user name decoded from %r to %r" % (self.user_name, strg) + self.user_name = strg + self.raw_user_name = False + return self.encoding + + def handle_codepage(self, data): + # DEBUG = 0 + codepage = unpack('= 2 + if self.biff_version >= 80: + option_flags, other_info =unpack("= 1 + blah2 = DEBUG or self.verbosity >= 2 + if self.biff_version >= 80: + num_refs = unpack("= 2: + logf = self.logfile + fprintf(logf, "FILEPASS:\n") + hex_char_dump(data, 0, len(data), base=0, fout=logf) + if self.biff_version >= 80: + kind1, = unpack('= 2 + bv = self.biff_version + if bv < 50: + return + self.derive_encoding() + # print + # hex_char_dump(data, 0, len(data), fout=self.logfile) + ( + option_flags, kb_shortcut, name_len, fmla_len, extsht_index, sheet_index, + menu_text_len, description_text_len, help_topic_text_len, status_bar_text_len, + ) = unpack("> nshift) + + macro_flag = " M"[nobj.macro] + if bv < 80: + internal_name, pos = unpack_string_update_pos(data, 14, self.encoding, known_len=name_len) + else: + internal_name, pos = unpack_unicode_update_pos(data, 14, known_len=name_len) + nobj.extn_sheet_num = extsht_index + nobj.excel_sheet_index = sheet_index + nobj.scope = None # patched up in the names_epilogue() method + if blah: + fprintf( + self.logfile, + "NAME[%d]:%s oflags=%d, name_len=%d, fmla_len=%d, extsht_index=%d, sheet_index=%d, name=%r\n", + name_index, macro_flag, option_flags, name_len, + fmla_len, extsht_index, sheet_index, internal_name) + name = internal_name + if nobj.builtin: + name = builtin_name_from_code.get(name, "??Unknown??") + if blah: print(" builtin: %s" % name, file=self.logfile) + nobj.name = name + nobj.raw_formula = data[pos:] + nobj.basic_formula_len = fmla_len + nobj.evaluated = 0 + if blah: + nobj.dump( + self.logfile, + header="--- handle_name: name[%d] ---" % name_index, + footer="-------------------", + ) + + def names_epilogue(self): + blah = self.verbosity >= 2 + f = self.logfile + if blah: + print("+++++ names_epilogue +++++", file=f) + print("_all_sheets_map", REPR(self._all_sheets_map), file=f) + print("_extnsht_name_from_num", REPR(self._extnsht_name_from_num), file=f) + print("_sheet_num_from_name", REPR(self._sheet_num_from_name), file=f) + num_names = len(self.name_obj_list) + for namex in range(num_names): + nobj = self.name_obj_list[namex] + # Convert from excel_sheet_index to scope. + # This is done here because in BIFF7 and earlier, the + # BOUNDSHEET records (from which _all_sheets_map is derived) + # come after the NAME records. + if self.biff_version >= 80: + sheet_index = nobj.excel_sheet_index + if sheet_index == 0: + intl_sheet_index = -1 # global + elif 1 <= sheet_index <= len(self._all_sheets_map): + intl_sheet_index = self._all_sheets_map[sheet_index-1] + if intl_sheet_index == -1: # maps to a macro or VBA sheet + intl_sheet_index = -2 # valid sheet reference but not useful + else: + # huh? + intl_sheet_index = -3 # invalid + elif 50 <= self.biff_version <= 70: + sheet_index = nobj.extn_sheet_num + if sheet_index == 0: + intl_sheet_index = -1 # global + else: + sheet_name = self._extnsht_name_from_num[sheet_index] + intl_sheet_index = self._sheet_num_from_name.get(sheet_name, -2) + nobj.scope = intl_sheet_index + + for namex in range(num_names): + nobj = self.name_obj_list[namex] + # Parse the formula ... + if nobj.macro or nobj.binary: continue + if nobj.evaluated: continue + evaluate_name_formula(self, nobj, namex, blah=blah) + + if self.verbosity >= 2: + print("---------- name object dump ----------", file=f) + for namex in range(num_names): + nobj = self.name_obj_list[namex] + nobj.dump(f, header="--- name[%d] ---" % namex) + print("--------------------------------------", file=f) + # + # Build some dicts for access to the name objects + # + name_and_scope_map = {} # (name.lower(), scope): Name_object + name_map = {} # name.lower() : list of Name_objects (sorted in scope order) + for namex in range(num_names): + nobj = self.name_obj_list[namex] + name_lcase = nobj.name.lower() + key = (name_lcase, nobj.scope) + if key in name_and_scope_map and self.verbosity: + fprintf(f, 'Duplicate entry %r in name_and_scope_map\n', key) + name_and_scope_map[key] = nobj + sort_data = (nobj.scope, namex, nobj) + # namex (a temp unique ID) ensures the Name objects will not + # be compared (fatal in py3) + if name_lcase in name_map: + name_map[name_lcase].append(sort_data) + else: + name_map[name_lcase] = [sort_data] + for key in name_map.keys(): + alist = name_map[key] + alist.sort() + name_map[key] = [x[2] for x in alist] + self.name_and_scope_map = name_and_scope_map + self.name_map = name_map + + def handle_obj(self, data): + # Not doing much handling at all. + # Worrying about embedded (BOF ... EOF) substreams is done elsewhere. + # DEBUG = 1 + obj_type, obj_id = unpack(' handle_obj type=%d id=0x%08x" % (obj_type, obj_id) + + def handle_supbook(self, data): + # aka EXTERNALBOOK in OOo docs + self._supbook_types.append(None) + blah = DEBUG or self.verbosity >= 2 + if blah: + print("SUPBOOK:", file=self.logfile) + hex_char_dump(data, 0, len(data), fout=self.logfile) + num_sheets = unpack("= 2: + fprintf(self.logfile, "SST: unique strings: %d\n", uniquestrings) + while 1: + code, nb, data = self.get_record_parts_conditional(XL_CONTINUE) + if code is None: + break + nbt += nb + if DEBUG >= 2: + fprintf(self.logfile, "CONTINUE: adding %d bytes to SST -> %d\n", nb, nbt) + strlist.append(data) + self._sharedstrings, rt_runlist = unpack_SST_table(strlist, uniquestrings) + if self.formatting_info: + self._rich_text_runlist_map = rt_runlist + if DEBUG: + t1 = time.time() + print("SST processing took %.2f seconds" % (t1 - t0, ), file=self.logfile) + + def handle_writeaccess(self, data): + DEBUG = 0 + if self.biff_version < 80: + if not self.encoding: + self.raw_user_name = True + self.user_name = data + return + strg = unpack_string(data, 0, self.encoding, lenlen=1) + else: + strg = unpack_unicode(data, 0, lenlen=2) + if DEBUG: fprintf(self.logfile, "WRITEACCESS: %d bytes; raw=%s %r\n", len(data), self.raw_user_name, strg) + strg = strg.rstrip() + self.user_name = strg + + def parse_globals(self): + # DEBUG = 0 + # no need to position, just start reading (after the BOF) + formatting.initialise_book(self) + while 1: + rc, length, data = self.get_record_parts() + if DEBUG: print("parse_globals: record code is 0x%04x" % rc, file=self.logfile) + if rc == XL_SST: + self.handle_sst(data) + elif rc == XL_FONT or rc == XL_FONT_B3B4: + self.handle_font(data) + elif rc == XL_FORMAT: # XL_FORMAT2 is BIFF <= 3.0, can't appear in globals + self.handle_format(data) + elif rc == XL_XF: + self.handle_xf(data) + elif rc == XL_BOUNDSHEET: + self.handle_boundsheet(data) + elif rc == XL_DATEMODE: + self.handle_datemode(data) + elif rc == XL_CODEPAGE: + self.handle_codepage(data) + elif rc == XL_COUNTRY: + self.handle_country(data) + elif rc == XL_EXTERNNAME: + self.handle_externname(data) + elif rc == XL_EXTERNSHEET: + self.handle_externsheet(data) + elif rc == XL_FILEPASS: + self.handle_filepass(data) + elif rc == XL_WRITEACCESS: + self.handle_writeaccess(data) + elif rc == XL_SHEETSOFFSET: + self.handle_sheetsoffset(data) + elif rc == XL_SHEETHDR: + self.handle_sheethdr(data) + elif rc == XL_SUPBOOK: + self.handle_supbook(data) + elif rc == XL_NAME: + self.handle_name(data) + elif rc == XL_PALETTE: + self.handle_palette(data) + elif rc == XL_STYLE: + self.handle_style(data) + elif rc & 0xff == 9 and self.verbosity: + fprintf(self.logfile, "*** Unexpected BOF at posn %d: 0x%04x len=%d data=%r\n", + self._position - length - 4, rc, length, data) + elif rc == XL_EOF: + self.xf_epilogue() + self.names_epilogue() + self.palette_epilogue() + if not self.encoding: + self.derive_encoding() + if self.biff_version == 45: + # DEBUG = 0 + if DEBUG: print("global EOF: position", self._position, file=self.logfile) + # if DEBUG: + # pos = self._position - 4 + # print repr(self.mem[pos:pos+40]) + return + else: + # if DEBUG: + # print >> self.logfile, "parse_globals: ignoring record code 0x%04x" % rc + pass + + def read(self, pos, length): + data = self.mem[pos:pos+length] + self._position = pos + len(data) + return data + + def getbof(self, rqd_stream): + # DEBUG = 1 + # if DEBUG: print >> self.logfile, "getbof(): position", self._position + if DEBUG: print("reqd: 0x%04x" % rqd_stream, file=self.logfile) + def bof_error(msg): + raise XLRDError('Unsupported format, or corrupt file: ' + msg) + savpos = self._position + opcode = self.get2bytes() + if opcode == MY_EOF: + bof_error('Expected BOF record; met end of file') + if opcode not in bofcodes: + bof_error('Expected BOF record; found %r' % self.mem[savpos:savpos+8]) + length = self.get2bytes() + if length == MY_EOF: + bof_error('Incomplete BOF record[1]; met end of file') + if not (4 <= length <= 20): + bof_error( + 'Invalid length (%d) for BOF record type 0x%04x' + % (length, opcode)) + padding = b'\0' * max(0, boflen[opcode] - length) + data = self.read(self._position, length); + if DEBUG: fprintf(self.logfile, "\ngetbof(): data=%r\n", data) + if len(data) < length: + bof_error('Incomplete BOF record[2]; met end of file') + data += padding + version1 = opcode >> 8 + version2, streamtype = unpack('= 2: + print("BOF: op=0x%04x vers=0x%04x stream=0x%04x buildid=%d buildyr=%d -> BIFF%d" \ + % (opcode, version2, streamtype, build, year, version), file=self.logfile) + got_globals = streamtype == XL_WORKBOOK_GLOBALS or ( + version == 45 and streamtype == XL_WORKBOOK_GLOBALS_4W) + if (rqd_stream == XL_WORKBOOK_GLOBALS and got_globals) or streamtype == rqd_stream: + return version + if version < 50 and streamtype == XL_WORKSHEET: + return version + if version >= 50 and streamtype == 0x0100: + bof_error("Workspace file -- no spreadsheet data") + bof_error( + 'BOF not workbook/worksheet: op=0x%04x vers=0x%04x strm=0x%04x build=%d year=%d -> BIFF%d' \ + % (opcode, version2, streamtype, build, year, version) + ) + +# === helper functions + +def expand_cell_address(inrow, incol): + # Ref : OOo docs, "4.3.4 Cell Addresses in BIFF8" + outrow = inrow + if incol & 0x8000: + if outrow >= 32768: + outrow -= 65536 + relrow = 1 + else: + relrow = 0 + outcol = incol & 0xFF + if incol & 0x4000: + if outcol >= 128: + outcol -= 256 + relcol = 1 + else: + relcol = 0 + return outrow, outcol, relrow, relcol + +def colname(colx, _A2Z="ABCDEFGHIJKLMNOPQRSTUVWXYZ"): + assert colx >= 0 + name = UNICODE_LITERAL('') + while 1: + quot, rem = divmod(colx, 26) + name = _A2Z[rem] + name + if not quot: + return name + colx = quot - 1 + +def display_cell_address(rowx, colx, relrow, relcol): + if relrow: + rowpart = "(*%s%d)" % ("+-"[rowx < 0], abs(rowx)) + else: + rowpart = "$%d" % (rowx+1,) + if relcol: + colpart = "(*%s%d)" % ("+-"[colx < 0], abs(colx)) + else: + colpart = "$" + colname(colx) + return colpart + rowpart + +def unpack_SST_table(datatab, nstrings): + "Return list of strings" + datainx = 0 + ndatas = len(datatab) + data = datatab[0] + datalen = len(data) + pos = 8 + strings = [] + strappend = strings.append + richtext_runs = {} + local_unpack = unpack + local_min = min + local_BYTES_ORD = BYTES_ORD + latin_1 = "latin_1" + for _unused_i in xrange(nstrings): + nchars = local_unpack('> 1, charsneed) + rawstrg = data[pos:pos+2*charsavail] + # if DEBUG: print "SST U16: nchars=%d pos=%d rawstrg=%r" % (nchars, pos, rawstrg) + try: + accstrg += unicode(rawstrg, "utf_16_le") + except: + # print "SST U16: nchars=%d pos=%d rawstrg=%r" % (nchars, pos, rawstrg) + # Probable cause: dodgy data e.g. unfinished surrogate pair. + # E.g. file unicode2.xls in pyExcelerator's examples has cells containing + # unichr(i) for i in range(0x100000) + # so this will include 0xD800 etc + raise + pos += 2*charsavail + else: + # Note: this is COMPRESSED (not ASCII!) encoding!!! + charsavail = local_min(datalen - pos, charsneed) + rawstrg = data[pos:pos+charsavail] + # if DEBUG: print "SST CMPRSD: nchars=%d pos=%d rawstrg=%r" % (nchars, pos, rawstrg) + accstrg += unicode(rawstrg, latin_1) + pos += charsavail + charsgot += charsavail + if charsgot == nchars: + break + datainx += 1 + data = datatab[datainx] + datalen = len(data) + options = local_BYTES_ORD(data[0]) + pos = 1 + + if rtcount: + runs = [] + for runindex in xrange(rtcount): + if pos == datalen: + pos = 0 + datainx += 1 + data = datatab[datainx] + datalen = len(data) + runs.append(local_unpack("= datalen: + # adjust to correct position in next record + pos = pos - datalen + datainx += 1 + if datainx < ndatas: + data = datatab[datainx] + datalen = len(data) + else: + assert _unused_i == nstrings - 1 + strappend(accstrg) + return strings, richtext_runs diff --git a/Assignment 3/packages/xlrd/compdoc.py b/Assignment 3/packages/xlrd/compdoc.py new file mode 100644 index 0000000..e434e8e --- /dev/null +++ b/Assignment 3/packages/xlrd/compdoc.py @@ -0,0 +1,473 @@ +# -*- coding: cp1252 -*- + +## +# Implements the minimal functionality required +# to extract a "Workbook" or "Book" stream (as one big string) +# from an OLE2 Compound Document file. +#

Copyright � 2005-2012 Stephen John Machin, Lingfo Pty Ltd

+#

This module is part of the xlrd package, which is released under a BSD-style licence.

+## + +# No part of the content of this file was derived from the works of David Giffin. + +# 2008-11-04 SJM Avoid assertion error when -1 used instead of -2 for first_SID of empty SCSS [Frank Hoffsuemmer] +# 2007-09-08 SJM Warning message if sector sizes are extremely large. +# 2007-05-07 SJM Meaningful exception instead of IndexError if a SAT (sector allocation table) is corrupted. +# 2007-04-22 SJM Missing "<" in a struct.unpack call => can't open files on bigendian platforms. + +from __future__ import print_function +import sys +from struct import unpack +from .timemachine import * +import array + +## +# Magic cookie that should appear in the first 8 bytes of the file. +SIGNATURE = b"\xD0\xCF\x11\xE0\xA1\xB1\x1A\xE1" + +EOCSID = -2 +FREESID = -1 +SATSID = -3 +MSATSID = -4 +EVILSID = -5 + +class CompDocError(Exception): + pass + +class DirNode(object): + + def __init__(self, DID, dent, DEBUG=0, logfile=sys.stdout): + # dent is the 128-byte directory entry + self.DID = DID + self.logfile = logfile + (cbufsize, self.etype, self.colour, self.left_DID, self.right_DID, + self.root_DID) = \ + unpack(' 20: # allows for 2**20 bytes i.e. 1MB + print("WARNING: sector size (2**%d) is preposterous; assuming 512 and continuing ..." \ + % ssz, file=logfile) + ssz = 9 + if sssz > ssz: + print("WARNING: short stream sector size (2**%d) is preposterous; assuming 64 and continuing ..." \ + % sssz, file=logfile) + sssz = 6 + self.sec_size = sec_size = 1 << ssz + self.short_sec_size = 1 << sssz + if self.sec_size != 512 or self.short_sec_size != 64: + print("@@@@ sec_size=%d short_sec_size=%d" % (self.sec_size, self.short_sec_size), file=logfile) + ( + SAT_tot_secs, self.dir_first_sec_sid, _unused, self.min_size_std_stream, + SSAT_first_sec_sid, SSAT_tot_secs, + MSATX_first_sec_sid, MSATX_tot_secs, + # ) = unpack(' 1: + print('MSATX: sid=%d (0x%08X)' % (sid, sid), file=logfile) + if sid >= mem_data_secs: + msg = "MSAT extension: accessing sector %d but only %d in file" % (sid, mem_data_secs) + if DEBUG > 1: + print(msg, file=logfile) + break + raise CompDocError(msg) + elif sid < 0: + raise CompDocError("MSAT extension: invalid sector id: %d" % sid) + if seen[sid]: + raise CompDocError("MSAT corruption: seen[%d] == %d" % (sid, seen[sid])) + seen[sid] = 1 + actual_MSATX_sectors += 1 + if DEBUG and actual_MSATX_sectors > expected_MSATX_sectors: + print("[1]===>>>", mem_data_secs, nent, SAT_sectors_reqd, expected_MSATX_sectors, actual_MSATX_sectors, file=logfile) + offset = 512 + sec_size * sid + MSAT.extend(unpack(fmt, mem[offset:offset+sec_size])) + sid = MSAT.pop() # last sector id is sid of next sector in the chain + + if DEBUG and actual_MSATX_sectors != expected_MSATX_sectors: + print("[2]===>>>", mem_data_secs, nent, SAT_sectors_reqd, expected_MSATX_sectors, actual_MSATX_sectors, file=logfile) + if DEBUG: + print("MSAT: len =", len(MSAT), file=logfile) + dump_list(MSAT, 10, logfile) + # + # === build the SAT === + # + self.SAT = [] + actual_SAT_sectors = 0 + dump_again = 0 + for msidx in xrange(len(MSAT)): + msid = MSAT[msidx] + if msid in (FREESID, EOCSID): + # Specification: the MSAT array may be padded with trailing FREESID entries. + # Toleration: a FREESID or EOCSID entry anywhere in the MSAT array will be ignored. + continue + if msid >= mem_data_secs: + if not trunc_warned: + print("WARNING *** File is truncated, or OLE2 MSAT is corrupt!!", file=logfile) + print("INFO: Trying to access sector %d but only %d available" \ + % (msid, mem_data_secs), file=logfile) + trunc_warned = 1 + MSAT[msidx] = EVILSID + dump_again = 1 + continue + elif msid < -2: + raise CompDocError("MSAT: invalid sector id: %d" % msid) + if seen[msid]: + raise CompDocError("MSAT extension corruption: seen[%d] == %d" % (msid, seen[msid])) + seen[msid] = 2 + actual_SAT_sectors += 1 + if DEBUG and actual_SAT_sectors > SAT_sectors_reqd: + print("[3]===>>>", mem_data_secs, nent, SAT_sectors_reqd, expected_MSATX_sectors, actual_MSATX_sectors, actual_SAT_sectors, msid, file=logfile) + offset = 512 + sec_size * msid + self.SAT.extend(unpack(fmt, mem[offset:offset+sec_size])) + + if DEBUG: + print("SAT: len =", len(self.SAT), file=logfile) + dump_list(self.SAT, 10, logfile) + # print >> logfile, "SAT ", + # for i, s in enumerate(self.SAT): + # print >> logfile, "entry: %4d offset: %6d, next entry: %4d" % (i, 512 + sec_size * i, s) + # print >> logfile, "%d:%d " % (i, s), + print(file=logfile) + if DEBUG and dump_again: + print("MSAT: len =", len(MSAT), file=logfile) + dump_list(MSAT, 10, logfile) + for satx in xrange(mem_data_secs, len(self.SAT)): + self.SAT[satx] = EVILSID + print("SAT: len =", len(self.SAT), file=logfile) + dump_list(self.SAT, 10, logfile) + # + # === build the directory === + # + dbytes = self._get_stream( + self.mem, 512, self.SAT, self.sec_size, self.dir_first_sec_sid, + name="directory", seen_id=3) + dirlist = [] + did = -1 + for pos in xrange(0, len(dbytes), 128): + did += 1 + dirlist.append(DirNode(did, dbytes[pos:pos+128], 0, logfile)) + self.dirlist = dirlist + _build_family_tree(dirlist, 0, dirlist[0].root_DID) # and stand well back ... + if DEBUG: + for d in dirlist: + d.dump(DEBUG) + # + # === get the SSCS === + # + sscs_dir = self.dirlist[0] + assert sscs_dir.etype == 5 # root entry + if sscs_dir.first_SID < 0 or sscs_dir.tot_size == 0: + # Problem reported by Frank Hoffsuemmer: some software was + # writing -1 instead of -2 (EOCSID) for the first_SID + # when the SCCS was empty. Not having EOCSID caused assertion + # failure in _get_stream. + # Solution: avoid calling _get_stream in any case when the + # SCSS appears to be empty. + self.SSCS = "" + else: + self.SSCS = self._get_stream( + self.mem, 512, self.SAT, sec_size, sscs_dir.first_SID, + sscs_dir.tot_size, name="SSCS", seen_id=4) + # if DEBUG: print >> logfile, "SSCS", repr(self.SSCS) + # + # === build the SSAT === + # + self.SSAT = [] + if SSAT_tot_secs > 0 and sscs_dir.tot_size == 0: + print("WARNING *** OLE2 inconsistency: SSCS size is 0 but SSAT size is non-zero", file=logfile) + if sscs_dir.tot_size > 0: + sid = SSAT_first_sec_sid + nsecs = SSAT_tot_secs + while sid >= 0 and nsecs > 0: + if seen[sid]: + raise CompDocError("SSAT corruption: seen[%d] == %d" % (sid, seen[sid])) + seen[sid] = 5 + nsecs -= 1 + start_pos = 512 + sid * sec_size + news = list(unpack(fmt, mem[start_pos:start_pos+sec_size])) + self.SSAT.extend(news) + sid = self.SAT[sid] + if DEBUG: print("SSAT last sid %d; remaining sectors %d" % (sid, nsecs), file=logfile) + assert nsecs == 0 and sid == EOCSID + if DEBUG: + print("SSAT", file=logfile) + dump_list(self.SSAT, 10, logfile) + if DEBUG: + print("seen", file=logfile) + dump_list(seen, 20, logfile) + + def _get_stream(self, mem, base, sat, sec_size, start_sid, size=None, name='', seen_id=None): + # print >> self.logfile, "_get_stream", base, sec_size, start_sid, size + sectors = [] + s = start_sid + if size is None: + # nothing to check against + while s >= 0: + if seen_id is not None: + if self.seen[s]: + raise CompDocError("%s corruption: seen[%d] == %d" % (name, s, self.seen[s])) + self.seen[s] = seen_id + start_pos = base + s * sec_size + sectors.append(mem[start_pos:start_pos+sec_size]) + try: + s = sat[s] + except IndexError: + raise CompDocError( + "OLE2 stream %r: sector allocation table invalid entry (%d)" % + (name, s) + ) + assert s == EOCSID + else: + todo = size + while s >= 0: + if seen_id is not None: + if self.seen[s]: + raise CompDocError("%s corruption: seen[%d] == %d" % (name, s, self.seen[s])) + self.seen[s] = seen_id + start_pos = base + s * sec_size + grab = sec_size + if grab > todo: + grab = todo + todo -= grab + sectors.append(mem[start_pos:start_pos+grab]) + try: + s = sat[s] + except IndexError: + raise CompDocError( + "OLE2 stream %r: sector allocation table invalid entry (%d)" % + (name, s) + ) + assert s == EOCSID + if todo != 0: + fprintf(self.logfile, + "WARNING *** OLE2 stream %r: expected size %d, actual size %d\n", + name, size, size - todo) + + return b''.join(sectors) + + def _dir_search(self, path, storage_DID=0): + # Return matching DirNode instance, or None + head = path[0] + tail = path[1:] + dl = self.dirlist + for child in dl[storage_DID].children: + if dl[child].name.lower() == head.lower(): + et = dl[child].etype + if et == 2: + return dl[child] + if et == 1: + if not tail: + raise CompDocError("Requested component is a 'storage'") + return self._dir_search(tail, child) + dl[child].dump(1) + raise CompDocError("Requested stream is not a 'user stream'") + return None + + ## + # Interrogate the compound document's directory; return the stream as a string if found, otherwise + # return None. + # @param qname Name of the desired stream e.g. u'Workbook'. Should be in Unicode or convertible thereto. + + def get_named_stream(self, qname): + d = self._dir_search(qname.split("/")) + if d is None: + return None + if d.tot_size >= self.min_size_std_stream: + return self._get_stream( + self.mem, 512, self.SAT, self.sec_size, d.first_SID, + d.tot_size, name=qname, seen_id=d.DID+6) + else: + return self._get_stream( + self.SSCS, 0, self.SSAT, self.short_sec_size, d.first_SID, + d.tot_size, name=qname + " (from SSCS)", seen_id=None) + + ## + # Interrogate the compound document's directory. + # If the named stream is not found, (None, 0, 0) will be returned. + # If the named stream is found and is contiguous within the original byte sequence ("mem") + # used when the document was opened, + # then (mem, offset_to_start_of_stream, length_of_stream) is returned. + # Otherwise a new string is built from the fragments and (new_string, 0, length_of_stream) is returned. + # @param qname Name of the desired stream e.g. u'Workbook'. Should be in Unicode or convertible thereto. + + def locate_named_stream(self, qname): + d = self._dir_search(qname.split("/")) + if d is None: + return (None, 0, 0) + if d.tot_size > self.mem_data_len: + raise CompDocError("%r stream length (%d bytes) > file data size (%d bytes)" + % (qname, d.tot_size, self.mem_data_len)) + if d.tot_size >= self.min_size_std_stream: + result = self._locate_stream( + self.mem, 512, self.SAT, self.sec_size, d.first_SID, + d.tot_size, qname, d.DID+6) + if self.DEBUG: + print("\nseen", file=self.logfile) + dump_list(self.seen, 20, self.logfile) + return result + else: + return ( + self._get_stream( + self.SSCS, 0, self.SSAT, self.short_sec_size, d.first_SID, + d.tot_size, qname + " (from SSCS)", None), + 0, + d.tot_size + ) + + def _locate_stream(self, mem, base, sat, sec_size, start_sid, expected_stream_size, qname, seen_id): + # print >> self.logfile, "_locate_stream", base, sec_size, start_sid, expected_stream_size + s = start_sid + if s < 0: + raise CompDocError("_locate_stream: start_sid (%d) is -ve" % start_sid) + p = -99 # dummy previous SID + start_pos = -9999 + end_pos = -8888 + slices = [] + tot_found = 0 + found_limit = (expected_stream_size + sec_size - 1) // sec_size + while s >= 0: + if self.seen[s]: + print("_locate_stream(%s): seen" % qname, file=self.logfile); dump_list(self.seen, 20, self.logfile) + raise CompDocError("%s corruption: seen[%d] == %d" % (qname, s, self.seen[s])) + self.seen[s] = seen_id + tot_found += 1 + if tot_found > found_limit: + raise CompDocError( + "%s: size exceeds expected %d bytes; corrupt?" + % (qname, found_limit * sec_size) + ) # Note: expected size rounded up to higher sector + if s == p+1: + # contiguous sectors + end_pos += sec_size + else: + # start new slice + if p >= 0: + # not first time + slices.append((start_pos, end_pos)) + start_pos = base + s * sec_size + end_pos = start_pos + sec_size + p = s + s = sat[s] + assert s == EOCSID + assert tot_found == found_limit + # print >> self.logfile, "_locate_stream(%s): seen" % qname; dump_list(self.seen, 20, self.logfile) + if not slices: + # The stream is contiguous ... just what we like! + return (mem, start_pos, expected_stream_size) + slices.append((start_pos, end_pos)) + # print >> self.logfile, "+++>>> %d fragments" % len(slices) + return (b''.join([mem[start_pos:end_pos] for start_pos, end_pos in slices]), 0, expected_stream_size) + +# ========================================================================================== +def x_dump_line(alist, stride, f, dpos, equal=0): + print("%5d%s" % (dpos, " ="[equal]), end=' ', file=f) + for value in alist[dpos:dpos + stride]: + print(str(value), end=' ', file=f) + print(file=f) + +def dump_list(alist, stride, f=sys.stdout): + def _dump_line(dpos, equal=0): + print("%5d%s" % (dpos, " ="[equal]), end=' ', file=f) + for value in alist[dpos:dpos + stride]: + print(str(value), end=' ', file=f) + print(file=f) + pos = None + oldpos = None + for pos in xrange(0, len(alist), stride): + if oldpos is None: + _dump_line(pos) + oldpos = pos + elif alist[pos:pos+stride] != alist[oldpos:oldpos+stride]: + if pos - oldpos > stride: + _dump_line(pos - stride, equal=1) + _dump_line(pos) + oldpos = pos + if oldpos is not None and pos is not None and pos != oldpos: + _dump_line(pos, equal=1) diff --git a/Assignment 3/packages/xlrd/doc/compdoc.html b/Assignment 3/packages/xlrd/doc/compdoc.html new file mode 100644 index 0000000..8a1e173 --- /dev/null +++ b/Assignment 3/packages/xlrd/doc/compdoc.html @@ -0,0 +1,69 @@ + + + + +The compdoc Module + + +

The compdoc Module

+

Implements the minimal functionality required +to extract a "Workbook" or "Book" stream (as one big string) +from an OLE2 Compound Document file. +

Copyright © 2005-2012 Stephen John Machin, Lingfo Pty Ltd

+

This module is part of the xlrd package, which is released under a BSD-style licence.

+

Module Contents

+
+
CompDoc(mem, logfile=sys.stdout, DEBUG=0) (class) [#]
+
+

Compound document handler.

+
+
mem
+
+The raw contents of the file, as a string, or as an mmap.mmap() object. The +only operation it needs to support is slicing.
+

+

For more information about this class, see The CompDoc Class.

+
+
SIGNATURE (variable) [#]
+
+

Magic cookie that should appear in the first 8 bytes of the file.

+
+
+

The CompDoc Class

+
+
CompDoc(mem, logfile=sys.stdout, DEBUG=0) (class) [#]
+
+

Compound document handler.

+
+
mem
+
+The raw contents of the file, as a string, or as an mmap.mmap() object. The +only operation it needs to support is slicing.
+

+
+
get_named_stream(qname) [#]
+
+

Interrogate the compound document's directory; return the stream as a string if found, otherwise +return None.

+
+
qname
+
+Name of the desired stream e.g. u'Workbook'. Should be in Unicode or convertible thereto.
+

+
+
locate_named_stream(qname) [#]
+
+

Interrogate the compound document's directory. +If the named stream is not found, (None, 0, 0) will be returned. +If the named stream is found and is contiguous within the original byte sequence ("mem") +used when the document was opened, +then (mem, offset_to_start_of_stream, length_of_stream) is returned. +Otherwise a new string is built from the fragments and (new_string, 0, length_of_stream) is returned.

+
+
qname
+
+Name of the desired stream e.g. u'Workbook'. Should be in Unicode or convertible thereto.
+

+
+
+ diff --git a/Assignment 3/packages/xlrd/doc/xlrd.html b/Assignment 3/packages/xlrd/doc/xlrd.html new file mode 100644 index 0000000..25113cc --- /dev/null +++ b/Assignment 3/packages/xlrd/doc/xlrd.html @@ -0,0 +1,2064 @@ + + + + +The xlrd Module + + +

The xlrd Module

+

A Python module for extracting data from MS Excel (TM) spreadsheet files. +

+Version 0.7.4 -- April 2012 +

+ +

General information

+ +

Acknowledgements

+ +

+Development of this module would not have been possible without the document +"OpenOffice.org's Documentation of the Microsoft Excel File Format" +("OOo docs" for short). +The latest version is available from OpenOffice.org in + PDF format +and + ODT format. +Small portions of the OOo docs are reproduced in this +document. A study of the OOo docs is recommended for those who wish a +deeper understanding of the Excel file layout than the xlrd docs can provide. +

+ +

Backporting to Python 2.1 was partially funded by + + Journyx - provider of timesheet and project accounting solutions. + +

+ +

Provision of formatting information in version 0.6.1 was funded by + + Simplistix Ltd. + +

+ +

Unicode

+ +

This module presents all text strings as Python unicode objects. +From Excel 97 onwards, text in Excel spreadsheets has been stored as Unicode. +Older files (Excel 95 and earlier) don't keep strings in Unicode; +a CODEPAGE record provides a codepage number (for example, 1252) which is +used by xlrd to derive the encoding (for same example: "cp1252") which is +used to translate to Unicode.

+ +

If the CODEPAGE record is missing (possible if the file was created +by third-party software), xlrd will assume that the encoding is ascii, and keep going. +If the actual encoding is not ascii, a UnicodeDecodeError exception will be raised and +you will need to determine the encoding yourself, and tell xlrd: +

+    book = xlrd.open_workbook(..., encoding_override="cp1252")
+

+

If the CODEPAGE record exists but is wrong (for example, the codepage +number is 1251, but the strings are actually encoded in koi8_r), +it can be overridden using the same mechanism. +The supplied runxlrd.py has a corresponding command-line argument, which +may be used for experimentation: +

+    runxlrd.py -e koi8_r 3rows myfile.xls
+

+

The first place to look for an encoding ("codec name") is + +the Python documentation. +

+ + +

Dates in Excel spreadsheets

+ +

In reality, there are no such things. What you have are floating point +numbers and pious hope. +There are several problems with Excel dates:

+ +

(1) Dates are not stored as a separate data type; they are stored as +floating point numbers and you have to rely on +(a) the "number format" applied to them in Excel and/or +(b) knowing which cells are supposed to have dates in them. +This module helps with (a) by inspecting the +format that has been applied to each number cell; +if it appears to be a date format, the cell +is classified as a date rather than a number. Feedback on this feature, +especially from non-English-speaking locales, would be appreciated.

+ +

(2) Excel for Windows stores dates by default as the number of +days (or fraction thereof) since 1899-12-31T00:00:00. Excel for +Macintosh uses a default start date of 1904-01-01T00:00:00. The date +system can be changed in Excel on a per-workbook basis (for example: +Tools -> Options -> Calculation, tick the "1904 date system" box). +This is of course a bad idea if there are already dates in the +workbook. There is no good reason to change it even if there are no +dates in the workbook. Which date system is in use is recorded in the +workbook. A workbook transported from Windows to Macintosh (or vice +versa) will work correctly with the host Excel. When using this +module's xldate_as_tuple function to convert numbers from a workbook, +you must use the datemode attribute of the Book object. If you guess, +or make a judgement depending on where you believe the workbook was +created, you run the risk of being 1462 days out of kilter.

+ +

Reference: +http://support.microsoft.com/default.aspx?scid=KB;EN-US;q180162

+ + +

(3) The Excel implementation of the Windows-default 1900-based date system works on the +incorrect premise that 1900 was a leap year. It interprets the number 60 as meaning 1900-02-29, +which is not a valid date. Consequently any number less than 61 is ambiguous. Example: is 59 the +result of 1900-02-28 entered directly, or is it 1900-03-01 minus 2 days? The OpenOffice.org Calc +program "corrects" the Microsoft problem; entering 1900-02-27 causes the number 59 to be stored. +Save as an XLS file, then open the file with Excel -- you'll see 1900-02-28 displayed.

+ +

Reference: http://support.microsoft.com/default.aspx?scid=kb;en-us;214326

+ +

(4) The Macintosh-default 1904-based date system counts 1904-01-02 as day 1 and 1904-01-01 as day zero. +Thus any number such that (0.0 <= number < 1.0) is ambiguous. Is 0.625 a time of day (15:00:00), +independent of the calendar, +or should it be interpreted as an instant on a particular day (1904-01-01T15:00:00)? +The xldate_* functions in this module +take the view that such a number is a calendar-independent time of day (like Python's datetime.time type) for both +date systems. This is consistent with more recent Microsoft documentation +(for example, the help file for Excel 2002 which says that the first day +in the 1904 date system is 1904-01-02). + +

(5) Usage of the Excel DATE() function may leave strange dates in a spreadsheet. Quoting the help file, +in respect of the 1900 date system: "If year is between 0 (zero) and 1899 (inclusive), +Excel adds that value to 1900 to calculate the year. For example, DATE(108,1,2) returns January 2, 2008 (1900+108)." +This gimmick, semi-defensible only for arguments up to 99 and only in the pre-Y2K-awareness era, +means that DATE(1899, 12, 31) is interpreted as 3799-12-31.

+ +

For further information, please refer to the documentation for the xldate_* functions.

+ +

Named references, constants, formulas, and macros

+ +

+A name is used to refer to a cell, a group of cells, a constant +value, a formula, or a macro. Usually the scope of a name is global +across the whole workbook. However it can be local to a worksheet. +For example, if the sales figures are in different cells in +different sheets, the user may define the name "Sales" in each +sheet. There are built-in names, like "Print_Area" and +"Print_Titles"; these two are naturally local to a sheet. +

+To inspect the names with a user interface like MS Excel, OOo Calc, +or Gnumeric, click on Insert/Names/Define. This will show the global +names, plus those local to the currently selected sheet. +

+A Book object provides two dictionaries (name_map and +name_and_scope_map) and a list (name_obj_list) which allow various +ways of accessing the Name objects. There is one Name object for +each NAME record found in the workbook. Name objects have many +attributes, several of which are relevant only when obj.macro is 1. +

+In the examples directory you will find namesdemo.xls which +showcases the many different ways that names can be used, and +xlrdnamesAPIdemo.py which offers 3 different queries for inspecting +the names in your files, and shows how to extract whatever a name is +referring to. There is currently one "convenience method", +Name.cell(), which extracts the value in the case where the name +refers to a single cell. More convenience methods are planned. The +source code for Name.cell (in __init__.py) is an extra source of +information on how the Name attributes hang together. +

+ +

Name information is not extracted from files older than +Excel 5.0 (Book.biff_version < 50)

+ +

Formatting

+ +

Introduction

+ +

This collection of features, new in xlrd version 0.6.1, is intended +to provide the information needed to (1) display/render spreadsheet contents +(say) on a screen or in a PDF file, and (2) copy spreadsheet data to another +file without losing the ability to display/render it.

+ +

The Palette; Colour Indexes

+ +

A colour is represented in Excel as a (red, green, blue) ("RGB") tuple +with each component in range(256). However it is not possible to access an +unlimited number of colours; each spreadsheet is limited to a palette of 64 different +colours (24 in Excel 3.0 and 4.0, 8 in Excel 2.0). Colours are referenced by an index +("colour index") into this palette. + +Colour indexes 0 to 7 represent 8 fixed built-in colours: black, white, red, green, blue, +yellow, magenta, and cyan.

+ +The remaining colours in the palette (8 to 63 in Excel 5.0 and later) +can be changed by the user. In the Excel 2003 UI, Tools/Options/Color presents a palette +of 7 rows of 8 colours. The last two rows are reserved for use in charts.
+The correspondence between this grid and the assigned +colour indexes is NOT left-to-right top-to-bottom.
+Indexes 8 to 15 correspond to changeable +parallels of the 8 fixed colours -- for example, index 7 is forever cyan; +index 15 starts off being cyan but can be changed by the user.
+ +The default colour for each index depends on the file version; tables of the defaults +are available in the source code. If the user changes one or more colours, +a PALETTE record appears in the XLS file -- it gives the RGB values for *all* changeable +indexes.
+Note that colours can be used in "number formats": "[CYAN]...." and "[COLOR8]...." refer +to colour index 7; "[COLOR16]...." will produce cyan +unless the user changes colour index 15 to something else.
+ +

In addition, there are several "magic" colour indexes used by Excel:
+0x18 (BIFF3-BIFF4), 0x40 (BIFF5-BIFF8): System window text colour for border lines +(used in XF, CF, and WINDOW2 records)
+0x19 (BIFF3-BIFF4), 0x41 (BIFF5-BIFF8): System window background colour for pattern background +(used in XF and CF records )
+0x43: System face colour (dialogue background colour)
+0x4D: System window text colour for chart border lines
+0x4E: System window background colour for chart areas
+0x4F: Automatic colour for chart border lines (seems to be always Black)
+0x50: System ToolTip background colour (used in note objects)
+0x51: System ToolTip text colour (used in note objects)
+0x7FFF: System window text colour for fonts (used in FONT and CF records)
+Note 0x7FFF appears to be the *default* colour index. It appears quite often in FONT +records.
+ +

Default Formatting

+ +Default formatting is applied to all empty cells (those not described by a cell record). +Firstly row default information (ROW record, Rowinfo class) is used if available. +Failing that, column default information (COLINFO record, Colinfo class) is used if available. +As a last resort the worksheet/workbook default cell format will be used; this +should always be present in an Excel file, +described by the XF record with the fixed index 15 (0-based). By default, it uses the +worksheet/workbook default cell style, described by the very first XF record (index 0). + +

Formatting features not included in xlrd version 0.6.1

+
    +
  • Rich text i.e. strings containing partial bold italic + and underlined text, change of font inside a string, etc. + See OOo docs s3.4 and s3.2. + Rich text is included in version 0.7.2
    • +
  • Asian phonetic text (known as "ruby"), used for Japanese furigana. See OOo docs + s3.4.2 (p15)
    • +
  • Conditional formatting. See OOo docs + s5.12, s6.21 (CONDFMT record), s6.16 (CF record)
    • +
  • Miscellaneous sheet-level and book-level items e.g. printing layout, screen panes.
    • +
  • Modern Excel file versions don't keep most of the built-in + "number formats" in the file; Excel loads formats according to the + user's locale. Currently xlrd's emulation of this is limited to + a hard-wired table that applies to the US English locale. This may mean + that currency symbols, date order, thousands separator, decimals separator, etc + are inappropriate. Note that this does not affect users who are copying XLS + files, only those who are visually rendering cells.
    • +
+ +

Loading worksheets on demand

+ +

This feature, new in version 0.7.1, is governed by the on_demand argument +to the open_workbook() function and allows saving memory and time by loading +only those sheets that the caller is interested in, and releasing sheets +when no longer required.

+ +

on_demand=False (default): No change. open_workbook() loads global data +and all sheets, releases resources no longer required (principally the +str or mmap object containing the Workbook stream), and returns.

+ +

on_demand=True and BIFF version < 5.0: A warning message is emitted, +on_demand is recorded as False, and the old process is followed.

+ +

on_demand=True and BIFF version >= 5.0: open_workbook() loads global +data and returns without releasing resources. At this stage, the only +information available about sheets is Book.nsheets and Book.sheet_names().

+ +

Book.sheet_by_name() and Book.sheet_by_index() will load the requested +sheet if it is not already loaded.

+ +

Book.sheets() will load all/any unloaded sheets.

+ +

The caller may save memory by calling +Book.unload_sheet(sheet_name_or_index) when finished with the sheet. +This applies irrespective of the state of on_demand.

+ +

The caller may re-load an unloaded sheet by calling Book.sheet_by_xxxx() + -- except if those required resources have been released (which will +have happened automatically when on_demand is false). This is the only +case where an exception will be raised.

+ +

The caller may query the state of a sheet: +Book.sheet_loaded(sheet_name_or_index) -> a bool

+ +

Book.release_resources() may used to save memory and close +any memory-mapped file before proceding to examine already-loaded +sheets. Once resources are released, no further sheets can be loaded.

+ +

When using on-demand, it is advisable to ensure that +Book.release_resources() is always called even if an exception +is raised in your own code; otherwise if the input file has been +memory-mapped, the mmap.mmap object will not be closed and you will +not be able to access the physical file until your Python process +terminates. This can be done by calling Book.release_resources() +explicitly in the finally suite of a try/finally block. +New in xlrd 0.7.2: the Book object is a "context manager", so if +using Python 2.5 or later, you can wrap your code in a "with" +statement.

+

Module Contents

+
+
BaseObject (class) [#]
+
+

Parent of almost all other classes in the package.

+

For more information about this class, see The BaseObject Class.

+
+
Book() (class) [#]
+
+

Contents of a "workbook".

+

For more information about this class, see The Book Class.

+
+
Cell(ctype, value, xf_index=None) (class) [#]
+
+

Contains the data for one cell.

+

For more information about this class, see The Cell Class.

+
+
cellname(rowx, colx) [#]
+
+

Utility function: (5, 7) => 'H6'

+
+
cellnameabs(rowx, colx, r1c1=0) [#]
+
+

Utility function: (5, 7) => '$H$6'

+
+
Colinfo (class) [#]
+
+

Width and default formatting information that applies to one or +more columns in a sheet.

+

For more information about this class, see The Colinfo Class.

+
+
colname(colx) [#]
+
+

Utility function: 7 => 'H', 27 => 'AB'

+
+
count_records(filename, outfile=sys.stdout) [#]
+
+

For debugging and analysis: summarise the file's BIFF records. +I.e. produce a sorted file of (record_name, count).

+
+
filename
+
+The path to the file to be summarised.
+
outfile
+
+An open file, to which the summary is written.
+

+
+
dump(filename, outfile=sys.stdout, unnumbered=False) [#]
+
+

For debugging: dump the file's BIFF records in char & hex. +

+
filename
+
+The path to the file to be dumped.
+
outfile
+
+An open file, to which the dump is written.
+
unnumbered
+
+If true, omit offsets (for meaningful diffs).
+

+
+
empty_cell (variable) [#]
+
+

There is one and only one instance of an empty cell -- it's a singleton. This is it. +You may use a test like "acell is empty_cell".

+
+
EqNeAttrs (class) [#]
+
+

This mixin class exists solely so that Format, Font, and XF....

+

For more information about this class, see The EqNeAttrs Class.

+
+
error_text_from_code (variable) [#]
+
+

This dictionary can be used to produce a text version of the internal codes +that Excel uses for error cells. Here are its contents: +

+0x00: '#NULL!',  # Intersection of two cell ranges is empty
+0x07: '#DIV/0!', # Division by zero
+0x0F: '#VALUE!', # Wrong type of operand
+0x17: '#REF!',   # Illegal or deleted cell reference
+0x1D: '#NAME?',  # Wrong function or range name
+0x24: '#NUM!',   # Value range overflow
+0x2A: '#N/A',    # Argument or function not available
+

+
+
Font (class) [#]
+
+

An Excel "font" contains the details of not only what is normally +considered a font, but also several other display attributes.

+

For more information about this class, see The Font Class.

+
+
Format(format_key, ty, format_str) (class) [#]
+
+

"Number format" information from a FORMAT record.

+

For more information about this class, see The Format Class.

+
+
Hyperlink (class) [#]
+
+

Contains the attributes of a hyperlink.

+

For more information about this class, see The Hyperlink Class.

+
+
Name (class) [#]
+
+

Information relating to a named reference, formula, macro, etc.

+

For more information about this class, see The Name Class.

+
+
Note (class) [#]
+
+

Represents a user "comment" or "note".

+

For more information about this class, see The Note Class.

+
+
open_workbook(filename=None, +logfile=sys.stdout, verbosity=0, use_mmap=USE_MMAP, +file_contents=None, +encoding_override=None, +formatting_info=False, on_demand=False, ragged_rows=False, +) [#]
+
+

Open a spreadsheet file for data extraction.

+
+
filename
+
+The path to the spreadsheet file to be opened.
+
logfile
+
+An open file to which messages and diagnostics are written.
+
verbosity
+
+Increases the volume of trace material written to the logfile.
+
use_mmap
+
+Whether to use the mmap module is determined heuristically. +Use this arg to override the result. Current heuristic: mmap is used if it exists.
+
file_contents
+
+... as a string or an mmap.mmap object or some other behave-alike object. +If file_contents is supplied, filename will not be used, except (possibly) in messages.
+
encoding_override
+
+Used to overcome missing or bad codepage information +in older-version files. Refer to discussion in the Unicode section above. +
-- New in version 0.6.0 + +
+
formatting_info
+
+Governs provision of a reference to an XF (eXtended Format) object +for each cell in the worksheet. +
Default is False. This is backwards compatible and saves memory. +"Blank" cells (those with their own formatting information but no data) are treated as empty +(by ignoring the file's BLANK and MULBLANK records). +It cuts off any bottom "margin" of rows of empty (and blank) cells and +any right "margin" of columns of empty (and blank) cells. +Only cell_value and cell_type are available. +
True provides all cells, including empty and blank cells. +XF information is available for each cell. +
-- New in version 0.6.1 + +
+
on_demand
+
+Governs whether sheets are all loaded initially or when demanded +by the caller. Please refer back to the section "Loading worksheets on demand" for details. +
-- New in version 0.7.1 + +
+
ragged_rows
+
+False (the default) means all rows are padded out with empty cells so that all +rows have the same size (Sheet.ncols). True means that there are no empty cells at the ends of rows. +This can result in substantial memory savings if rows are of widely varying sizes. See also the +Sheet.row_len() method. +
-- New in version 0.7.2 + +
+
Returns:
+
+An instance of the Book class.
+

+
+
Operand(akind=None, avalue=None, arank=0, atext='?') (class) [#]
+
+

Used in evaluating formulas.

+

For more information about this class, see The Operand Class.

+
+
rangename3d(book, ref3d) [#]
+
+

Utility function: +
Ref3D((1, 4, 5, 20, 7, 10)) => 'Sheet2:Sheet3!$H$6:$J$20' +

+
rangename3drel(book, ref3d, browx=None, bcolx=None, r1c1=0) [#]
+
+

Utility function: +
Ref3D(coords=(0, 1, -32, -22, -13, 13), relflags=(0, 0, 1, 1, 1, 1)) +R1C1 mode => 'Sheet1!R[-32]C[-13]:R[-23]C[12]' +A1 mode => depends on base cell (browx, bcolx) +

+
Ref3D(atuple) (class) [#]
+
+

Represents an absolute or relative 3-dimensional reference to a box +of one or more cells.

+

For more information about this class, see The Ref3D Class.

+
+
Rowinfo() (class) [#]
+
+

Height and default formatting information that applies to a row in a sheet.

+

For more information about this class, see The Rowinfo Class.

+
+
Sheet(book, position, name, number) (class) [#]
+
+

Contains the data for one worksheet.

+

For more information about this class, see The Sheet Class.

+
+
XF (class) [#]
+
+

eXtended Formatting information for cells, rows, columns and styles.

+

For more information about this class, see The XF Class.

+
+
XFAlignment (class) [#]
+
+

A collection of the alignment and similar attributes of an XF record.

+

For more information about this class, see The XFAlignment Class.

+
+
XFBackground (class) [#]
+
+

A collection of the background-related attributes of an XF record.

+

For more information about this class, see The XFBackground Class.

+
+
XFBorder (class) [#]
+
+

A collection of the border-related attributes of an XF record.

+

For more information about this class, see The XFBorder Class.

+
+
XFProtection (class) [#]
+
+

A collection of the protection-related attributes of an XF record.

+

For more information about this class, see The XFProtection Class.

+
+
xldate_as_tuple(xldate, datemode) [#]
+
+

Convert an Excel number (presumed to represent a date, a datetime or a time) into +a tuple suitable for feeding to datetime or mx.DateTime constructors.

+
+
xldate
+
+The Excel number
+
datemode
+
+0: 1900-based, 1: 1904-based. +
WARNING: when using this function to +interpret the contents of a workbook, you should pass in the Book.datemode +attribute of that workbook. Whether +the workbook has ever been anywhere near a Macintosh is irrelevant. +
+
Returns:
+
+Gregorian (year, month, day, hour, minute, nearest_second). +
Special case: if 0.0 <= xldate < 1.0, it is assumed to represent a time; +(0, 0, 0, hour, minute, second) will be returned. +
Note: 1904-01-01 is not regarded as a valid date in the datemode 1 system; its "serial number" +is zero. +
+
Raises XLDateNegative:
+xldate < 0.00 +
+
Raises XLDateAmbiguous:
+The 1900 leap-year problem (datemode == 0 and 1.0 <= xldate < 61.0) +
+
Raises XLDateTooLarge:
+Gregorian year 10000 or later
+
Raises XLDateBadDatemode:
+datemode arg is neither 0 nor 1
+
Raises XLDateError:
+Covers the 4 specific errors
+

+
+
xldate_from_date_tuple((year, month, day), datemode) [#]
+
+

Convert a date tuple (year, month, day) to an Excel date.

+
+
year
+
+Gregorian year.
+
month
+
+1 <= month <= 12 +
+
day
+
+1 <= day <= last day of that (year, month) +
+
datemode
+
+0: 1900-based, 1: 1904-based.
+
Raises XLDateAmbiguous:
+The 1900 leap-year problem (datemode == 0 and 1.0 <= xldate < 61.0) +
+
Raises XLDateBadDatemode:
+datemode arg is neither 0 nor 1
+
Raises XLDateBadTuple:
+(year, month, day) is too early/late or has invalid component(s)
+
Raises XLDateError:
+Covers the specific errors
+

+
+
xldate_from_datetime_tuple(datetime_tuple, datemode) [#]
+
+

Convert a datetime tuple (year, month, day, hour, minute, second) to an Excel date value. +For more details, refer to other xldate_from_*_tuple functions.

+
+
datetime_tuple
+
+(year, month, day, hour, minute, second)
+
datemode
+
+0: 1900-based, 1: 1904-based.
+

+
+
xldate_from_time_tuple((hour, minute, second)) [#]
+
+

Convert a time tuple (hour, minute, second) to an Excel "date" value (fraction of a day).

+
+
hour
+
+0 <= hour < 24 +
+
minute
+
+0 <= minute < 60 +
+
second
+
+0 <= second < 60 +
+
Raises XLDateBadTuple:
+Out-of-range hour, minute, or second
+

+
+
+

The BaseObject Class

+
+
BaseObject (class) [#]
+
+

Parent of almost all other classes in the package. Defines a common "dump" method +for debugging.

+
+
dump(f=None, header=None, footer=None, indent=0) [#]
+
+
+
f
+
+open file object, to which the dump is written
+
header
+
+text to write before the dump
+
footer
+
+text to write after the dump
+
indent
+
+number of leading spaces (for recursive calls)
+

+
+
+

The Book Class

+
+
Book() (class) [#]
+
+

Contents of a "workbook". +

WARNING: You don't call this class yourself. You use the Book object that +was returned when you called xlrd.open_workbook("myfile.xls").

+
+
biff_version [#]
+
+

Version of BIFF (Binary Interchange File Format) used to create the file. +Latest is 8.0 (represented here as 80), introduced with Excel 97. +Earliest supported by this module: 2.0 (represented as 20).

+
+
codepage [#]
+
+

An integer denoting the character set used for strings in this file. +For BIFF 8 and later, this will be 1200, meaning Unicode; more precisely, UTF_16_LE. +For earlier versions, this is used to derive the appropriate Python encoding +to be used to convert to Unicode. +Examples: 1252 -> 'cp1252', 10000 -> 'mac_roman'

+
+
colour_map [#]
+
+

This provides definitions for colour indexes. Please refer to the +above section "The Palette; Colour Indexes" for an explanation +of how colours are represented in Excel.
+Colour indexes into the palette map into (red, green, blue) tuples. +"Magic" indexes e.g. 0x7FFF map to None. +colour_map is what you need if you want to render cells on screen or in a PDF +file. If you are writing an output XLS file, use palette_record. +
-- New in version 0.6.1. Extracted only if open_workbook(..., formatting_info=True) +

+
countries [#]
+
+

A tuple containing the (telephone system) country code for:
+ [0]: the user-interface setting when the file was created.
+ [1]: the regional settings.
+Example: (1, 61) meaning (USA, Australia). +This information may give a clue to the correct encoding for an unknown codepage. +For a long list of observed values, refer to the OpenOffice.org documentation for +the COUNTRY record. +

+
datemode [#]
+
+

Which date system was in force when this file was last saved.
+ 0 => 1900 system (the Excel for Windows default).
+ 1 => 1904 system (the Excel for Macintosh default).
+

+
encoding [#]
+
+

The encoding that was derived from the codepage.

+
+
font_list [#]
+
+

A list of Font class instances, each corresponding to a FONT record. +
-- New in version 0.6.1 +

+
format_list [#]
+
+

A list of Format objects, each corresponding to a FORMAT record, in +the order that they appear in the input file. +It does not contain builtin formats. +If you are creating an output file using (for example) pyExcelerator, +use this list. +The collection to be used for all visual rendering purposes is format_map. +
-- New in version 0.6.1 +

+
format_map [#]
+
+

The mapping from XF.format_key to Format object. +
-- New in version 0.6.1 +

+
load_time_stage_1 [#]
+
+

Time in seconds to extract the XLS image as a contiguous string (or mmap equivalent).

+
+
load_time_stage_2 [#]
+
+

Time in seconds to parse the data from the contiguous string (or mmap equivalent).

+
+
name_and_scope_map [#]
+
+

A mapping from (lower_case_name, scope) to a single Name object. +
-- New in version 0.6.0 +

+
name_map [#]
+
+

A mapping from lower_case_name to a list of Name objects. The list is +sorted in scope order. Typically there will be one item (of global scope) +in the list. +
-- New in version 0.6.0 +

+
name_obj_list [#]
+
+

List containing a Name object for each NAME record in the workbook. +
-- New in version 0.6.0 +

+
nsheets [#]
+
+

The number of worksheets present in the workbook file. +This information is available even when no sheets have yet been loaded.

+
+
palette_record [#]
+
+

If the user has changed any of the colours in the standard palette, the XLS +file will contain a PALETTE record with 56 (16 for Excel 4.0 and earlier) +RGB values in it, and this list will be e.g. [(r0, b0, g0), ..., (r55, b55, g55)]. +Otherwise this list will be empty. This is what you need if you are +writing an output XLS file. If you want to render cells on screen or in a PDF +file, use colour_map. +
-- New in version 0.6.1. Extracted only if open_workbook(..., formatting_info=True) +

+
release_resources() [#]
+
+

This method has a dual purpose. You can call it to release +memory-consuming objects and (possibly) a memory-mapped file +(mmap.mmap object) when you have finished loading sheets in +on_demand mode, but still require the Book object to examine the +loaded sheets. It is also called automatically (a) when open_workbook +raises an exception and (b) if you are using a "with" statement, when +the "with" block is exited. Calling this method multiple times on the +same object has no ill effect.

+
+
sheet_by_index(sheetx) [#]
+
+
+
sheetx
+
+Sheet index in range(nsheets)
+
Returns:
+
+An object of the Sheet class
+

+
+
sheet_by_name(sheet_name) [#]
+
+
+
sheet_name
+
+Name of sheet required
+
Returns:
+
+An object of the Sheet class
+

+
+
sheet_loaded(sheet_name_or_index) [#]
+
+
+
sheet_name_or_index
+
+Name or index of sheet enquired upon
+
Returns:
+
+true if sheet is loaded, false otherwise +
-- New in version 0.7.1 +
+

+
+
sheet_names() [#]
+
+
+
Returns:
+
+A list of the names of all the worksheets in the workbook file. +This information is available even when no sheets have yet been loaded.
+

+
+
sheets() [#]
+
+
+
Returns:
+
+A list of all sheets in the book. +All sheets not already loaded will be loaded.
+

+
+
style_name_map [#]
+
+

This provides access via name to the extended format information for +both built-in styles and user-defined styles.
+It maps name to (built_in, xf_index), where:
+name is either the name of a user-defined style, +or the name of one of the built-in styles. Known built-in names are +Normal, RowLevel_1 to RowLevel_7, +ColLevel_1 to ColLevel_7, Comma, Currency, Percent, "Comma [0]", +"Currency [0]", Hyperlink, and "Followed Hyperlink".
+built_in 1 = built-in style, 0 = user-defined
+xf_index is an index into Book.xf_list.
+References: OOo docs s6.99 (STYLE record); Excel UI Format/Style +
-- New in version 0.6.1; since 0.7.4, extracted only if +open_workbook(..., formatting_info=True) +

+
unload_sheet(sheet_name_or_index) [#]
+
+
+
sheet_name_or_index
+
+Name or index of sheet to be unloaded. +
-- New in version 0.7.1 +
+

+
+
user_name [#]
+
+

What (if anything) is recorded as the name of the last user to save the file.

+
+
xf_list [#]
+
+

A list of XF class instances, each corresponding to an XF record. +
-- New in version 0.6.1 +

+
+

The Cell Class

+
+
Cell(ctype, value, xf_index=None) (class) [#]
+
+

Contains the data for one cell.

+ +

WARNING: You don't call this class yourself. You access Cell objects +via methods of the Sheet object(s) that you found in the Book object that +was returned when you called xlrd.open_workbook("myfile.xls").

+

Cell objects have three attributes: ctype is an int, value +(which depends on ctype) and xf_index. +If "formatting_info" is not enabled when the workbook is opened, xf_index will be None. +The following table describes the types of cells and how their values +are represented in Python.

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Type symbolType numberPython value
XL_CELL_EMPTY0empty string u''
XL_CELL_TEXT1a Unicode string
XL_CELL_NUMBER2float
XL_CELL_DATE3float
XL_CELL_BOOLEAN4int; 1 means TRUE, 0 means FALSE
XL_CELL_ERROR5int representing internal Excel codes; for a text representation, +refer to the supplied dictionary error_text_from_code
XL_CELL_BLANK6empty string u''. Note: this type will appear only when +open_workbook(..., formatting_info=True) is used.
+

+

+
+

The Colinfo Class

+
+
Colinfo (class) [#]
+
+

Width and default formatting information that applies to one or +more columns in a sheet. Derived from COLINFO records. + +

Here is the default hierarchy for width, according to the OOo docs: + +
"""In BIFF3, if a COLINFO record is missing for a column, +the width specified in the record DEFCOLWIDTH is used instead. + +
In BIFF4-BIFF7, the width set in this [COLINFO] record is only used, +if the corresponding bit for this column is cleared in the GCW +record, otherwise the column width set in the DEFCOLWIDTH record +is used (the STANDARDWIDTH record is always ignored in this case [see footnote!]). + +
In BIFF8, if a COLINFO record is missing for a column, +the width specified in the record STANDARDWIDTH is used. +If this [STANDARDWIDTH] record is also missing, +the column width of the record DEFCOLWIDTH is used instead.""" +
+ +Footnote: The docs on the GCW record say this: +"""
+If a bit is set, the corresponding column uses the width set in the STANDARDWIDTH +record. If a bit is cleared, the corresponding column uses the width set in the +COLINFO record for this column. +
If a bit is set, and the worksheet does not contain the STANDARDWIDTH record, or if +the bit is cleared, and the worksheet does not contain the COLINFO record, the DEFCOLWIDTH +record of the worksheet will be used instead. +
"""
+At the moment (2007-01-17) xlrd is going with the GCW version of the story. +Reference to the source may be useful: see the computed_column_width(colx) method +of the Sheet class. +
-- New in version 0.6.1 +

+
+
bit1_flag [#]
+
+

Value of a 1-bit flag whose purpose is unknown +but is often seen set to 1

+
+
collapsed [#]
+
+

1 = column is collapsed

+
+
hidden [#]
+
+

1 = column is hidden

+
+
outline_level [#]
+
+

Outline level of the column, in range(7). +(0 = no outline)

+
+
width [#]
+
+

Width of the column in 1/256 of the width of the zero character, +using default font (first FONT record in the file).

+
+
xf_index [#]
+
+

XF index to be used for formatting empty cells.

+
+
+

The EqNeAttrs Class

+
+
EqNeAttrs (class) [#]
+
+

This mixin class exists solely so that Format, Font, and XF.... objects +can be compared by value of their attributes.

+
+
+

The Font Class

+
+
Font (class) [#]
+
+

An Excel "font" contains the details of not only what is normally +considered a font, but also several other display attributes. +Items correspond to those in the Excel UI's Format/Cells/Font tab. +
-- New in version 0.6.1 +

+
bold [#]
+
+

1 = Characters are bold. Redundant; see "weight" attribute.

+
+
character_set [#]
+
+

Values: 0 = ANSI Latin, 1 = System default, 2 = Symbol, +77 = Apple Roman, +128 = ANSI Japanese Shift-JIS, +129 = ANSI Korean (Hangul), +130 = ANSI Korean (Johab), +134 = ANSI Chinese Simplified GBK, +136 = ANSI Chinese Traditional BIG5, +161 = ANSI Greek, +162 = ANSI Turkish, +163 = ANSI Vietnamese, +177 = ANSI Hebrew, +178 = ANSI Arabic, +186 = ANSI Baltic, +204 = ANSI Cyrillic, +222 = ANSI Thai, +238 = ANSI Latin II (Central European), +255 = OEM Latin I

+
+
colour_index [#]
+
+

An explanation of "colour index" is given in the Formatting +section at the start of this document.

+
+
escapement [#]
+
+

1 = Superscript, 2 = Subscript.

+
+
family [#]
+
+

0 = None (unknown or don't care)
+1 = Roman (variable width, serifed)
+2 = Swiss (variable width, sans-serifed)
+3 = Modern (fixed width, serifed or sans-serifed)
+4 = Script (cursive)
+5 = Decorative (specialised, for example Old English, Fraktur) +

+
font_index [#]
+
+

The 0-based index used to refer to this Font() instance. +Note that index 4 is never used; xlrd supplies a dummy place-holder.

+
+
height [#]
+
+

Height of the font (in twips). A twip = 1/20 of a point.

+
+
italic [#]
+
+

1 = Characters are italic.

+
+
name [#]
+
+

The name of the font. Example: u"Arial"

+
+
outline [#]
+
+

1 = Font is outline style (Macintosh only)

+
+
shadow [#]
+
+

1 = Font is shadow style (Macintosh only)

+
+
struck_out [#]
+
+

1 = Characters are struck out.

+
+
underline_type [#]
+
+

0 = None
+1 = Single; 0x21 (33) = Single accounting
+2 = Double; 0x22 (34) = Double accounting +

+
underlined [#]
+
+

1 = Characters are underlined. Redundant; see "underline_type" attribute.

+
+
weight [#]
+
+

Font weight (100-1000). Standard values are 400 for normal text +and 700 for bold text.

+
+
+

The Format Class

+
+
Format(format_key, ty, format_str) (class) [#]
+
+

"Number format" information from a FORMAT record. +
-- New in version 0.6.1 +

+
format_key [#]
+
+

The key into Book.format_map

+
+
format_str [#]
+
+

The format string

+
+
type [#]
+
+

A classification that has been inferred from the format string. +Currently, this is used only to distinguish between numbers and dates. +
Values: +
FUN = 0 # unknown +
FDT = 1 # date +
FNU = 2 # number +
FGE = 3 # general +
FTX = 4 # text +

+
+

The Hyperlink Class

+
+
Hyperlink (class) [#]
+
+

Contains the attributes of a hyperlink. +Hyperlink objects are accessible through Sheet.hyperlink_list +and Sheet.hyperlink_map. +
-- New in version 0.7.2 +

+
+
desc [#]
+
+

Description ... this is displayed in the cell, +and should be identical to the cell value. Unicode string, or None. It seems +impossible NOT to have a description created by the Excel UI.

+
+
fcolx [#]
+
+

Index of first column

+
+
frowx [#]
+
+

Index of first row

+
+
lcolx [#]
+
+

Index of last column

+
+
lrowx [#]
+
+

Index of last row

+
+
quicktip [#]
+
+

The text of the "quick tip" displayed when the cursor +hovers over the hyperlink.

+
+
target [#]
+
+

Target frame. Unicode string. Note: I have not seen a case of this. +It seems impossible to create one in the Excel UI.

+
+
textmark [#]
+
+

"Textmark": the piece after the "#" in +"http://docs.python.org/library#struct_module", or the Sheet1!A1:Z99 +part when type is "workbook".

+
+
type [#]
+
+

Type of hyperlink. Unicode string, one of 'url', 'unc', +'local file', 'workbook', 'unknown'

+
+
url_or_path [#]
+
+

The URL or file-path, depending in the type. Unicode string, except +in the rare case of a local but non-existent file with non-ASCII +characters in the name, in which case only the "8.3" filename is available, +as a bytes (3.x) or str (2.x) string, with unknown encoding. +

+
+

The Name Class

+
+
Name (class) [#]
+
+

Information relating to a named reference, formula, macro, etc. +
-- New in version 0.6.0 +
-- Name information is not extracted from files older than +Excel 5.0 (Book.biff_version < 50) +

+
area2d(clipped=True) [#]
+
+

This is a convenience method for the use case where the name +refers to one rectangular area in one worksheet.

+
+
clipped
+
+If true (the default), the returned rectangle is clipped +to fit in (0, sheet.nrows, 0, sheet.ncols) -- it is guaranteed that +0 <= rowxlo <= rowxhi <= sheet.nrows and that the number of usable rows +in the area (which may be zero) is rowxhi - rowxlo; likewise for columns. +
+
Returns:
+
+a tuple (sheet_object, rowxlo, rowxhi, colxlo, colxhi).
+
Raises XLRDError:
+The name is not a constant absolute reference +to a single area in a single sheet.
+

+
+
binary [#]
+
+

0 = Formula definition; 1 = Binary data
No examples have been sighted. +

+
builtin [#]
+
+

0 = User-defined name; 1 = Built-in name +(common examples: Print_Area, Print_Titles; see OOo docs for full list)

+
+
cell() [#]
+
+

This is a convenience method for the frequent use case where the name +refers to a single cell.

+
+
Returns:
+
+An instance of the Cell class.
+
Raises XLRDError:
+The name is not a constant absolute reference +to a single cell.
+

+
+
complex [#]
+
+

0 = Simple formula; 1 = Complex formula (array formula or user defined)
+No examples have been sighted. +

+
func [#]
+
+

0 = Command macro; 1 = Function macro. Relevant only if macro == 1

+
+
funcgroup [#]
+
+

Function group. Relevant only if macro == 1; see OOo docs for values.

+
+
hidden [#]
+
+

0 = Visible; 1 = Hidden

+
+
macro [#]
+
+

0 = Standard name; 1 = Macro name

+
+
name [#]
+
+

A Unicode string. If builtin, decoded as per OOo docs.

+
+
name_index [#]
+
+

The index of this object in book.name_obj_list

+
+
raw_formula [#]
+
+

An 8-bit string.

+
+
result [#]
+
+

The result of evaluating the formula, if any. +If no formula, or evaluation of the formula encountered problems, +the result is None. Otherwise the result is a single instance of the +Operand class.

+
+
scope [#]
+
+

-1: The name is global (visible in all calculation sheets).
+-2: The name belongs to a macro sheet or VBA sheet.
+-3: The name is invalid.
+0 <= scope < book.nsheets: The name is local to the sheet whose index is scope. +

+
vbasic [#]
+
+

0 = Sheet macro; 1 = VisualBasic macro. Relevant only if macro == 1

+
+
+

The Note Class

+
+
Note (class) [#]
+
+

Represents a user "comment" or "note". +Note objects are accessible through Sheet.cell_note_map. +
-- New in version 0.7.2 +

+
+
author [#]
+
+

Author of note

+
+
col_hidden [#]
+
+

True if the containing column is hidden

+
+
colx [#]
+
+

Column index

+
+
rich_text_runlist [#]
+
+

List of (offset_in_string, font_index) tuples. +Unlike Sheet.rich_text_runlist_map, the first offset should always be 0. +

+
row_hidden [#]
+
+

True if the containing row is hidden

+
+
rowx [#]
+
+

Row index

+
+
show [#]
+
+

True if note is always shown

+
+
text [#]
+
+

Text of the note

+
+
+

The Operand Class

+
+
Operand(akind=None, avalue=None, arank=0, atext='?') (class) [#]
+
+

Used in evaluating formulas. +The following table describes the kinds and how their values +are represented.

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Kind symbolKind numberValue representation
oBOOL3integer: 0 => False; 1 => True
oERR4None, or an int error code (same as XL_CELL_ERROR in the Cell class). +
oMSNG5Used by Excel as a placeholder for a missing (not supplied) function +argument. Should *not* appear as a final formula result. Value is None.
oNUM2A float. Note that there is no way of distinguishing dates.
oREF-1The value is either None or a non-empty list of +absolute Ref3D instances.
+
oREL-2The value is None or a non-empty list of +fully or partially relative Ref3D instances. +
oSTRG1A Unicode string.
oUNK0The kind is unknown or ambiguous. The value is None
+

+

+
kind [#]
+
+

oUNK means that the kind of operand is not known unambiguously.

+
+
text [#]
+
+

The reconstituted text of the original formula. Function names will be +in English irrespective of the original language, which doesn't seem +to be recorded anywhere. The separator is ",", not ";" or whatever else +might be more appropriate for the end-user's locale; patches welcome.

+
+
value [#]
+
+

None means that the actual value of the operand is a variable +(depends on cell data), not a constant.

+
+
+

The Ref3D Class

+
+
Ref3D(atuple) (class) [#]
+
+

Represents an absolute or relative 3-dimensional reference to a box +of one or more cells.
+-- New in version 0.6.0 +

+ +

The coords attribute is a tuple of the form:
+(shtxlo, shtxhi, rowxlo, rowxhi, colxlo, colxhi)
+where 0 <= thingxlo <= thingx < thingxhi.
+Note that it is quite possible to have thingx > nthings; for example +Print_Titles could have colxhi == 256 and/or rowxhi == 65536 +irrespective of how many columns/rows are actually used in the worksheet. +The caller will need to decide how to handle this situation. +Keyword: IndexError :-) +

+ +

The components of the coords attribute are also available as individual +attributes: shtxlo, shtxhi, rowxlo, rowxhi, colxlo, and colxhi.

+ +

The relflags attribute is a 6-tuple of flags which indicate whether +the corresponding (sheet|row|col)(lo|hi) is relative (1) or absolute (0).
+Note that there is necessarily no information available as to what cell(s) +the reference could possibly be relative to. The caller must decide what if +any use to make of oREL operands. Note also that a partially relative +reference may well be a typo. +For example, define name A1Z10 as $a$1:$z10 (missing $ after z) +while the cursor is on cell Sheet3!A27.
+The resulting Ref3D instance will have coords = (2, 3, 0, -16, 0, 26) +and relflags = (0, 0, 0, 1, 0, 0).
+So far, only one possibility of a sheet-relative component in +a reference has been noticed: a 2D reference located in the "current sheet". +
This will appear as coords = (0, 1, ...) and relflags = (1, 1, ...). +

+
+

The Rowinfo Class

+
+
Rowinfo() (class) [#]
+
+

Height and default formatting information that applies to a row in a sheet. +Derived from ROW records. +
-- New in version 0.6.1

+ +

height: Height of the row, in twips. One twip == 1/20 of a point.

+ +

has_default_height: 0 = Row has custom height; 1 = Row has default height.

+ +

outline_level: Outline level of the row (0 to 7)

+ +

outline_group_starts_ends: 1 = Outline group starts or ends here (depending on where the +outline buttons are located, see WSBOOL record [TODO ??]), +and is collapsed

+ +

hidden: 1 = Row is hidden (manually, or by a filter or outline group)

+ +

height_mismatch: 1 = Row height and default font height do not match

+ +

has_default_xf_index: 1 = the xf_index attribute is usable; 0 = ignore it

+ +

xf_index: Index to default XF record for empty cells in this row. +Don't use this if has_default_xf_index == 0.

+ +

additional_space_above: This flag is set, if the upper border of at least one cell in this row +or if the lower border of at least one cell in the row above is +formatted with a thick line style. Thin and medium line styles are not +taken into account.

+ +

additional_space_below: This flag is set, if the lower border of at least one cell in this row +or if the upper border of at least one cell in the row below is +formatted with a medium or thick line style. Thin line styles are not +taken into account.

+
+
+

The Sheet Class

+
+
Sheet(book, position, name, number) (class) [#]
+
+

Contains the data for one worksheet.

+ +

In the cell access functions, "rowx" is a row index, counting from zero, and "colx" is a +column index, counting from zero. +Negative values for row/column indexes and slice positions are supported in the expected fashion.

+ +

For information about cell types and cell values, refer to the documentation of the Cell class.

+ +

WARNING: You don't call this class yourself. You access Sheet objects via the Book object that +was returned when you called xlrd.open_workbook("myfile.xls").

+
+
book [#]
+
+

A reference to the Book object to which this sheet belongs. +Example usage: some_sheet.book.datemode

+
+
cell(rowx, colx) [#]
+
+

Cell object in the given row and column. +

+
cell_note_map [#]
+
+

A sparse mapping from (rowx, colx) to a Note object. +Cells not containing a note ("comment") are not mapped. +
-- New in version 0.7.2

+
+
cell_type(rowx, colx) [#]
+
+

Type of the cell in the given row and column. +Refer to the documentation of the Cell class. +

+
cell_value(rowx, colx) [#]
+
+

Value of the cell in the given row and column.

+
+
cell_xf_index(rowx, colx) [#]
+
+

XF index of the cell in the given row and column. +This is an index into Book.xf_list. +
-- New in version 0.6.1 +

+
col(colx) [#]
+
+

Returns a sequence of the Cell objects in the given column. +

+
col_label_ranges [#]
+
+

List of address ranges of cells containing column labels. +These are set up in Excel by Insert > Name > Labels > Columns. +
-- New in version 0.6.0 +
How to deconstruct the list: +

+for crange in thesheet.col_label_ranges:
+    rlo, rhi, clo, chi = crange
+    for rx in xrange(rlo, rhi):
+        for cx in xrange(clo, chi):
+            print "Column label at (rowx=%d, colx=%d) is %r" \
+                (rx, cx, thesheet.cell_value(rx, cx))
+
+

+
col_slice(colx, start_rowx=0, end_rowx=None) [#]
+
+

Returns a slice of the Cell objects in the given column. +

+
col_types(colx, start_rowx=0, end_rowx=None) [#]
+
+

Returns a slice of the types of the cells in the given column.

+
+
col_values(colx, start_rowx=0, end_rowx=None) [#]
+
+

Returns a slice of the values of the cells in the given column.

+
+
colinfo_map [#]
+
+

The map from a column index to a Colinfo object. Often there is an entry +in COLINFO records for all column indexes in range(257). +Note that xlrd ignores the entry for the non-existent +257th column. On the other hand, there may be no entry for unused columns. +
-- New in version 0.6.1. Populated only if open_workbook(formatting_info=True). +

+
computed_column_width(colx) [#]
+
+

Determine column display width. +
-- New in version 0.6.1 +
+

+
colx
+
+Index of the queried column, range 0 to 255. +Note that it is possible to find out the width that will be used to display +columns with no cell information e.g. column IV (colx=255).
+
Returns:
+
+The column width that will be used for displaying +the given column by Excel, in units of 1/256th of the width of a +standard character (the digit zero in the first font).
+

+
+
default_additional_space_above [#]
+
+

Default value to be used for a row if there is +no ROW record for that row. +From the optional DEFAULTROWHEIGHT record. +

+
default_additional_space_below [#]
+
+

Default value to be used for a row if there is +no ROW record for that row. +From the optional DEFAULTROWHEIGHT record. +

+
default_row_height [#]
+
+

Default value to be used for a row if there is +no ROW record for that row. +From the optional DEFAULTROWHEIGHT record. +

+
default_row_height_mismatch [#]
+
+

Default value to be used for a row if there is +no ROW record for that row. +From the optional DEFAULTROWHEIGHT record. +

+
default_row_hidden [#]
+
+

Default value to be used for a row if there is +no ROW record for that row. +From the optional DEFAULTROWHEIGHT record. +

+
defcolwidth [#]
+
+

Default column width from DEFCOLWIDTH record, else None. +From the OOo docs:
+"""Column width in characters, using the width of the zero character +from default font (first FONT record in the file). Excel adds some +extra space to the default width, depending on the default font and +default font size. The algorithm how to exactly calculate the resulting +column width is not known.
+Example: The default width of 8 set in this record results in a column +width of 8.43 using Arial font with a size of 10 points."""
+For the default hierarchy, refer to the Colinfo class. +
-- New in version 0.6.1 +

+
gcw [#]
+
+

A 256-element tuple corresponding to the contents of the GCW record for this sheet. +If no such record, treat as all bits zero. +Applies to BIFF4-7 only. See docs of the Colinfo class for discussion. +

+
has_pane_record [#]
+
+

Boolean specifying if a PANE record was present, ignore unless you're xlutils.copy

+
+
horizontal_page_breaks [#]
+
+

A list of the horizontal page breaks in this sheet. +Breaks are tuples in the form (index of row after break, start col index, end col index). +Populated only if open_workbook(formatting_info=True). +
-- New in version 0.7.2 +

+
horz_split_first_visible [#]
+
+

Index of first visible row in bottom frozen/split pane

+
+
horz_split_pos [#]
+
+

Number of rows in top pane (frozen panes; for split panes, see comments below in code)

+
+
hyperlink_list [#]
+
+

A list of Hyperlink objects corresponding to HLINK records found +in the worksheet.
-- New in version 0.7.2

+
+
hyperlink_map [#]
+
+

A sparse mapping from (rowx, colx) to an item in hyperlink_list. +Cells not covered by a hyperlink are not mapped. +It is possible using the Excel UI to set up a hyperlink that +covers a larger-than-1x1 rectangle of cells. +Hyperlink rectangles may overlap (Excel doesn't check). +When a multiply-covered cell is clicked on, the hyperlink that is activated +(and the one that is mapped here) is the last in hyperlink_list. +
-- New in version 0.7.2

+
+
merged_cells [#]
+
+

List of address ranges of cells which have been merged. +These are set up in Excel by Format > Cells > Alignment, then ticking +the "Merge cells" box. +
-- New in version 0.6.1. Extracted only if open_workbook(formatting_info=True). +
How to deconstruct the list: +

+for crange in thesheet.merged_cells:
+    rlo, rhi, clo, chi = crange
+    for rowx in xrange(rlo, rhi):
+        for colx in xrange(clo, chi):
+            # cell (rlo, clo) (the top left one) will carry the data
+            # and formatting info; the remainder will be recorded as
+            # blank cells, but a renderer will apply the formatting info
+            # for the top left cell (e.g. border, pattern) to all cells in
+            # the range.
+
+

+
name [#]
+
+

Name of sheet.

+
+
ncols [#]
+
+

Nominal number of columns in sheet. It is 1 + the maximum column index +found, ignoring trailing empty cells. See also open_workbook(ragged_rows=?) +and Sheet.row_len(row_index). +

+
nrows [#]
+
+

Number of rows in sheet. A row index is in range(thesheet.nrows).

+
+
rich_text_runlist_map [#]
+
+

Mapping of (rowx, colx) to list of (offset, font_index) tuples. The offset +defines where in the string the font begins to be used. +Offsets are expected to be in ascending order. +If the first offset is not zero, the meaning is that the cell's XF's font should +be used from offset 0. +
This is a sparse mapping. There is no entry for cells that are not formatted with +rich text. +
How to use: +

+runlist = thesheet.rich_text_runlist_map.get((rowx, colx))
+if runlist:
+    for offset, font_index in runlist:
+        # do work here.
+        pass
+
+Populated only if open_workbook(formatting_info=True). +
-- New in version 0.7.2. +
  +

+
row(rowx) [#]
+
+

Returns a sequence of the Cell objects in the given row. +

+
row_label_ranges [#]
+
+

List of address ranges of cells containing row labels. +For more details, see col_label_ranges above. +
-- New in version 0.6.0 +

+
row_len(rowx) [#]
+
+

Returns the effective number of cells in the given row. For use with +open_workbook(ragged_rows=True) which is likely to produce rows +with fewer than ncols cells. +
-- New in version 0.7.2 +

+
row_slice(rowx, start_colx=0, end_colx=None) [#]
+
+

Returns a slice of the Cell objects in the given row. +

+
row_types(rowx, start_colx=0, end_colx=None) [#]
+
+

Returns a slice of the types +of the cells in the given row.

+
+
row_values(rowx, start_colx=0, end_colx=None) [#]
+
+

Returns a slice of the values +of the cells in the given row.

+
+
rowinfo_map [#]
+
+

The map from a row index to a Rowinfo object. Note that it is possible +to have missing entries -- at least one source of XLS files doesn't +bother writing ROW records. +
-- New in version 0.6.1. Populated only if open_workbook(formatting_info=True). +

+
split_active_pane [#]
+
+

Frozen panes: ignore it. Split panes: explanation and diagrams in OOo docs.

+
+
standardwidth [#]
+
+

Default column width from STANDARDWIDTH record, else None. +From the OOo docs:
+"""Default width of the columns in 1/256 of the width of the zero +character, using default font (first FONT record in the file)."""
+For the default hierarchy, refer to the Colinfo class. +
-- New in version 0.6.1 +

+
vert_split_first_visible [#]
+
+

Index of first visible column in right frozen/split pane

+
+
vert_split_pos [#]
+
+

Number of columns in left pane (frozen panes; for split panes, see comments below in code)

+
+
vertical_page_breaks [#]
+
+

A list of the vertical page breaks in this sheet. +Breaks are tuples in the form (index of col after break, start row index, end row index). +Populated only if open_workbook(formatting_info=True). +
-- New in version 0.7.2 +

+
visibility [#]
+
+

Visibility of the sheet. 0 = visible, 1 = hidden (can be unhidden +by user -- Format/Sheet/Unhide), 2 = "very hidden" (can be unhidden +only by VBA macro).

+
+
+

The XF Class

+
+
XF (class) [#]
+
+

eXtended Formatting information for cells, rows, columns and styles. +
-- New in version 0.6.1 + +

Each of the 6 flags below describes the validity of +a specific group of attributes. +
+In cell XFs, flag==0 means the attributes of the parent style XF are used, +(but only if the attributes are valid there); flag==1 means the attributes +of this XF are used.
+In style XFs, flag==0 means the attribute setting is valid; flag==1 means +the attribute should be ignored.
+Note that the API +provides both "raw" XFs and "computed" XFs -- in the latter case, cell XFs +have had the above inheritance mechanism applied. +

+
+
_alignment_flag [#]
+
+
+
_background_flag [#]
+
+
+
_border_flag [#]
+
+
+
_font_flag [#]
+
+
+
_format_flag [#]
+
+
+
_protection_flag [#]
+
+

  +

+
alignment [#]
+
+

An instance of an XFAlignment object.

+
+
background [#]
+
+

An instance of an XFBackground object.

+
+
border [#]
+
+

An instance of an XFBorder object.

+
+
font_index [#]
+
+

Index into Book.font_list

+
+
format_key [#]
+
+

Key into Book.format_map +

+Warning: OOo docs on the XF record call this "Index to FORMAT record". +It is not an index in the Python sense. It is a key to a map. +It is true only for Excel 4.0 and earlier files +that the key into format_map from an XF instance +is the same as the index into format_list, and only +if the index is less than 164. +

+
+
is_style [#]
+
+

0 = cell XF, 1 = style XF

+
+
parent_style_index [#]
+
+

cell XF: Index into Book.xf_list +of this XF's style XF
+style XF: 0xFFF +

+
protection [#]
+
+

An instance of an XFProtection object.

+
+
xf_index [#]
+
+

Index into Book.xf_list

+
+
+

The XFAlignment Class

+
+
XFAlignment (class) [#]
+
+

A collection of the alignment and similar attributes of an XF record. +Items correspond to those in the Excel UI's Format/Cells/Alignment tab. +
-- New in version 0.6.1 +

+
hor_align [#]
+
+

Values: section 6.115 (p 214) of OOo docs

+
+
indent_level [#]
+
+

A number in range(15).

+
+
rotation [#]
+
+

Values: section 6.115 (p 215) of OOo docs.
+Note: file versions BIFF7 and earlier use the documented +"orientation" attribute; this will be mapped (without loss) +into "rotation". +

+
shrink_to_fit [#]
+
+

1 = shrink font size to fit text into cell.

+
+
text_direction [#]
+
+

0 = according to context; 1 = left-to-right; 2 = right-to-left

+
+
text_wrapped [#]
+
+

1 = text is wrapped at right margin

+
+
vert_align [#]
+
+

Values: section 6.115 (p 215) of OOo docs

+
+
+

The XFBackground Class

+
+
XFBackground (class) [#]
+
+

A collection of the background-related attributes of an XF record. +Items correspond to those in the Excel UI's Format/Cells/Patterns tab. +An explanation of "colour index" is given in the Formatting +section at the start of this document. +
-- New in version 0.6.1 +

+
background_colour_index [#]
+
+

See section 3.11 of the OOo docs.

+
+
fill_pattern [#]
+
+

See section 3.11 of the OOo docs.

+
+
pattern_colour_index [#]
+
+

See section 3.11 of the OOo docs.

+
+
+

The XFBorder Class

+
+
XFBorder (class) [#]
+
+

A collection of the border-related attributes of an XF record. +Items correspond to those in the Excel UI's Format/Cells/Border tab.

+

An explanations of "colour index" is given in the Formatting +section at the start of this document. +There are five line style attributes; possible values and the +associated meanings are: +0 = No line, +1 = Thin, +2 = Medium, +3 = Dashed, +4 = Dotted, +5 = Thick, +6 = Double, +7 = Hair, +8 = Medium dashed, +9 = Thin dash-dotted, +10 = Medium dash-dotted, +11 = Thin dash-dot-dotted, +12 = Medium dash-dot-dotted, +13 = Slanted medium dash-dotted. +The line styles 8 to 13 appear in BIFF8 files (Excel 97 and later) only. +For pictures of the line styles, refer to OOo docs s3.10 (p22) +"Line Styles for Cell Borders (BIFF3-BIFF8)".

+
-- New in version 0.6.1 +
+
bottom_colour_index [#]
+
+

The colour index for the cell's bottom line

+
+
bottom_line_style [#]
+
+

The line style for the cell's bottom line

+
+
diag_colour_index [#]
+
+

The colour index for the cell's diagonal lines, if any

+
+
diag_down [#]
+
+

1 = draw a diagonal from top left to bottom right

+
+
diag_line_style [#]
+
+

The line style for the cell's diagonal lines, if any

+
+
diag_up [#]
+
+

1 = draw a diagonal from bottom left to top right

+
+
left_colour_index [#]
+
+

The colour index for the cell's left line

+
+
left_line_style [#]
+
+

The line style for the cell's left line

+
+
right_colour_index [#]
+
+

The colour index for the cell's right line

+
+
right_line_style [#]
+
+

The line style for the cell's right line

+
+
top_colour_index [#]
+
+

The colour index for the cell's top line

+
+
top_line_style [#]
+
+

The line style for the cell's top line

+
+
+

The XFProtection Class

+
+
XFProtection (class) [#]
+
+

A collection of the protection-related attributes of an XF record. +Items correspond to those in the Excel UI's Format/Cells/Protection tab. +Note the OOo docs include the "cell or style" bit +in this bundle of attributes. +This is incorrect; the bit is used in determining which bundles to use. +
-- New in version 0.6.1 +

+
cell_locked [#]
+
+

1 = Cell is prevented from being changed, moved, resized, or deleted +(only if the sheet is protected).

+
+
formula_hidden [#]
+
+

1 = Hide formula so that it doesn't appear in the formula bar when +the cell is selected (only if the sheet is protected).

+
+
+ diff --git a/Assignment 3/packages/xlrd/examples/namesdemo.xls b/Assignment 3/packages/xlrd/examples/namesdemo.xls new file mode 100644 index 0000000..8a16865 Binary files /dev/null and b/Assignment 3/packages/xlrd/examples/namesdemo.xls differ diff --git a/Assignment 3/packages/xlrd/examples/xlrdnameAPIdemo.py b/Assignment 3/packages/xlrd/examples/xlrdnameAPIdemo.py new file mode 100644 index 0000000..b29a827 --- /dev/null +++ b/Assignment 3/packages/xlrd/examples/xlrdnameAPIdemo.py @@ -0,0 +1,179 @@ +# -*- coding: cp1252 -*- + +## +# Module/script example of the xlrd API for extracting information +# about named references, named constants, etc. +# +#

Copyright © 2006 Stephen John Machin, Lingfo Pty Ltd

+#

This module is part of the xlrd package, which is released under a BSD-style licence.

+## +from __future__ import print_function + +import xlrd +from xlrd.timemachine import REPR +import sys +import glob + +def scope_as_string(book, scope): + if 0 <= scope < book.nsheets: + return "sheet #%d (%r)" % (scope, REPR(book.sheet_names()[scope])) + if scope == -1: + return "Global" + if scope == -2: + return "Macro/VBA" + return "Unknown scope value (%r)" % REPR(scope) + +def do_scope_query(book, scope_strg, show_contents=0, f=sys.stdout): + try: + qscope = int(scope_strg) + except ValueError: + if scope_strg == "*": + qscope = None # means "all' + else: + # so assume it's a sheet name ... + qscope = book.sheet_names().index(scope_strg) + print("%r => %d" % (scope_strg, qscope), file=f) + for nobj in book.name_obj_list: + if qscope is None or nobj.scope == qscope: + show_name_object(book, nobj, show_contents, f) + +def show_name_details(book, name, show_contents=0, f=sys.stdout): + """ + book -- Book object obtained from xlrd.open_workbook(). + name -- The name that's being investigated. + show_contents -- 0: Don't; 1: Non-empty cells only; 2: All cells + f -- Open output file handle. + """ + name_lcase = name.lower() # Excel names are case-insensitive. + nobj_list = book.name_map.get(name_lcase) + if not nobj_list: + print("%r: unknown name" % name, file=f) + return + for nobj in nobj_list: + show_name_object(book, nobj, show_contents, f) + +def show_name_details_in_scope( + book, name, scope_strg, show_contents=0, f=sys.stdout, + ): + try: + scope = int(scope_strg) + except ValueError: + # so assume it's a sheet name ... + scope = book.sheet_names().index(scope_strg) + print("%r => %d" % (scope_strg, scope), file=f) + name_lcase = name.lower() # Excel names are case-insensitive. + while 1: + nobj = book.name_and_scope_map.get((name_lcase, scope)) + if nobj: + break + print("Name %s not found in scope %d" % (REPR(name), scope), file=f) + if scope == -1: + return + scope = -1 # Try again with global scope + print("Name %s found in scope %d" % (REPR(name), scope), file=f) + show_name_object(book, nobj, show_contents, f) + +def showable_cell_value(celltype, cellvalue, datemode): + if celltype == xlrd.XL_CELL_DATE: + try: + showval = xlrd.xldate_as_tuple(cellvalue, datemode) + except xlrd.XLDateError as e: + showval = "%s:%s" % (type(e).__name__, e) + elif celltype == xlrd.XL_CELL_ERROR: + showval = xlrd.error_text_from_code.get( + cellvalue, '' % cellvalue) + else: + showval = cellvalue + return showval + +def show_name_object(book, nobj, show_contents=0, f=sys.stdout): + print("\nName: %s, scope: %s (%s)" \ + % (REPR(nobj.name), REPR(nobj.scope), scope_as_string(book, nobj.scope)), file=f) + res = nobj.result + print("Formula eval result: %s" % REPR(res), file=f) + if res is None: + return + # result should be an instance of the Operand class + kind = res.kind + value = res.value + if kind >= 0: + # A scalar, or unknown ... you've seen all there is to see. + pass + elif kind == xlrd.oREL: + # A list of Ref3D objects representing *relative* ranges + for i in range(len(value)): + ref3d = value[i] + print("Range %d: %s ==> %s"% (i, REPR(ref3d.coords), REPR(xlrd.rangename3drel(book, ref3d))), file=f) + elif kind == xlrd.oREF: + # A list of Ref3D objects + for i in range(len(value)): + ref3d = value[i] + print("Range %d: %s ==> %s"% (i, REPR(ref3d.coords), REPR(xlrd.rangename3d(book, ref3d))), file=f) + if not show_contents: + continue + datemode = book.datemode + for shx in range(ref3d.shtxlo, ref3d.shtxhi): + sh = book.sheet_by_index(shx) + print(" Sheet #%d (%s)" % (shx, sh.name), file=f) + rowlim = min(ref3d.rowxhi, sh.nrows) + collim = min(ref3d.colxhi, sh.ncols) + for rowx in range(ref3d.rowxlo, rowlim): + for colx in range(ref3d.colxlo, collim): + cty = sh.cell_type(rowx, colx) + if cty == xlrd.XL_CELL_EMPTY and show_contents == 1: + continue + cval = sh.cell_value(rowx, colx) + sval = showable_cell_value(cty, cval, datemode) + print(" (%3d,%3d) %-5s: %s" + % (rowx, colx, xlrd.cellname(rowx, colx), REPR(sval)), file=f) + +if __name__ == "__main__": + def usage(): + text = """ +usage: xlrdnameAIPdemo.py glob_pattern name scope show_contents + +where: + "glob_pattern" designates a set of files + "name" is a name or '*' (all names) + "scope" is -1 (global) or a sheet number + or a sheet name or * (all scopes) + "show_contents" is one of 0 (no show), + 1 (only non-empty cells), or 2 (all cells) + +Examples (script name and glob_pattern arg omitted for brevity) + [Searching through book.name_obj_list] + * * 0 lists all names + * * 1 lists all names, showing referenced non-empty cells + * 1 0 lists all names local to the 2nd sheet + * Northern 0 lists all names local to the 'Northern' sheet + * -1 0 lists all names with global scope + [Initial direct access through book.name_map] + Sales * 0 lists all occurrences of "Sales" in any scope + [Direct access through book.name_and_scope_map] + Revenue -1 0 checks if "Revenue" exists in global scope + +""" + sys.stdout.write(text) + + if len(sys.argv) != 5: + usage() + sys.exit(0) + arg_pattern = sys.argv[1] # glob pattern e.g. "foo*.xls" + arg_name = sys.argv[2] # see below + arg_scope = sys.argv[3] # see below + arg_show_contents = int(sys.argv[4]) # 0: no show, 1: only non-empty cells, + # 2: all cells + for fname in glob.glob(arg_pattern): + book = xlrd.open_workbook(fname) + if arg_name == "*": + # Examine book.name_obj_list to find all names + # in a given scope ("*" => all scopes) + do_scope_query(book, arg_scope, arg_show_contents) + elif arg_scope == "*": + # Using book.name_map to find all usage of a name. + show_name_details(book, arg_name, arg_show_contents) + else: + # Using book.name_and_scope_map to find which if any instances + # of a name are visible in the given scope, which can be supplied + # as -1 (global) or a sheet number or a sheet name. + show_name_details_in_scope(book, arg_name, arg_scope, arg_show_contents) diff --git a/Assignment 3/packages/xlrd/formatting.py b/Assignment 3/packages/xlrd/formatting.py new file mode 100644 index 0000000..f044915 --- /dev/null +++ b/Assignment 3/packages/xlrd/formatting.py @@ -0,0 +1,1262 @@ +# -*- coding: cp1252 -*- + +## +# Module for formatting information. +# +#

Copyright © 2005-2012 Stephen John Machin, Lingfo Pty Ltd

+#

This module is part of the xlrd package, which is released under +# a BSD-style licence.

+## + +# No part of the content of this file was derived from the works of David Giffin. + +from __future__ import print_function + +DEBUG = 0 +import re +from struct import unpack +from .timemachine import * +from .biffh import BaseObject, unpack_unicode, unpack_string, \ + upkbits, upkbitsL, fprintf, \ + FUN, FDT, FNU, FGE, FTX, XL_CELL_NUMBER, XL_CELL_DATE, \ + XL_FORMAT, XL_FORMAT2, \ + XLRDError + +_cellty_from_fmtty = { + FNU: XL_CELL_NUMBER, + FUN: XL_CELL_NUMBER, + FGE: XL_CELL_NUMBER, + FDT: XL_CELL_DATE, + FTX: XL_CELL_NUMBER, # Yes, a number can be formatted as text. + } + +excel_default_palette_b5 = ( + ( 0, 0, 0), (255, 255, 255), (255, 0, 0), ( 0, 255, 0), + ( 0, 0, 255), (255, 255, 0), (255, 0, 255), ( 0, 255, 255), + (128, 0, 0), ( 0, 128, 0), ( 0, 0, 128), (128, 128, 0), + (128, 0, 128), ( 0, 128, 128), (192, 192, 192), (128, 128, 128), + (153, 153, 255), (153, 51, 102), (255, 255, 204), (204, 255, 255), + (102, 0, 102), (255, 128, 128), ( 0, 102, 204), (204, 204, 255), + ( 0, 0, 128), (255, 0, 255), (255, 255, 0), ( 0, 255, 255), + (128, 0, 128), (128, 0, 0), ( 0, 128, 128), ( 0, 0, 255), + ( 0, 204, 255), (204, 255, 255), (204, 255, 204), (255, 255, 153), + (153, 204, 255), (255, 153, 204), (204, 153, 255), (227, 227, 227), + ( 51, 102, 255), ( 51, 204, 204), (153, 204, 0), (255, 204, 0), + (255, 153, 0), (255, 102, 0), (102, 102, 153), (150, 150, 150), + ( 0, 51, 102), ( 51, 153, 102), ( 0, 51, 0), ( 51, 51, 0), + (153, 51, 0), (153, 51, 102), ( 51, 51, 153), ( 51, 51, 51), + ) + +excel_default_palette_b2 = excel_default_palette_b5[:16] + +# Following table borrowed from Gnumeric 1.4 source. +# Checked against OOo docs and MS docs. +excel_default_palette_b8 = ( # (red, green, blue) + ( 0, 0, 0), (255,255,255), (255, 0, 0), ( 0,255, 0), # 0 + ( 0, 0,255), (255,255, 0), (255, 0,255), ( 0,255,255), # 4 + (128, 0, 0), ( 0,128, 0), ( 0, 0,128), (128,128, 0), # 8 + (128, 0,128), ( 0,128,128), (192,192,192), (128,128,128), # 12 + (153,153,255), (153, 51,102), (255,255,204), (204,255,255), # 16 + (102, 0,102), (255,128,128), ( 0,102,204), (204,204,255), # 20 + ( 0, 0,128), (255, 0,255), (255,255, 0), ( 0,255,255), # 24 + (128, 0,128), (128, 0, 0), ( 0,128,128), ( 0, 0,255), # 28 + ( 0,204,255), (204,255,255), (204,255,204), (255,255,153), # 32 + (153,204,255), (255,153,204), (204,153,255), (255,204,153), # 36 + ( 51,102,255), ( 51,204,204), (153,204, 0), (255,204, 0), # 40 + (255,153, 0), (255,102, 0), (102,102,153), (150,150,150), # 44 + ( 0, 51,102), ( 51,153,102), ( 0, 51, 0), ( 51, 51, 0), # 48 + (153, 51, 0), (153, 51,102), ( 51, 51,153), ( 51, 51, 51), # 52 + ) + +default_palette = { + 80: excel_default_palette_b8, + 70: excel_default_palette_b5, + 50: excel_default_palette_b5, + 45: excel_default_palette_b2, + 40: excel_default_palette_b2, + 30: excel_default_palette_b2, + 21: excel_default_palette_b2, + 20: excel_default_palette_b2, + } + +""" +00H = Normal +01H = RowLevel_lv (see next field) +02H = ColLevel_lv (see next field) +03H = Comma +04H = Currency +05H = Percent +06H = Comma [0] (BIFF4-BIFF8) +07H = Currency [0] (BIFF4-BIFF8) +08H = Hyperlink (BIFF8) +09H = Followed Hyperlink (BIFF8) +""" +built_in_style_names = [ + "Normal", + "RowLevel_", + "ColLevel_", + "Comma", + "Currency", + "Percent", + "Comma [0]", + "Currency [0]", + "Hyperlink", + "Followed Hyperlink", + ] + +def initialise_colour_map(book): + book.colour_map = {} + book.colour_indexes_used = {} + if not book.formatting_info: + return + # Add the 8 invariant colours + for i in xrange(8): + book.colour_map[i] = excel_default_palette_b8[i] + # Add the default palette depending on the version + dpal = default_palette[book.biff_version] + ndpal = len(dpal) + for i in xrange(ndpal): + book.colour_map[i+8] = dpal[i] + # Add the specials -- None means the RGB value is not known + # System window text colour for border lines + book.colour_map[ndpal+8] = None + # System window background colour for pattern background + book.colour_map[ndpal+8+1] = None # + for ci in ( + 0x51, # System ToolTip text colour (used in note objects) + 0x7FFF, # 32767, system window text colour for fonts + ): + book.colour_map[ci] = None + +def nearest_colour_index(colour_map, rgb, debug=0): + # General purpose function. Uses Euclidean distance. + # So far used only for pre-BIFF8 WINDOW2 record. + # Doesn't have to be fast. + # Doesn't have to be fancy. + best_metric = 3 * 256 * 256 + best_colourx = 0 + for colourx, cand_rgb in colour_map.items(): + if cand_rgb is None: + continue + metric = 0 + for v1, v2 in zip(rgb, cand_rgb): + metric += (v1 - v2) * (v1 - v2) + if metric < best_metric: + best_metric = metric + best_colourx = colourx + if metric == 0: + break + if 0 and debug: + print("nearest_colour_index for %r is %r -> %r; best_metric is %d" \ + % (rgb, best_colourx, colour_map[best_colourx], best_metric)) + return best_colourx + +## +# This mixin class exists solely so that Format, Font, and XF.... objects +# can be compared by value of their attributes. +class EqNeAttrs(object): + + def __eq__(self, other): + return self.__dict__ == other.__dict__ + + def __ne__(self, other): + return self.__dict__ != other.__dict__ + +## +# An Excel "font" contains the details of not only what is normally +# considered a font, but also several other display attributes. +# Items correspond to those in the Excel UI's Format/Cells/Font tab. +#
-- New in version 0.6.1 +class Font(BaseObject, EqNeAttrs): + ## + # 1 = Characters are bold. Redundant; see "weight" attribute. + bold = 0 + ## + # Values: 0 = ANSI Latin, 1 = System default, 2 = Symbol, + # 77 = Apple Roman, + # 128 = ANSI Japanese Shift-JIS, + # 129 = ANSI Korean (Hangul), + # 130 = ANSI Korean (Johab), + # 134 = ANSI Chinese Simplified GBK, + # 136 = ANSI Chinese Traditional BIG5, + # 161 = ANSI Greek, + # 162 = ANSI Turkish, + # 163 = ANSI Vietnamese, + # 177 = ANSI Hebrew, + # 178 = ANSI Arabic, + # 186 = ANSI Baltic, + # 204 = ANSI Cyrillic, + # 222 = ANSI Thai, + # 238 = ANSI Latin II (Central European), + # 255 = OEM Latin I + character_set = 0 + ## + # An explanation of "colour index" is given in the Formatting + # section at the start of this document. + colour_index = 0 + ## + # 1 = Superscript, 2 = Subscript. + escapement = 0 + ## + # 0 = None (unknown or don't care)
+ # 1 = Roman (variable width, serifed)
+ # 2 = Swiss (variable width, sans-serifed)
+ # 3 = Modern (fixed width, serifed or sans-serifed)
+ # 4 = Script (cursive)
+ # 5 = Decorative (specialised, for example Old English, Fraktur) + family = 0 + ## + # The 0-based index used to refer to this Font() instance. + # Note that index 4 is never used; xlrd supplies a dummy place-holder. + font_index = 0 + ## + # Height of the font (in twips). A twip = 1/20 of a point. + height = 0 + ## + # 1 = Characters are italic. + italic = 0 + ## + # The name of the font. Example: u"Arial" + name = UNICODE_LITERAL("") + ## + # 1 = Characters are struck out. + struck_out = 0 + ## + # 0 = None
+ # 1 = Single; 0x21 (33) = Single accounting
+ # 2 = Double; 0x22 (34) = Double accounting + underline_type = 0 + ## + # 1 = Characters are underlined. Redundant; see "underline_type" attribute. + underlined = 0 + ## + # Font weight (100-1000). Standard values are 400 for normal text + # and 700 for bold text. + weight = 400 + ## + # 1 = Font is outline style (Macintosh only) + outline = 0 + ## + # 1 = Font is shadow style (Macintosh only) + shadow = 0 + + # No methods ... + +def handle_efont(book, data): # BIFF2 only + if not book.formatting_info: + return + book.font_list[-1].colour_index = unpack('= 2 + bv = book.biff_version + k = len(book.font_list) + if k == 4: + f = Font() + f.name = UNICODE_LITERAL('Dummy Font') + f.font_index = k + book.font_list.append(f) + k += 1 + f = Font() + f.font_index = k + book.font_list.append(f) + if bv >= 50: + ( + f.height, option_flags, f.colour_index, f.weight, + f.escapement, f.underline_type, f.family, + f.character_set, + ) = unpack('> 1 + f.underlined = (option_flags & 4) >> 2 + f.struck_out = (option_flags & 8) >> 3 + f.outline = (option_flags & 16) >> 4 + f.shadow = (option_flags & 32) >> 5 + if bv >= 80: + f.name = unpack_unicode(data, 14, lenlen=1) + else: + f.name = unpack_string(data, 14, book.encoding, lenlen=1) + elif bv >= 30: + f.height, option_flags, f.colour_index = unpack('> 1 + f.underlined = (option_flags & 4) >> 2 + f.struck_out = (option_flags & 8) >> 3 + f.outline = (option_flags & 16) >> 4 + f.shadow = (option_flags & 32) >> 5 + f.name = unpack_string(data, 6, book.encoding, lenlen=1) + # Now cook up the remaining attributes ... + f.weight = [400, 700][f.bold] + f.escapement = 0 # None + f.underline_type = f.underlined # None or Single + f.family = 0 # Unknown / don't care + f.character_set = 1 # System default (0 means "ANSI Latin") + else: # BIFF2 + f.height, option_flags = unpack('> 1 + f.underlined = (option_flags & 4) >> 2 + f.struck_out = (option_flags & 8) >> 3 + f.outline = 0 + f.shadow = 0 + f.name = unpack_string(data, 4, book.encoding, lenlen=1) + # Now cook up the remaining attributes ... + f.weight = [400, 700][f.bold] + f.escapement = 0 # None + f.underline_type = f.underlined # None or Single + f.family = 0 # Unknown / don't care + f.character_set = 1 # System default (0 means "ANSI Latin") + if blah: + f.dump( + book.logfile, + header="--- handle_font: font[%d] ---" % f.font_index, + footer="-------------------", + ) + +# === "Number formats" === + +## +# "Number format" information from a FORMAT record. +#
-- New in version 0.6.1 +class Format(BaseObject, EqNeAttrs): + ## + # The key into Book.format_map + format_key = 0 + ## + # A classification that has been inferred from the format string. + # Currently, this is used only to distinguish between numbers and dates. + #
Values: + #
FUN = 0 # unknown + #
FDT = 1 # date + #
FNU = 2 # number + #
FGE = 3 # general + #
FTX = 4 # text + type = FUN + ## + # The format string + format_str = UNICODE_LITERAL('') + + def __init__(self, format_key, ty, format_str): + self.format_key = format_key + self.type = ty + self.format_str = format_str + +std_format_strings = { + # "std" == "standard for US English locale" + # #### TODO ... a lot of work to tailor these to the user's locale. + # See e.g. gnumeric-1.x.y/src/formats.c + 0x00: "General", + 0x01: "0", + 0x02: "0.00", + 0x03: "#,##0", + 0x04: "#,##0.00", + 0x05: "$#,##0_);($#,##0)", + 0x06: "$#,##0_);[Red]($#,##0)", + 0x07: "$#,##0.00_);($#,##0.00)", + 0x08: "$#,##0.00_);[Red]($#,##0.00)", + 0x09: "0%", + 0x0a: "0.00%", + 0x0b: "0.00E+00", + 0x0c: "# ?/?", + 0x0d: "# ??/??", + 0x0e: "m/d/yy", + 0x0f: "d-mmm-yy", + 0x10: "d-mmm", + 0x11: "mmm-yy", + 0x12: "h:mm AM/PM", + 0x13: "h:mm:ss AM/PM", + 0x14: "h:mm", + 0x15: "h:mm:ss", + 0x16: "m/d/yy h:mm", + 0x25: "#,##0_);(#,##0)", + 0x26: "#,##0_);[Red](#,##0)", + 0x27: "#,##0.00_);(#,##0.00)", + 0x28: "#,##0.00_);[Red](#,##0.00)", + 0x29: "_(* #,##0_);_(* (#,##0);_(* \"-\"_);_(@_)", + 0x2a: "_($* #,##0_);_($* (#,##0);_($* \"-\"_);_(@_)", + 0x2b: "_(* #,##0.00_);_(* (#,##0.00);_(* \"-\"??_);_(@_)", + 0x2c: "_($* #,##0.00_);_($* (#,##0.00);_($* \"-\"??_);_(@_)", + 0x2d: "mm:ss", + 0x2e: "[h]:mm:ss", + 0x2f: "mm:ss.0", + 0x30: "##0.0E+0", + 0x31: "@", + } + +fmt_code_ranges = [ # both-inclusive ranges of "standard" format codes + # Source: the openoffice.org doc't + # and the OOXML spec Part 4, section 3.8.30 + ( 0, 0, FGE), + ( 1, 13, FNU), + (14, 22, FDT), + (27, 36, FDT), # CJK date formats + (37, 44, FNU), + (45, 47, FDT), + (48, 48, FNU), + (49, 49, FTX), + # Gnumeric assumes (or assumed) that built-in formats finish at 49, not at 163 + (50, 58, FDT), # CJK date formats + (59, 62, FNU), # Thai number (currency?) formats + (67, 70, FNU), # Thai number (currency?) formats + (71, 81, FDT), # Thai date formats + ] + +std_format_code_types = {} +for lo, hi, ty in fmt_code_ranges: + for x in xrange(lo, hi+1): + std_format_code_types[x] = ty +del lo, hi, ty, x + +date_chars = UNICODE_LITERAL('ymdhs') # year, month/minute, day, hour, second +date_char_dict = {} +for _c in date_chars + date_chars.upper(): + date_char_dict[_c] = 5 +del _c, date_chars + +skip_char_dict = {} +for _c in UNICODE_LITERAL('$-+/(): '): + skip_char_dict[_c] = 1 + +num_char_dict = { + UNICODE_LITERAL('0'): 5, + UNICODE_LITERAL('#'): 5, + UNICODE_LITERAL('?'): 5, + } + +non_date_formats = { + UNICODE_LITERAL('0.00E+00'):1, + UNICODE_LITERAL('##0.0E+0'):1, + UNICODE_LITERAL('General') :1, + UNICODE_LITERAL('GENERAL') :1, # OOo Calc 1.1.4 does this. + UNICODE_LITERAL('general') :1, # pyExcelerator 0.6.3 does this. + UNICODE_LITERAL('@') :1, + } + +fmt_bracketed_sub = re.compile(r'\[[^]]*\]').sub + +# Boolean format strings (actual cases) +# u'"Yes";"Yes";"No"' +# u'"True";"True";"False"' +# u'"On";"On";"Off"' + +def is_date_format_string(book, fmt): + # Heuristics: + # Ignore "text" and [stuff in square brackets (aarrgghh -- see below)]. + # Handle backslashed-escaped chars properly. + # E.g. hh\hmm\mss\s should produce a display like 23h59m59s + # Date formats have one or more of ymdhs (caseless) in them. + # Numeric formats have # and 0. + # N.B. u'General"."' hence get rid of "text" first. + # TODO: Find where formats are interpreted in Gnumeric + # TODO: u'[h]\\ \\h\\o\\u\\r\\s' ([h] means don't care about hours > 23) + state = 0 + s = '' + + for c in fmt: + if state == 0: + if c == UNICODE_LITERAL('"'): + state = 1 + elif c in UNICODE_LITERAL(r"\_*"): + state = 2 + elif c in skip_char_dict: + pass + else: + s += c + elif state == 1: + if c == UNICODE_LITERAL('"'): + state = 0 + elif state == 2: + # Ignore char after backslash, underscore or asterisk + state = 0 + assert 0 <= state <= 2 + if book.verbosity >= 4: + print("is_date_format_string: reduced format is %s" % REPR(s), file=book.logfile) + s = fmt_bracketed_sub('', s) + if s in non_date_formats: + return False + state = 0 + separator = ";" + got_sep = 0 + date_count = num_count = 0 + for c in s: + if c in date_char_dict: + date_count += date_char_dict[c] + elif c in num_char_dict: + num_count += num_char_dict[c] + elif c == separator: + got_sep = 1 + # print num_count, date_count, repr(fmt) + if date_count and not num_count: + return True + if num_count and not date_count: + return False + if date_count: + if book.verbosity: + fprintf(book.logfile, + 'WARNING *** is_date_format: ambiguous d=%d n=%d fmt=%r\n', + date_count, num_count, fmt) + elif not got_sep: + if book.verbosity: + fprintf(book.logfile, + "WARNING *** format %r produces constant result\n", + fmt) + return date_count > num_count + +def handle_format(self, data, rectype=XL_FORMAT): + DEBUG = 0 + bv = self.biff_version + if rectype == XL_FORMAT2: + bv = min(bv, 30) + if not self.encoding: + self.derive_encoding() + strpos = 2 + if bv >= 50: + fmtkey = unpack('= 80: + unistrg = unpack_unicode(data, 2) + else: + unistrg = unpack_string(data, strpos, self.encoding, lenlen=1) + blah = DEBUG or self.verbosity >= 3 + if blah: + fprintf(self.logfile, + "FORMAT: count=%d fmtkey=0x%04x (%d) s=%r\n", + self.actualfmtcount, fmtkey, fmtkey, unistrg) + is_date_s = self.is_date_format_string(unistrg) + ty = [FGE, FDT][is_date_s] + if not(fmtkey > 163 or bv < 50): + # user_defined if fmtkey > 163 + # N.B. Gnumeric incorrectly starts these at 50 instead of 164 :-( + # if earlier than BIFF 5, standard info is useless + std_ty = std_format_code_types.get(fmtkey, FUN) + # print "std ty", std_ty + is_date_c = std_ty == FDT + if self.verbosity and 0 < fmtkey < 50 and (is_date_c ^ is_date_s): + DEBUG = 2 + fprintf(self.logfile, + "WARNING *** Conflict between " + "std format key %d and its format string %r\n", + fmtkey, unistrg) + if DEBUG == 2: + fprintf(self.logfile, + "ty: %d; is_date_c: %r; is_date_s: %r; fmt_strg: %r", + ty, is_date_c, is_date_s, unistrg) + fmtobj = Format(fmtkey, ty, unistrg) + if blah: + fmtobj.dump(self.logfile, + header="--- handle_format [%d] ---" % (self.actualfmtcount-1, )) + self.format_map[fmtkey] = fmtobj + self.format_list.append(fmtobj) + +# ============================================================================= + +def handle_palette(book, data): + if not book.formatting_info: + return + blah = DEBUG or book.verbosity >= 2 + n_colours, = unpack('= 50] + if ((DEBUG or book.verbosity >= 1) + and n_colours != expected_n_colours): + fprintf(book.logfile, + "NOTE *** Expected %d colours in PALETTE record, found %d\n", + expected_n_colours, n_colours) + elif blah: + fprintf(book.logfile, + "PALETTE record with %d colours\n", n_colours) + fmt = '> 8) & 0xff + blue = (c >> 16) & 0xff + old_rgb = book.colour_map[8+i] + new_rgb = (red, green, blue) + book.palette_record.append(new_rgb) + book.colour_map[8+i] = new_rgb + if blah: + if new_rgb != old_rgb: + print("%2d: %r -> %r" % (i, old_rgb, new_rgb), file=book.logfile) + +def palette_epilogue(book): + # Check colour indexes in fonts etc. + # This must be done here as FONT records + # come *before* the PALETTE record :-( + for font in book.font_list: + if font.font_index == 4: # the missing font record + continue + cx = font.colour_index + if cx == 0x7fff: # system window text colour + continue + if cx in book.colour_map: + book.colour_indexes_used[cx] = 1 + elif book.verbosity: + print("Size of colour table:", len(book.colour_map), file=book.logfile) + fprintf(book.logfile, "*** Font #%d (%r): colour index 0x%04x is unknown\n", + font.font_index, font.name, cx) + if book.verbosity >= 1: + used = sorted(book.colour_indexes_used.keys()) + print("\nColour indexes used:\n%r\n" % used, file=book.logfile) + +def handle_style(book, data): + if not book.formatting_info: + return + blah = DEBUG or book.verbosity >= 2 + bv = book.biff_version + flag_and_xfx, built_in_id, level = unpack('= 80: + try: + name = unpack_unicode(data, 2, lenlen=2) + except UnicodeDecodeError: + print("STYLE: built_in=%d xf_index=%d built_in_id=%d level=%d" \ + % (built_in, xf_index, built_in_id, level), file=book.logfile) + print("raw bytes:", repr(data[2:]), file=book.logfile) + raise + else: + name = unpack_string(data, 2, book.encoding, lenlen=1) + if blah and not name: + print("WARNING *** A user-defined style has a zero-length name", file=book.logfile) + book.style_name_map[name] = (built_in, xf_index) + if blah: + fprintf(book.logfile, "STYLE: built_in=%d xf_index=%d built_in_id=%d level=%d name=%r\n", + built_in, xf_index, built_in_id, level, name) + +def check_colour_indexes_in_obj(book, obj, orig_index): + alist = sorted(obj.__dict__.items()) + for attr, nobj in alist: + if hasattr(nobj, 'dump'): + check_colour_indexes_in_obj(book, nobj, orig_index) + elif attr.find('colour_index') >= 0: + if nobj in book.colour_map: + book.colour_indexes_used[nobj] = 1 + continue + oname = obj.__class__.__name__ + print("*** xf #%d : %s.%s = 0x%04x (unknown)" \ + % (orig_index, oname, attr, nobj), file=book.logfile) + +def fill_in_standard_formats(book): + for x in std_format_code_types.keys(): + if x not in book.format_map: + ty = std_format_code_types[x] + # Note: many standard format codes (mostly CJK date formats) have + # format strings that vary by locale; xlrd does not (yet) + # handle those; the type (date or numeric) is recorded but the fmt_str will be None. + fmt_str = std_format_strings.get(x) + fmtobj = Format(x, ty, fmt_str) + book.format_map[x] = fmtobj + +def handle_xf(self, data): + ### self is a Book instance + # DEBUG = 0 + blah = DEBUG or self.verbosity >= 3 + bv = self.biff_version + xf = XF() + xf.alignment = XFAlignment() + xf.alignment.indent_level = 0 + xf.alignment.shrink_to_fit = 0 + xf.alignment.text_direction = 0 + xf.border = XFBorder() + xf.border.diag_up = 0 + xf.border.diag_down = 0 + xf.border.diag_colour_index = 0 + xf.border.diag_line_style = 0 # no line + xf.background = XFBackground() + xf.protection = XFProtection() + # fill in the known standard formats + if bv >= 50 and not self.xfcount: + # i.e. do this once before we process the first XF record + fill_in_standard_formats(self) + if bv >= 80: + unpack_fmt = '> 2 + for attr_stem in \ + "format font alignment border background protection".split(): + attr = "_" + attr_stem + "_flag" + setattr(xf, attr, reg & 1) + reg >>= 1 + upkbitsL(xf.border, pkd_brdbkg1, ( + (0, 0x0000000f, 'left_line_style'), + (4, 0x000000f0, 'right_line_style'), + (8, 0x00000f00, 'top_line_style'), + (12, 0x0000f000, 'bottom_line_style'), + (16, 0x007f0000, 'left_colour_index'), + (23, 0x3f800000, 'right_colour_index'), + (30, 0x40000000, 'diag_down'), + (31, 0x80000000, 'diag_up'), + )) + upkbits(xf.border, pkd_brdbkg2, ( + (0, 0x0000007F, 'top_colour_index'), + (7, 0x00003F80, 'bottom_colour_index'), + (14, 0x001FC000, 'diag_colour_index'), + (21, 0x01E00000, 'diag_line_style'), + )) + upkbitsL(xf.background, pkd_brdbkg2, ( + (26, 0xFC000000, 'fill_pattern'), + )) + upkbits(xf.background, pkd_brdbkg3, ( + (0, 0x007F, 'pattern_colour_index'), + (7, 0x3F80, 'background_colour_index'), + )) + elif bv >= 50: + unpack_fmt = '> 2 + for attr_stem in \ + "format font alignment border background protection".split(): + attr = "_" + attr_stem + "_flag" + setattr(xf, attr, reg & 1) + reg >>= 1 + upkbitsL(xf.background, pkd_brdbkg1, ( + ( 0, 0x0000007F, 'pattern_colour_index'), + ( 7, 0x00003F80, 'background_colour_index'), + (16, 0x003F0000, 'fill_pattern'), + )) + upkbitsL(xf.border, pkd_brdbkg1, ( + (22, 0x01C00000, 'bottom_line_style'), + (25, 0xFE000000, 'bottom_colour_index'), + )) + upkbits(xf.border, pkd_brdbkg2, ( + ( 0, 0x00000007, 'top_line_style'), + ( 3, 0x00000038, 'left_line_style'), + ( 6, 0x000001C0, 'right_line_style'), + ( 9, 0x0000FE00, 'top_colour_index'), + (16, 0x007F0000, 'left_colour_index'), + (23, 0x3F800000, 'right_colour_index'), + )) + elif bv >= 40: + unpack_fmt = '> 6 + xf.alignment.rotation = [0, 255, 90, 180][orientation] + reg = pkd_used >> 2 + for attr_stem in \ + "format font alignment border background protection".split(): + attr = "_" + attr_stem + "_flag" + setattr(xf, attr, reg & 1) + reg >>= 1 + upkbits(xf.background, pkd_bkg_34, ( + ( 0, 0x003F, 'fill_pattern'), + ( 6, 0x07C0, 'pattern_colour_index'), + (11, 0xF800, 'background_colour_index'), + )) + upkbitsL(xf.border, pkd_brd_34, ( + ( 0, 0x00000007, 'top_line_style'), + ( 3, 0x000000F8, 'top_colour_index'), + ( 8, 0x00000700, 'left_line_style'), + (11, 0x0000F800, 'left_colour_index'), + (16, 0x00070000, 'bottom_line_style'), + (19, 0x00F80000, 'bottom_colour_index'), + (24, 0x07000000, 'right_line_style'), + (27, 0xF8000000, 'right_colour_index'), + )) + elif bv == 30: + unpack_fmt = '> 2 + for attr_stem in \ + "format font alignment border background protection".split(): + attr = "_" + attr_stem + "_flag" + setattr(xf, attr, reg & 1) + reg >>= 1 + upkbits(xf.background, pkd_bkg_34, ( + ( 0, 0x003F, 'fill_pattern'), + ( 6, 0x07C0, 'pattern_colour_index'), + (11, 0xF800, 'background_colour_index'), + )) + upkbitsL(xf.border, pkd_brd_34, ( + ( 0, 0x00000007, 'top_line_style'), + ( 3, 0x000000F8, 'top_colour_index'), + ( 8, 0x00000700, 'left_line_style'), + (11, 0x0000F800, 'left_colour_index'), + (16, 0x00070000, 'bottom_line_style'), + (19, 0x00F80000, 'bottom_colour_index'), + (24, 0x07000000, 'right_line_style'), + (27, 0xF8000000, 'right_colour_index'), + )) + xf.alignment.vert_align = 2 # bottom + xf.alignment.rotation = 0 + elif bv == 21: + #### Warning: incomplete treatment; formatting_info not fully supported. + #### Probably need to offset incoming BIFF2 XF[n] to BIFF8-like XF[n+16], + #### and create XF[0:16] like the standard ones in BIFF8 + #### *AND* add 16 to all XF references in cell records :-( + (xf.font_index, format_etc, halign_etc) = unpack('= 3 + blah1 = DEBUG or self.verbosity >= 1 + if blah: + fprintf(self.logfile, "xf_epilogue called ...\n") + + def check_same(book_arg, xf_arg, parent_arg, attr): + # the _arg caper is to avoid a Warning msg from Python 2.1 :-( + if getattr(xf_arg, attr) != getattr(parent_arg, attr): + fprintf(book_arg.logfile, + "NOTE !!! XF[%d] parent[%d] %s different\n", + xf_arg.xf_index, parent_arg.xf_index, attr) + + for xfx in xrange(num_xfs): + xf = self.xf_list[xfx] + if xf.format_key not in self.format_map: + msg = "ERROR *** XF[%d] unknown format key (%d, 0x%04x)\n" + fprintf(self.logfile, msg, + xf.xf_index, xf.format_key, xf.format_key) + xf.format_key = 0 + + fmt = self.format_map[xf.format_key] + cellty = _cellty_from_fmtty[fmt.type] + self._xf_index_to_xl_type_map[xf.xf_index] = cellty + # Now for some assertions etc + if not self.formatting_info: + continue + if xf.is_style: + continue + if not(0 <= xf.parent_style_index < num_xfs): + if blah1: + fprintf(self.logfile, + "WARNING *** XF[%d]: is_style=%d but parent_style_index=%d\n", + xf.xf_index, xf.is_style, xf.parent_style_index) + # make it conform + xf.parent_style_index = 0 + if self.biff_version >= 30: + if blah1: + if xf.parent_style_index == xf.xf_index: + fprintf(self.logfile, + "NOTE !!! XF[%d]: parent_style_index is also %d\n", + xf.xf_index, xf.parent_style_index) + elif not self.xf_list[xf.parent_style_index].is_style: + fprintf(self.logfile, + "NOTE !!! XF[%d]: parent_style_index is %d; style flag not set\n", + xf.xf_index, xf.parent_style_index) + if blah1 and xf.parent_style_index > xf.xf_index: + fprintf(self.logfile, + "NOTE !!! XF[%d]: parent_style_index is %d; out of order?\n", + xf.xf_index, xf.parent_style_index) + parent = self.xf_list[xf.parent_style_index] + if not xf._alignment_flag and not parent._alignment_flag: + if blah1: check_same(self, xf, parent, 'alignment') + if not xf._background_flag and not parent._background_flag: + if blah1: check_same(self, xf, parent, 'background') + if not xf._border_flag and not parent._border_flag: + if blah1: check_same(self, xf, parent, 'border') + if not xf._protection_flag and not parent._protection_flag: + if blah1: check_same(self, xf, parent, 'protection') + if not xf._format_flag and not parent._format_flag: + if blah1 and xf.format_key != parent.format_key: + fprintf(self.logfile, + "NOTE !!! XF[%d] fmtk=%d, parent[%d] fmtk=%r\n%r / %r\n", + xf.xf_index, xf.format_key, parent.xf_index, parent.format_key, + self.format_map[xf.format_key].format_str, + self.format_map[parent.format_key].format_str) + if not xf._font_flag and not parent._font_flag: + if blah1 and xf.font_index != parent.font_index: + fprintf(self.logfile, + "NOTE !!! XF[%d] fontx=%d, parent[%d] fontx=%r\n", + xf.xf_index, xf.font_index, parent.xf_index, parent.font_index) + +def initialise_book(book): + initialise_colour_map(book) + book._xf_epilogue_done = 0 + methods = ( + handle_font, + handle_efont, + handle_format, + is_date_format_string, + handle_palette, + palette_epilogue, + handle_style, + handle_xf, + xf_epilogue, + ) + for method in methods: + setattr(book.__class__, method.__name__, method) + +## +#

A collection of the border-related attributes of an XF record. +# Items correspond to those in the Excel UI's Format/Cells/Border tab.

+#

An explanations of "colour index" is given in the Formatting +# section at the start of this document. +# There are five line style attributes; possible values and the +# associated meanings are: +# 0 = No line, +# 1 = Thin, +# 2 = Medium, +# 3 = Dashed, +# 4 = Dotted, +# 5 = Thick, +# 6 = Double, +# 7 = Hair, +# 8 = Medium dashed, +# 9 = Thin dash-dotted, +# 10 = Medium dash-dotted, +# 11 = Thin dash-dot-dotted, +# 12 = Medium dash-dot-dotted, +# 13 = Slanted medium dash-dotted. +# The line styles 8 to 13 appear in BIFF8 files (Excel 97 and later) only. +# For pictures of the line styles, refer to OOo docs s3.10 (p22) +# "Line Styles for Cell Borders (BIFF3-BIFF8)".

+#
-- New in version 0.6.1 +class XFBorder(BaseObject, EqNeAttrs): + + ## + # The colour index for the cell's top line + top_colour_index = 0 + ## + # The colour index for the cell's bottom line + bottom_colour_index = 0 + ## + # The colour index for the cell's left line + left_colour_index = 0 + ## + # The colour index for the cell's right line + right_colour_index = 0 + ## + # The colour index for the cell's diagonal lines, if any + diag_colour_index = 0 + ## + # The line style for the cell's top line + top_line_style = 0 + ## + # The line style for the cell's bottom line + bottom_line_style = 0 + ## + # The line style for the cell's left line + left_line_style = 0 + ## + # The line style for the cell's right line + right_line_style = 0 + ## + # The line style for the cell's diagonal lines, if any + diag_line_style = 0 + ## + # 1 = draw a diagonal from top left to bottom right + diag_down = 0 + ## + # 1 = draw a diagonal from bottom left to top right + diag_up = 0 + +## +# A collection of the background-related attributes of an XF record. +# Items correspond to those in the Excel UI's Format/Cells/Patterns tab. +# An explanation of "colour index" is given in the Formatting +# section at the start of this document. +#
-- New in version 0.6.1 +class XFBackground(BaseObject, EqNeAttrs): + + ## + # See section 3.11 of the OOo docs. + fill_pattern = 0 + ## + # See section 3.11 of the OOo docs. + background_colour_index = 0 + ## + # See section 3.11 of the OOo docs. + pattern_colour_index = 0 + +## +# A collection of the alignment and similar attributes of an XF record. +# Items correspond to those in the Excel UI's Format/Cells/Alignment tab. +#
-- New in version 0.6.1 + +class XFAlignment(BaseObject, EqNeAttrs): + + ## + # Values: section 6.115 (p 214) of OOo docs + hor_align = 0 + ## + # Values: section 6.115 (p 215) of OOo docs + vert_align = 0 + ## + # Values: section 6.115 (p 215) of OOo docs.
+ # Note: file versions BIFF7 and earlier use the documented + # "orientation" attribute; this will be mapped (without loss) + # into "rotation". + rotation = 0 + ## + # 1 = text is wrapped at right margin + text_wrapped = 0 + ## + # A number in range(15). + indent_level = 0 + ## + # 1 = shrink font size to fit text into cell. + shrink_to_fit = 0 + ## + # 0 = according to context; 1 = left-to-right; 2 = right-to-left + text_direction = 0 + +## +# A collection of the protection-related attributes of an XF record. +# Items correspond to those in the Excel UI's Format/Cells/Protection tab. +# Note the OOo docs include the "cell or style" bit +# in this bundle of attributes. +# This is incorrect; the bit is used in determining which bundles to use. +#
-- New in version 0.6.1 + +class XFProtection(BaseObject, EqNeAttrs): + + ## + # 1 = Cell is prevented from being changed, moved, resized, or deleted + # (only if the sheet is protected). + cell_locked = 0 + ## + # 1 = Hide formula so that it doesn't appear in the formula bar when + # the cell is selected (only if the sheet is protected). + formula_hidden = 0 + +## +# eXtended Formatting information for cells, rows, columns and styles. +#
-- New in version 0.6.1 +# +#

Each of the 6 flags below describes the validity of +# a specific group of attributes. +#
+# In cell XFs, flag==0 means the attributes of the parent style XF are used, +# (but only if the attributes are valid there); flag==1 means the attributes +# of this XF are used.
+# In style XFs, flag==0 means the attribute setting is valid; flag==1 means +# the attribute should be ignored.
+# Note that the API +# provides both "raw" XFs and "computed" XFs -- in the latter case, cell XFs +# have had the above inheritance mechanism applied. +#

+ +class XF(BaseObject): + + ## + # 0 = cell XF, 1 = style XF + is_style = 0 + ## + # cell XF: Index into Book.xf_list + # of this XF's style XF
+ # style XF: 0xFFF + parent_style_index = 0 + ## + # + _format_flag = 0 + ## + # + _font_flag = 0 + ## + # + _alignment_flag = 0 + ## + # + _border_flag = 0 + ## + # + _background_flag = 0 + ## + #   + _protection_flag = 0 + ## + # Index into Book.xf_list + xf_index = 0 + ## + # Index into Book.font_list + font_index = 0 + ## + # Key into Book.format_map + #

+ # Warning: OOo docs on the XF record call this "Index to FORMAT record". + # It is not an index in the Python sense. It is a key to a map. + # It is true only for Excel 4.0 and earlier files + # that the key into format_map from an XF instance + # is the same as the index into format_list, and only + # if the index is less than 164. + #

+ format_key = 0 + ## + # An instance of an XFProtection object. + protection = None + ## + # An instance of an XFBackground object. + background = None + ## + # An instance of an XFAlignment object. + alignment = None + ## + # An instance of an XFBorder object. + border = None diff --git a/Assignment 3/packages/xlrd/formula.py b/Assignment 3/packages/xlrd/formula.py new file mode 100644 index 0000000..7c56aa4 --- /dev/null +++ b/Assignment 3/packages/xlrd/formula.py @@ -0,0 +1,2179 @@ +# -*- coding: cp1252 -*- + +## +# Module for parsing/evaluating Microsoft Excel formulas. +# +#

Copyright © 2005-2012 Stephen John Machin, Lingfo Pty Ltd

+#

This module is part of the xlrd package, which is released under +# a BSD-style licence.

+## + +# No part of the content of this file was derived from the works of David Giffin. + +from __future__ import print_function +import copy +from struct import unpack +from .timemachine import * +from .biffh import unpack_unicode_update_pos, unpack_string_update_pos, \ + XLRDError, hex_char_dump, error_text_from_code, BaseObject + +__all__ = [ + 'oBOOL', 'oERR', 'oNUM', 'oREF', 'oREL', 'oSTRG', 'oUNK', + 'decompile_formula', + 'dump_formula', + 'evaluate_name_formula', + 'okind_dict', + 'rangename3d', 'rangename3drel', 'cellname', 'cellnameabs', 'colname', + 'FMLA_TYPE_CELL', + 'FMLA_TYPE_SHARED', + 'FMLA_TYPE_ARRAY', + 'FMLA_TYPE_COND_FMT', + 'FMLA_TYPE_DATA_VAL', + 'FMLA_TYPE_NAME', + ] + +FMLA_TYPE_CELL = 1 +FMLA_TYPE_SHARED = 2 +FMLA_TYPE_ARRAY = 4 +FMLA_TYPE_COND_FMT = 8 +FMLA_TYPE_DATA_VAL = 16 +FMLA_TYPE_NAME = 32 +ALL_FMLA_TYPES = 63 + + +FMLA_TYPEDESCR_MAP = { + 1 : 'CELL', + 2 : 'SHARED', + 4 : 'ARRAY', + 8 : 'COND-FMT', + 16: 'DATA-VAL', + 32: 'NAME', + } + +_TOKEN_NOT_ALLOWED = { + 0x01: ALL_FMLA_TYPES - FMLA_TYPE_CELL, # tExp + 0x02: ALL_FMLA_TYPES - FMLA_TYPE_CELL, # tTbl + 0x0F: FMLA_TYPE_SHARED + FMLA_TYPE_COND_FMT + FMLA_TYPE_DATA_VAL, # tIsect + 0x10: FMLA_TYPE_SHARED + FMLA_TYPE_COND_FMT + FMLA_TYPE_DATA_VAL, # tUnion/List + 0x11: FMLA_TYPE_SHARED + FMLA_TYPE_COND_FMT + FMLA_TYPE_DATA_VAL, # tRange + 0x20: FMLA_TYPE_SHARED + FMLA_TYPE_COND_FMT + FMLA_TYPE_DATA_VAL, # tArray + 0x23: FMLA_TYPE_SHARED, # tName + 0x39: FMLA_TYPE_SHARED + FMLA_TYPE_COND_FMT + FMLA_TYPE_DATA_VAL, # tNameX + 0x3A: FMLA_TYPE_SHARED + FMLA_TYPE_COND_FMT + FMLA_TYPE_DATA_VAL, # tRef3d + 0x3B: FMLA_TYPE_SHARED + FMLA_TYPE_COND_FMT + FMLA_TYPE_DATA_VAL, # tArea3d + 0x2C: FMLA_TYPE_CELL + FMLA_TYPE_ARRAY, # tRefN + 0x2D: FMLA_TYPE_CELL + FMLA_TYPE_ARRAY, # tAreaN + # plus weird stuff like tMem* + }.get + +oBOOL = 3 +oERR = 4 +oMSNG = 5 # tMissArg +oNUM = 2 +oREF = -1 +oREL = -2 +oSTRG = 1 +oUNK = 0 + +okind_dict = { + -2: "oREL", + -1: "oREF", + 0 : "oUNK", + 1 : "oSTRG", + 2 : "oNUM", + 3 : "oBOOL", + 4 : "oERR", + 5 : "oMSNG", + } + +listsep = ',' #### probably should depend on locale + + +# sztabN[opcode] -> the number of bytes to consume. +# -1 means variable +# -2 means this opcode not implemented in this version. +# Which N to use? Depends on biff_version; see szdict. +sztab0 = [-2, 4, 4, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -2, -1, 8, 4, 2, 2, 3, 9, 8, 2, 3, 8, 4, 7, 5, 5, 5, 2, 4, 7, 4, 7, 2, 2, -2, -2, -2, -2, -2, -2, -2, -2, 3, -2, -2, -2, -2, -2, -2, -2] +sztab1 = [-2, 5, 5, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -2, -1, 11, 5, 2, 2, 3, 9, 9, 2, 3, 11, 4, 7, 7, 7, 7, 3, 4, 7, 4, 7, 3, 3, -2, -2, -2, -2, -2, -2, -2, -2, 3, -2, -2, -2, -2, -2, -2, -2] +sztab2 = [-2, 5, 5, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -2, -1, 11, 5, 2, 2, 3, 9, 9, 3, 4, 11, 4, 7, 7, 7, 7, 3, 4, 7, 4, 7, 3, 3, -2, -2, -2, -2, -2, -2, -2, -2, -2, -2, -2, -2, -2, -2, -2, -2] +sztab3 = [-2, 5, 5, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -2, -1, -2, -2, 2, 2, 3, 9, 9, 3, 4, 15, 4, 7, 7, 7, 7, 3, 4, 7, 4, 7, 3, 3, -2, -2, -2, -2, -2, -2, -2, -2, -2, 25, 18, 21, 18, 21, -2, -2] +sztab4 = [-2, 5, 5, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, -1, -2, -2, 2, 2, 3, 9, 9, 3, 4, 5, 5, 9, 7, 7, 7, 3, 5, 9, 5, 9, 3, 3, -2, -2, -2, -2, -2, -2, -2, -2, -2, 7, 7, 11, 7, 11, -2, -2] + +szdict = { + 20 : sztab0, + 21 : sztab0, + 30 : sztab1, + 40 : sztab2, + 45 : sztab2, + 50 : sztab3, + 70 : sztab3, + 80 : sztab4, + } + +# For debugging purposes ... the name for each opcode +# (without the prefix "t" used on OOo docs) +onames = ['Unk00', 'Exp', 'Tbl', 'Add', 'Sub', 'Mul', 'Div', 'Power', 'Concat', 'LT', 'LE', 'EQ', 'GE', 'GT', 'NE', 'Isect', 'List', 'Range', 'Uplus', 'Uminus', 'Percent', 'Paren', 'MissArg', 'Str', 'Extended', 'Attr', 'Sheet', 'EndSheet', 'Err', 'Bool', 'Int', 'Num', 'Array', 'Func', 'FuncVar', 'Name', 'Ref', 'Area', 'MemArea', 'MemErr', 'MemNoMem', 'MemFunc', 'RefErr', 'AreaErr', 'RefN', 'AreaN', 'MemAreaN', 'MemNoMemN', '', '', '', '', '', '', '', '', 'FuncCE', 'NameX', 'Ref3d', 'Area3d', 'RefErr3d', 'AreaErr3d', '', ''] + +func_defs = { + # index: (name, min#args, max#args, flags, #known_args, return_type, kargs) + 0 : ('COUNT', 0, 30, 0x04, 1, 'V', 'R'), + 1 : ('IF', 2, 3, 0x04, 3, 'V', 'VRR'), + 2 : ('ISNA', 1, 1, 0x02, 1, 'V', 'V'), + 3 : ('ISERROR', 1, 1, 0x02, 1, 'V', 'V'), + 4 : ('SUM', 0, 30, 0x04, 1, 'V', 'R'), + 5 : ('AVERAGE', 1, 30, 0x04, 1, 'V', 'R'), + 6 : ('MIN', 1, 30, 0x04, 1, 'V', 'R'), + 7 : ('MAX', 1, 30, 0x04, 1, 'V', 'R'), + 8 : ('ROW', 0, 1, 0x04, 1, 'V', 'R'), + 9 : ('COLUMN', 0, 1, 0x04, 1, 'V', 'R'), + 10 : ('NA', 0, 0, 0x02, 0, 'V', ''), + 11 : ('NPV', 2, 30, 0x04, 2, 'V', 'VR'), + 12 : ('STDEV', 1, 30, 0x04, 1, 'V', 'R'), + 13 : ('DOLLAR', 1, 2, 0x04, 1, 'V', 'V'), + 14 : ('FIXED', 2, 3, 0x04, 3, 'V', 'VVV'), + 15 : ('SIN', 1, 1, 0x02, 1, 'V', 'V'), + 16 : ('COS', 1, 1, 0x02, 1, 'V', 'V'), + 17 : ('TAN', 1, 1, 0x02, 1, 'V', 'V'), + 18 : ('ATAN', 1, 1, 0x02, 1, 'V', 'V'), + 19 : ('PI', 0, 0, 0x02, 0, 'V', ''), + 20 : ('SQRT', 1, 1, 0x02, 1, 'V', 'V'), + 21 : ('EXP', 1, 1, 0x02, 1, 'V', 'V'), + 22 : ('LN', 1, 1, 0x02, 1, 'V', 'V'), + 23 : ('LOG10', 1, 1, 0x02, 1, 'V', 'V'), + 24 : ('ABS', 1, 1, 0x02, 1, 'V', 'V'), + 25 : ('INT', 1, 1, 0x02, 1, 'V', 'V'), + 26 : ('SIGN', 1, 1, 0x02, 1, 'V', 'V'), + 27 : ('ROUND', 2, 2, 0x02, 2, 'V', 'VV'), + 28 : ('LOOKUP', 2, 3, 0x04, 2, 'V', 'VR'), + 29 : ('INDEX', 2, 4, 0x0c, 4, 'R', 'RVVV'), + 30 : ('REPT', 2, 2, 0x02, 2, 'V', 'VV'), + 31 : ('MID', 3, 3, 0x02, 3, 'V', 'VVV'), + 32 : ('LEN', 1, 1, 0x02, 1, 'V', 'V'), + 33 : ('VALUE', 1, 1, 0x02, 1, 'V', 'V'), + 34 : ('TRUE', 0, 0, 0x02, 0, 'V', ''), + 35 : ('FALSE', 0, 0, 0x02, 0, 'V', ''), + 36 : ('AND', 1, 30, 0x04, 1, 'V', 'R'), + 37 : ('OR', 1, 30, 0x04, 1, 'V', 'R'), + 38 : ('NOT', 1, 1, 0x02, 1, 'V', 'V'), + 39 : ('MOD', 2, 2, 0x02, 2, 'V', 'VV'), + 40 : ('DCOUNT', 3, 3, 0x02, 3, 'V', 'RRR'), + 41 : ('DSUM', 3, 3, 0x02, 3, 'V', 'RRR'), + 42 : ('DAVERAGE', 3, 3, 0x02, 3, 'V', 'RRR'), + 43 : ('DMIN', 3, 3, 0x02, 3, 'V', 'RRR'), + 44 : ('DMAX', 3, 3, 0x02, 3, 'V', 'RRR'), + 45 : ('DSTDEV', 3, 3, 0x02, 3, 'V', 'RRR'), + 46 : ('VAR', 1, 30, 0x04, 1, 'V', 'R'), + 47 : ('DVAR', 3, 3, 0x02, 3, 'V', 'RRR'), + 48 : ('TEXT', 2, 2, 0x02, 2, 'V', 'VV'), + 49 : ('LINEST', 1, 4, 0x04, 4, 'A', 'RRVV'), + 50 : ('TREND', 1, 4, 0x04, 4, 'A', 'RRRV'), + 51 : ('LOGEST', 1, 4, 0x04, 4, 'A', 'RRVV'), + 52 : ('GROWTH', 1, 4, 0x04, 4, 'A', 'RRRV'), + 56 : ('PV', 3, 5, 0x04, 5, 'V', 'VVVVV'), + 57 : ('FV', 3, 5, 0x04, 5, 'V', 'VVVVV'), + 58 : ('NPER', 3, 5, 0x04, 5, 'V', 'VVVVV'), + 59 : ('PMT', 3, 5, 0x04, 5, 'V', 'VVVVV'), + 60 : ('RATE', 3, 6, 0x04, 6, 'V', 'VVVVVV'), + 61 : ('MIRR', 3, 3, 0x02, 3, 'V', 'RVV'), + 62 : ('IRR', 1, 2, 0x04, 2, 'V', 'RV'), + 63 : ('RAND', 0, 0, 0x0a, 0, 'V', ''), + 64 : ('MATCH', 2, 3, 0x04, 3, 'V', 'VRR'), + 65 : ('DATE', 3, 3, 0x02, 3, 'V', 'VVV'), + 66 : ('TIME', 3, 3, 0x02, 3, 'V', 'VVV'), + 67 : ('DAY', 1, 1, 0x02, 1, 'V', 'V'), + 68 : ('MONTH', 1, 1, 0x02, 1, 'V', 'V'), + 69 : ('YEAR', 1, 1, 0x02, 1, 'V', 'V'), + 70 : ('WEEKDAY', 1, 2, 0x04, 2, 'V', 'VV'), + 71 : ('HOUR', 1, 1, 0x02, 1, 'V', 'V'), + 72 : ('MINUTE', 1, 1, 0x02, 1, 'V', 'V'), + 73 : ('SECOND', 1, 1, 0x02, 1, 'V', 'V'), + 74 : ('NOW', 0, 0, 0x0a, 0, 'V', ''), + 75 : ('AREAS', 1, 1, 0x02, 1, 'V', 'R'), + 76 : ('ROWS', 1, 1, 0x02, 1, 'V', 'R'), + 77 : ('COLUMNS', 1, 1, 0x02, 1, 'V', 'R'), + 78 : ('OFFSET', 3, 5, 0x04, 5, 'R', 'RVVVV'), + 82 : ('SEARCH', 2, 3, 0x04, 3, 'V', 'VVV'), + 83 : ('TRANSPOSE', 1, 1, 0x02, 1, 'A', 'A'), + 86 : ('TYPE', 1, 1, 0x02, 1, 'V', 'V'), + 92 : ('SERIESSUM', 4, 4, 0x02, 4, 'V', 'VVVA'), + 97 : ('ATAN2', 2, 2, 0x02, 2, 'V', 'VV'), + 98 : ('ASIN', 1, 1, 0x02, 1, 'V', 'V'), + 99 : ('ACOS', 1, 1, 0x02, 1, 'V', 'V'), + 100: ('CHOOSE', 2, 30, 0x04, 2, 'V', 'VR'), + 101: ('HLOOKUP', 3, 4, 0x04, 4, 'V', 'VRRV'), + 102: ('VLOOKUP', 3, 4, 0x04, 4, 'V', 'VRRV'), + 105: ('ISREF', 1, 1, 0x02, 1, 'V', 'R'), + 109: ('LOG', 1, 2, 0x04, 2, 'V', 'VV'), + 111: ('CHAR', 1, 1, 0x02, 1, 'V', 'V'), + 112: ('LOWER', 1, 1, 0x02, 1, 'V', 'V'), + 113: ('UPPER', 1, 1, 0x02, 1, 'V', 'V'), + 114: ('PROPER', 1, 1, 0x02, 1, 'V', 'V'), + 115: ('LEFT', 1, 2, 0x04, 2, 'V', 'VV'), + 116: ('RIGHT', 1, 2, 0x04, 2, 'V', 'VV'), + 117: ('EXACT', 2, 2, 0x02, 2, 'V', 'VV'), + 118: ('TRIM', 1, 1, 0x02, 1, 'V', 'V'), + 119: ('REPLACE', 4, 4, 0x02, 4, 'V', 'VVVV'), + 120: ('SUBSTITUTE', 3, 4, 0x04, 4, 'V', 'VVVV'), + 121: ('CODE', 1, 1, 0x02, 1, 'V', 'V'), + 124: ('FIND', 2, 3, 0x04, 3, 'V', 'VVV'), + 125: ('CELL', 1, 2, 0x0c, 2, 'V', 'VR'), + 126: ('ISERR', 1, 1, 0x02, 1, 'V', 'V'), + 127: ('ISTEXT', 1, 1, 0x02, 1, 'V', 'V'), + 128: ('ISNUMBER', 1, 1, 0x02, 1, 'V', 'V'), + 129: ('ISBLANK', 1, 1, 0x02, 1, 'V', 'V'), + 130: ('T', 1, 1, 0x02, 1, 'V', 'R'), + 131: ('N', 1, 1, 0x02, 1, 'V', 'R'), + 140: ('DATEVALUE', 1, 1, 0x02, 1, 'V', 'V'), + 141: ('TIMEVALUE', 1, 1, 0x02, 1, 'V', 'V'), + 142: ('SLN', 3, 3, 0x02, 3, 'V', 'VVV'), + 143: ('SYD', 4, 4, 0x02, 4, 'V', 'VVVV'), + 144: ('DDB', 4, 5, 0x04, 5, 'V', 'VVVVV'), + 148: ('INDIRECT', 1, 2, 0x0c, 2, 'R', 'VV'), + 162: ('CLEAN', 1, 1, 0x02, 1, 'V', 'V'), + 163: ('MDETERM', 1, 1, 0x02, 1, 'V', 'A'), + 164: ('MINVERSE', 1, 1, 0x02, 1, 'A', 'A'), + 165: ('MMULT', 2, 2, 0x02, 2, 'A', 'AA'), + 167: ('IPMT', 4, 6, 0x04, 6, 'V', 'VVVVVV'), + 168: ('PPMT', 4, 6, 0x04, 6, 'V', 'VVVVVV'), + 169: ('COUNTA', 0, 30, 0x04, 1, 'V', 'R'), + 183: ('PRODUCT', 0, 30, 0x04, 1, 'V', 'R'), + 184: ('FACT', 1, 1, 0x02, 1, 'V', 'V'), + 189: ('DPRODUCT', 3, 3, 0x02, 3, 'V', 'RRR'), + 190: ('ISNONTEXT', 1, 1, 0x02, 1, 'V', 'V'), + 193: ('STDEVP', 1, 30, 0x04, 1, 'V', 'R'), + 194: ('VARP', 1, 30, 0x04, 1, 'V', 'R'), + 195: ('DSTDEVP', 3, 3, 0x02, 3, 'V', 'RRR'), + 196: ('DVARP', 3, 3, 0x02, 3, 'V', 'RRR'), + 197: ('TRUNC', 1, 2, 0x04, 2, 'V', 'VV'), + 198: ('ISLOGICAL', 1, 1, 0x02, 1, 'V', 'V'), + 199: ('DCOUNTA', 3, 3, 0x02, 3, 'V', 'RRR'), + 204: ('USDOLLAR', 1, 2, 0x04, 2, 'V', 'VV'), + 205: ('FINDB', 2, 3, 0x04, 3, 'V', 'VVV'), + 206: ('SEARCHB', 2, 3, 0x04, 3, 'V', 'VVV'), + 207: ('REPLACEB', 4, 4, 0x02, 4, 'V', 'VVVV'), + 208: ('LEFTB', 1, 2, 0x04, 2, 'V', 'VV'), + 209: ('RIGHTB', 1, 2, 0x04, 2, 'V', 'VV'), + 210: ('MIDB', 3, 3, 0x02, 3, 'V', 'VVV'), + 211: ('LENB', 1, 1, 0x02, 1, 'V', 'V'), + 212: ('ROUNDUP', 2, 2, 0x02, 2, 'V', 'VV'), + 213: ('ROUNDDOWN', 2, 2, 0x02, 2, 'V', 'VV'), + 214: ('ASC', 1, 1, 0x02, 1, 'V', 'V'), + 215: ('DBCS', 1, 1, 0x02, 1, 'V', 'V'), + 216: ('RANK', 2, 3, 0x04, 3, 'V', 'VRV'), + 219: ('ADDRESS', 2, 5, 0x04, 5, 'V', 'VVVVV'), + 220: ('DAYS360', 2, 3, 0x04, 3, 'V', 'VVV'), + 221: ('TODAY', 0, 0, 0x0a, 0, 'V', ''), + 222: ('VDB', 5, 7, 0x04, 7, 'V', 'VVVVVVV'), + 227: ('MEDIAN', 1, 30, 0x04, 1, 'V', 'R'), + 228: ('SUMPRODUCT', 1, 30, 0x04, 1, 'V', 'A'), + 229: ('SINH', 1, 1, 0x02, 1, 'V', 'V'), + 230: ('COSH', 1, 1, 0x02, 1, 'V', 'V'), + 231: ('TANH', 1, 1, 0x02, 1, 'V', 'V'), + 232: ('ASINH', 1, 1, 0x02, 1, 'V', 'V'), + 233: ('ACOSH', 1, 1, 0x02, 1, 'V', 'V'), + 234: ('ATANH', 1, 1, 0x02, 1, 'V', 'V'), + 235: ('DGET', 3, 3, 0x02, 3, 'V', 'RRR'), + 244: ('INFO', 1, 1, 0x02, 1, 'V', 'V'), + 247: ('DB', 4, 5, 0x04, 5, 'V', 'VVVVV'), + 252: ('FREQUENCY', 2, 2, 0x02, 2, 'A', 'RR'), + 261: ('ERROR.TYPE', 1, 1, 0x02, 1, 'V', 'V'), + 269: ('AVEDEV', 1, 30, 0x04, 1, 'V', 'R'), + 270: ('BETADIST', 3, 5, 0x04, 1, 'V', 'V'), + 271: ('GAMMALN', 1, 1, 0x02, 1, 'V', 'V'), + 272: ('BETAINV', 3, 5, 0x04, 1, 'V', 'V'), + 273: ('BINOMDIST', 4, 4, 0x02, 4, 'V', 'VVVV'), + 274: ('CHIDIST', 2, 2, 0x02, 2, 'V', 'VV'), + 275: ('CHIINV', 2, 2, 0x02, 2, 'V', 'VV'), + 276: ('COMBIN', 2, 2, 0x02, 2, 'V', 'VV'), + 277: ('CONFIDENCE', 3, 3, 0x02, 3, 'V', 'VVV'), + 278: ('CRITBINOM', 3, 3, 0x02, 3, 'V', 'VVV'), + 279: ('EVEN', 1, 1, 0x02, 1, 'V', 'V'), + 280: ('EXPONDIST', 3, 3, 0x02, 3, 'V', 'VVV'), + 281: ('FDIST', 3, 3, 0x02, 3, 'V', 'VVV'), + 282: ('FINV', 3, 3, 0x02, 3, 'V', 'VVV'), + 283: ('FISHER', 1, 1, 0x02, 1, 'V', 'V'), + 284: ('FISHERINV', 1, 1, 0x02, 1, 'V', 'V'), + 285: ('FLOOR', 2, 2, 0x02, 2, 'V', 'VV'), + 286: ('GAMMADIST', 4, 4, 0x02, 4, 'V', 'VVVV'), + 287: ('GAMMAINV', 3, 3, 0x02, 3, 'V', 'VVV'), + 288: ('CEILING', 2, 2, 0x02, 2, 'V', 'VV'), + 289: ('HYPGEOMDIST', 4, 4, 0x02, 4, 'V', 'VVVV'), + 290: ('LOGNORMDIST', 3, 3, 0x02, 3, 'V', 'VVV'), + 291: ('LOGINV', 3, 3, 0x02, 3, 'V', 'VVV'), + 292: ('NEGBINOMDIST', 3, 3, 0x02, 3, 'V', 'VVV'), + 293: ('NORMDIST', 4, 4, 0x02, 4, 'V', 'VVVV'), + 294: ('NORMSDIST', 1, 1, 0x02, 1, 'V', 'V'), + 295: ('NORMINV', 3, 3, 0x02, 3, 'V', 'VVV'), + 296: ('NORMSINV', 1, 1, 0x02, 1, 'V', 'V'), + 297: ('STANDARDIZE', 3, 3, 0x02, 3, 'V', 'VVV'), + 298: ('ODD', 1, 1, 0x02, 1, 'V', 'V'), + 299: ('PERMUT', 2, 2, 0x02, 2, 'V', 'VV'), + 300: ('POISSON', 3, 3, 0x02, 3, 'V', 'VVV'), + 301: ('TDIST', 3, 3, 0x02, 3, 'V', 'VVV'), + 302: ('WEIBULL', 4, 4, 0x02, 4, 'V', 'VVVV'), + 303: ('SUMXMY2', 2, 2, 0x02, 2, 'V', 'AA'), + 304: ('SUMX2MY2', 2, 2, 0x02, 2, 'V', 'AA'), + 305: ('SUMX2PY2', 2, 2, 0x02, 2, 'V', 'AA'), + 306: ('CHITEST', 2, 2, 0x02, 2, 'V', 'AA'), + 307: ('CORREL', 2, 2, 0x02, 2, 'V', 'AA'), + 308: ('COVAR', 2, 2, 0x02, 2, 'V', 'AA'), + 309: ('FORECAST', 3, 3, 0x02, 3, 'V', 'VAA'), + 310: ('FTEST', 2, 2, 0x02, 2, 'V', 'AA'), + 311: ('INTERCEPT', 2, 2, 0x02, 2, 'V', 'AA'), + 312: ('PEARSON', 2, 2, 0x02, 2, 'V', 'AA'), + 313: ('RSQ', 2, 2, 0x02, 2, 'V', 'AA'), + 314: ('STEYX', 2, 2, 0x02, 2, 'V', 'AA'), + 315: ('SLOPE', 2, 2, 0x02, 2, 'V', 'AA'), + 316: ('TTEST', 4, 4, 0x02, 4, 'V', 'AAVV'), + 317: ('PROB', 3, 4, 0x04, 3, 'V', 'AAV'), + 318: ('DEVSQ', 1, 30, 0x04, 1, 'V', 'R'), + 319: ('GEOMEAN', 1, 30, 0x04, 1, 'V', 'R'), + 320: ('HARMEAN', 1, 30, 0x04, 1, 'V', 'R'), + 321: ('SUMSQ', 0, 30, 0x04, 1, 'V', 'R'), + 322: ('KURT', 1, 30, 0x04, 1, 'V', 'R'), + 323: ('SKEW', 1, 30, 0x04, 1, 'V', 'R'), + 324: ('ZTEST', 2, 3, 0x04, 2, 'V', 'RV'), + 325: ('LARGE', 2, 2, 0x02, 2, 'V', 'RV'), + 326: ('SMALL', 2, 2, 0x02, 2, 'V', 'RV'), + 327: ('QUARTILE', 2, 2, 0x02, 2, 'V', 'RV'), + 328: ('PERCENTILE', 2, 2, 0x02, 2, 'V', 'RV'), + 329: ('PERCENTRANK', 2, 3, 0x04, 2, 'V', 'RV'), + 330: ('MODE', 1, 30, 0x04, 1, 'V', 'A'), + 331: ('TRIMMEAN', 2, 2, 0x02, 2, 'V', 'RV'), + 332: ('TINV', 2, 2, 0x02, 2, 'V', 'VV'), + 336: ('CONCATENATE', 0, 30, 0x04, 1, 'V', 'V'), + 337: ('POWER', 2, 2, 0x02, 2, 'V', 'VV'), + 342: ('RADIANS', 1, 1, 0x02, 1, 'V', 'V'), + 343: ('DEGREES', 1, 1, 0x02, 1, 'V', 'V'), + 344: ('SUBTOTAL', 2, 30, 0x04, 2, 'V', 'VR'), + 345: ('SUMIF', 2, 3, 0x04, 3, 'V', 'RVR'), + 346: ('COUNTIF', 2, 2, 0x02, 2, 'V', 'RV'), + 347: ('COUNTBLANK', 1, 1, 0x02, 1, 'V', 'R'), + 350: ('ISPMT', 4, 4, 0x02, 4, 'V', 'VVVV'), + 351: ('DATEDIF', 3, 3, 0x02, 3, 'V', 'VVV'), + 352: ('DATESTRING', 1, 1, 0x02, 1, 'V', 'V'), + 353: ('NUMBERSTRING', 2, 2, 0x02, 2, 'V', 'VV'), + 354: ('ROMAN', 1, 2, 0x04, 2, 'V', 'VV'), + 358: ('GETPIVOTDATA', 2, 2, 0x02, 2, 'V', 'RV'), + 359: ('HYPERLINK', 1, 2, 0x04, 2, 'V', 'VV'), + 360: ('PHONETIC', 1, 1, 0x02, 1, 'V', 'V'), + 361: ('AVERAGEA', 1, 30, 0x04, 1, 'V', 'R'), + 362: ('MAXA', 1, 30, 0x04, 1, 'V', 'R'), + 363: ('MINA', 1, 30, 0x04, 1, 'V', 'R'), + 364: ('STDEVPA', 1, 30, 0x04, 1, 'V', 'R'), + 365: ('VARPA', 1, 30, 0x04, 1, 'V', 'R'), + 366: ('STDEVA', 1, 30, 0x04, 1, 'V', 'R'), + 367: ('VARA', 1, 30, 0x04, 1, 'V', 'R'), + 368: ('BAHTTEXT', 1, 1, 0x02, 1, 'V', 'V'), + 369: ('THAIDAYOFWEEK', 1, 1, 0x02, 1, 'V', 'V'), + 370: ('THAIDIGIT', 1, 1, 0x02, 1, 'V', 'V'), + 371: ('THAIMONTHOFYEAR', 1, 1, 0x02, 1, 'V', 'V'), + 372: ('THAINUMSOUND', 1, 1, 0x02, 1, 'V', 'V'), + 373: ('THAINUMSTRING', 1, 1, 0x02, 1, 'V', 'V'), + 374: ('THAISTRINGLENGTH', 1, 1, 0x02, 1, 'V', 'V'), + 375: ('ISTHAIDIGIT', 1, 1, 0x02, 1, 'V', 'V'), + 376: ('ROUNDBAHTDOWN', 1, 1, 0x02, 1, 'V', 'V'), + 377: ('ROUNDBAHTUP', 1, 1, 0x02, 1, 'V', 'V'), + 378: ('THAIYEAR', 1, 1, 0x02, 1, 'V', 'V'), + 379: ('RTD', 2, 5, 0x04, 1, 'V', 'V'), + } + +tAttrNames = { + 0x00: "Skip??", # seen in SAMPLES.XLS which shipped with Excel 5.0 + 0x01: "Volatile", + 0x02: "If", + 0x04: "Choose", + 0x08: "Skip", + 0x10: "Sum", + 0x20: "Assign", + 0x40: "Space", + 0x41: "SpaceVolatile", + } + +error_opcodes = set([0x07, 0x08, 0x0A, 0x0B, 0x1C, 0x1D, 0x2F]) + +tRangeFuncs = (min, max, min, max, min, max) +tIsectFuncs = (max, min, max, min, max, min) + +def do_box_funcs(box_funcs, boxa, boxb): + return tuple([ + func(numa, numb) + for func, numa, numb in zip(box_funcs, boxa.coords, boxb.coords) + ]) + +def adjust_cell_addr_biff8(rowval, colval, reldelta, browx=None, bcolx=None): + row_rel = (colval >> 15) & 1 + col_rel = (colval >> 14) & 1 + rowx = rowval + colx = colval & 0xff + if reldelta: + if row_rel and rowx >= 32768: + rowx -= 65536 + if col_rel and colx >= 128: + colx -= 256 + else: + if row_rel: + rowx -= browx + if col_rel: + colx -= bcolx + return rowx, colx, row_rel, col_rel + +def adjust_cell_addr_biff_le7( + rowval, colval, reldelta, browx=None, bcolx=None): + row_rel = (rowval >> 15) & 1 + col_rel = (rowval >> 14) & 1 + rowx = rowval & 0x3fff + colx = colval + if reldelta: + if row_rel and rowx >= 8192: + rowx -= 16384 + if col_rel and colx >= 128: + colx -= 256 + else: + if row_rel: + rowx -= browx + if col_rel: + colx -= bcolx + return rowx, colx, row_rel, col_rel + +def get_cell_addr(data, pos, bv, reldelta, browx=None, bcolx=None): + if bv >= 80: + rowval, colval = unpack("= 80: + row1val, row2val, col1val, col2val = unpack(" addins %r" % (refx, info), file=bk.logfile) + assert ref_first_sheetx == 0xFFFE == ref_last_sheetx + return (-5, -5) + if ref_recordx != bk._supbook_locals_inx: + if blah: + print("/// get_externsheet_local_range(refx=%d) -> external %r" % (refx, info), file=bk.logfile) + return (-4, -4) # external reference + if ref_first_sheetx == 0xFFFE == ref_last_sheetx: + if blah: + print("/// get_externsheet_local_range(refx=%d) -> unspecified sheet %r" % (refx, info), file=bk.logfile) + return (-1, -1) # internal reference, any sheet + if ref_first_sheetx == 0xFFFF == ref_last_sheetx: + if blah: + print("/// get_externsheet_local_range(refx=%d) -> deleted sheet(s)" % (refx, ), file=bk.logfile) + return (-2, -2) # internal reference, deleted sheet(s) + nsheets = len(bk._all_sheets_map) + if not(0 <= ref_first_sheetx <= ref_last_sheetx < nsheets): + if blah: + print("/// get_externsheet_local_range(refx=%d) -> %r" % (refx, info), file=bk.logfile) + print("--- first/last sheet not in range(%d)" % nsheets, file=bk.logfile) + return (-102, -102) # stuffed up somewhere :-( + xlrd_sheetx1 = bk._all_sheets_map[ref_first_sheetx] + xlrd_sheetx2 = bk._all_sheets_map[ref_last_sheetx] + if not(0 <= xlrd_sheetx1 <= xlrd_sheetx2): + return (-3, -3) # internal reference, but to a macro sheet + return xlrd_sheetx1, xlrd_sheetx2 + +def get_externsheet_local_range_b57( + bk, raw_extshtx, ref_first_sheetx, ref_last_sheetx, blah=0): + if raw_extshtx > 0: + if blah: + print("/// get_externsheet_local_range_b57(raw_extshtx=%d) -> external" % raw_extshtx, file=bk.logfile) + return (-4, -4) # external reference + if ref_first_sheetx == -1 and ref_last_sheetx == -1: + return (-2, -2) # internal reference, deleted sheet(s) + nsheets = len(bk._all_sheets_map) + if not(0 <= ref_first_sheetx <= ref_last_sheetx < nsheets): + if blah: + print("/// get_externsheet_local_range_b57(%d, %d, %d) -> ???" \ + % (raw_extshtx, ref_first_sheetx, ref_last_sheetx), file=bk.logfile) + print("--- first/last sheet not in range(%d)" % nsheets, file=bk.logfile) + return (-103, -103) # stuffed up somewhere :-( + xlrd_sheetx1 = bk._all_sheets_map[ref_first_sheetx] + xlrd_sheetx2 = bk._all_sheets_map[ref_last_sheetx] + if not(0 <= xlrd_sheetx1 <= xlrd_sheetx2): + return (-3, -3) # internal reference, but to a macro sheet + return xlrd_sheetx1, xlrd_sheetx2 + +class FormulaError(Exception): + pass + + +## +# Used in evaluating formulas. +# The following table describes the kinds and how their values +# are represented.

+# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +#
Kind symbolKind numberValue representation
oBOOL3integer: 0 => False; 1 => True
oERR4None, or an int error code (same as XL_CELL_ERROR in the Cell class). +#
oMSNG5Used by Excel as a placeholder for a missing (not supplied) function +# argument. Should *not* appear as a final formula result. Value is None.
oNUM2A float. Note that there is no way of distinguishing dates.
oREF-1The value is either None or a non-empty list of +# absolute Ref3D instances.
+#
oREL-2The value is None or a non-empty list of +# fully or partially relative Ref3D instances. +#
oSTRG1A Unicode string.
oUNK0The kind is unknown or ambiguous. The value is None
+#

+ +class Operand(object): + + ## + # None means that the actual value of the operand is a variable + # (depends on cell data), not a constant. + value = None + ## + # oUNK means that the kind of operand is not known unambiguously. + kind = oUNK + ## + # The reconstituted text of the original formula. Function names will be + # in English irrespective of the original language, which doesn't seem + # to be recorded anywhere. The separator is ",", not ";" or whatever else + # might be more appropriate for the end-user's locale; patches welcome. + text = '?' + + def __init__(self, akind=None, avalue=None, arank=0, atext='?'): + if akind is not None: + self.kind = akind + if avalue is not None: + self.value = avalue + self.rank = arank + # rank is an internal gizmo (operator precedence); + # it's used in reconstructing formula text. + self.text = atext + + def __repr__(self): + kind_text = okind_dict.get(self.kind, "?Unknown kind?") + return "Operand(kind=%s, value=%r, text=%r)" \ + % (kind_text, self.value, self.text) + +## +#

Represents an absolute or relative 3-dimensional reference to a box +# of one or more cells.
+# -- New in version 0.6.0 +#

+# +#

The coords attribute is a tuple of the form:
+# (shtxlo, shtxhi, rowxlo, rowxhi, colxlo, colxhi)
+# where 0 <= thingxlo <= thingx < thingxhi.
+# Note that it is quite possible to have thingx > nthings; for example +# Print_Titles could have colxhi == 256 and/or rowxhi == 65536 +# irrespective of how many columns/rows are actually used in the worksheet. +# The caller will need to decide how to handle this situation. +# Keyword: IndexError :-) +#

+# +#

The components of the coords attribute are also available as individual +# attributes: shtxlo, shtxhi, rowxlo, rowxhi, colxlo, and colxhi.

+# +#

The relflags attribute is a 6-tuple of flags which indicate whether +# the corresponding (sheet|row|col)(lo|hi) is relative (1) or absolute (0).
+# Note that there is necessarily no information available as to what cell(s) +# the reference could possibly be relative to. The caller must decide what if +# any use to make of oREL operands. Note also that a partially relative +# reference may well be a typo. +# For example, define name A1Z10 as $a$1:$z10 (missing $ after z) +# while the cursor is on cell Sheet3!A27.
+# The resulting Ref3D instance will have coords = (2, 3, 0, -16, 0, 26) +# and relflags = (0, 0, 0, 1, 0, 0).
+# So far, only one possibility of a sheet-relative component in +# a reference has been noticed: a 2D reference located in the "current sheet". +#
This will appear as coords = (0, 1, ...) and relflags = (1, 1, ...). + +class Ref3D(tuple): + + def __init__(self, atuple): + self.coords = atuple[0:6] + self.relflags = atuple[6:12] + if not self.relflags: + self.relflags = (0, 0, 0, 0, 0, 0) + (self.shtxlo, self.shtxhi, + self.rowxlo, self.rowxhi, + self.colxlo, self.colxhi) = self.coords + + def __repr__(self): + if not self.relflags or self.relflags == (0, 0, 0, 0, 0, 0): + return "Ref3D(coords=%r)" % (self.coords, ) + else: + return "Ref3D(coords=%r, relflags=%r)" \ + % (self.coords, self.relflags) + +tAdd = 0x03 +tSub = 0x04 +tMul = 0x05 +tDiv = 0x06 +tPower = 0x07 +tConcat = 0x08 +tLT, tLE, tEQ, tGE, tGT, tNE = range(0x09, 0x0F) + +import operator as opr + +def nop(x): + return x + +def _opr_pow(x, y): return x ** y + +def _opr_lt(x, y): return x < y +def _opr_le(x, y): return x <= y +def _opr_eq(x, y): return x == y +def _opr_ge(x, y): return x >= y +def _opr_gt(x, y): return x > y +def _opr_ne(x, y): return x != y + +def num2strg(num): + """Attempt to emulate Excel's default conversion + from number to string. + """ + s = str(num) + if s.endswith(".0"): + s = s[:-2] + return s + +_arith_argdict = {oNUM: nop, oSTRG: float} +_cmp_argdict = {oNUM: nop, oSTRG: nop} +# Seems no conversions done on relops; in Excel, "1" > 9 produces TRUE. +_strg_argdict = {oNUM:num2strg, oSTRG:nop} +binop_rules = { + tAdd: (_arith_argdict, oNUM, opr.add, 30, '+'), + tSub: (_arith_argdict, oNUM, opr.sub, 30, '-'), + tMul: (_arith_argdict, oNUM, opr.mul, 40, '*'), + tDiv: (_arith_argdict, oNUM, opr.truediv, 40, '/'), + tPower: (_arith_argdict, oNUM, _opr_pow, 50, '^',), + tConcat:(_strg_argdict, oSTRG, opr.add, 20, '&'), + tLT: (_cmp_argdict, oBOOL, _opr_lt, 10, '<'), + tLE: (_cmp_argdict, oBOOL, _opr_le, 10, '<='), + tEQ: (_cmp_argdict, oBOOL, _opr_eq, 10, '='), + tGE: (_cmp_argdict, oBOOL, _opr_ge, 10, '>='), + tGT: (_cmp_argdict, oBOOL, _opr_gt, 10, '>'), + tNE: (_cmp_argdict, oBOOL, _opr_ne, 10, '<>'), + } + +unop_rules = { + 0x13: (lambda x: -x, 70, '-', ''), # unary minus + 0x12: (lambda x: x, 70, '+', ''), # unary plus + 0x14: (lambda x: x / 100.0, 60, '', '%'),# percent + } + +LEAF_RANK = 90 +FUNC_RANK = 90 + +STACK_ALARM_LEVEL = 5 +STACK_PANIC_LEVEL = 10 + +def evaluate_name_formula(bk, nobj, namex, blah=0, level=0): + if level > STACK_ALARM_LEVEL: + blah = 1 + data = nobj.raw_formula + fmlalen = nobj.basic_formula_len + bv = bk.biff_version + reldelta = 1 # All defined name formulas use "Method B" [OOo docs] + if blah: + print("::: evaluate_name_formula %r %r %d %d %r level=%d" \ + % (namex, nobj.name, fmlalen, bv, data, level), file=bk.logfile) + hex_char_dump(data, 0, fmlalen, fout=bk.logfile) + if level > STACK_PANIC_LEVEL: + raise XLRDError("Excessive indirect references in NAME formula") + sztab = szdict[bv] + pos = 0 + stack = [] + any_rel = 0 + any_err = 0 + any_external = 0 + unk_opnd = Operand(oUNK, None) + error_opnd = Operand(oERR, None) + spush = stack.append + + def do_binop(opcd, stk): + assert len(stk) >= 2 + bop = stk.pop() + aop = stk.pop() + argdict, result_kind, func, rank, sym = binop_rules[opcd] + otext = ''.join([ + '('[:aop.rank < rank], + aop.text, + ')'[:aop.rank < rank], + sym, + '('[:bop.rank < rank], + bop.text, + ')'[:bop.rank < rank], + ]) + resop = Operand(result_kind, None, rank, otext) + try: + bconv = argdict[bop.kind] + aconv = argdict[aop.kind] + except KeyError: + stk.append(resop) + return + if bop.value is None or aop.value is None: + stk.append(resop) + return + bval = bconv(bop.value) + aval = aconv(aop.value) + result = func(aval, bval) + if result_kind == oBOOL: + result = 1 if result else 0 + resop.value = result + stk.append(resop) + + def do_unaryop(opcode, result_kind, stk): + assert len(stk) >= 1 + aop = stk.pop() + val = aop.value + func, rank, sym1, sym2 = unop_rules[opcode] + otext = ''.join([ + sym1, + '('[:aop.rank < rank], + aop.text, + ')'[:aop.rank < rank], + sym2, + ]) + if val is not None: + val = func(val) + stk.append(Operand(result_kind, val, rank, otext)) + + def not_in_name_formula(op_arg, oname_arg): + msg = "ERROR *** Token 0x%02x (%s) found in NAME formula" \ + % (op_arg, oname_arg) + raise FormulaError(msg) + + if fmlalen == 0: + stack = [unk_opnd] + + while 0 <= pos < fmlalen: + op = BYTES_ORD(data[pos]) + opcode = op & 0x1f + optype = (op & 0x60) >> 5 + if optype: + opx = opcode + 32 + else: + opx = opcode + oname = onames[opx] # + [" RVA"][optype] + sz = sztab[opx] + if blah: + print("Pos:%d Op:0x%02x Name:t%s Sz:%d opcode:%02xh optype:%02xh" \ + % (pos, op, oname, sz, opcode, optype), file=bk.logfile) + print("Stack =", stack, file=bk.logfile) + if sz == -2: + msg = 'ERROR *** Unexpected token 0x%02x ("%s"); biff_version=%d' \ + % (op, oname, bv) + raise FormulaError(msg) + if not optype: + if 0x00 <= opcode <= 0x02: # unk_opnd, tExp, tTbl + not_in_name_formula(op, oname) + elif 0x03 <= opcode <= 0x0E: + # Add, Sub, Mul, Div, Power + # tConcat + # tLT, ..., tNE + do_binop(opcode, stack) + elif opcode == 0x0F: # tIsect + if blah: print("tIsect pre", stack, file=bk.logfile) + assert len(stack) >= 2 + bop = stack.pop() + aop = stack.pop() + sym = ' ' + rank = 80 ########## check ####### + otext = ''.join([ + '('[:aop.rank < rank], + aop.text, + ')'[:aop.rank < rank], + sym, + '('[:bop.rank < rank], + bop.text, + ')'[:bop.rank < rank], + ]) + res = Operand(oREF) + res.text = otext + if bop.kind == oERR or aop.kind == oERR: + res.kind = oERR + elif bop.kind == oUNK or aop.kind == oUNK: + # This can happen with undefined + # (go search in the current sheet) labels. + # For example =Bob Sales + # Each label gets a NAME record with an empty formula (!) + # Evaluation of the tName token classifies it as oUNK + # res.kind = oREF + pass + elif bop.kind == oREF == aop.kind: + if aop.value is not None and bop.value is not None: + assert len(aop.value) == 1 + assert len(bop.value) == 1 + coords = do_box_funcs( + tIsectFuncs, aop.value[0], bop.value[0]) + res.value = [Ref3D(coords)] + elif bop.kind == oREL == aop.kind: + res.kind = oREL + if aop.value is not None and bop.value is not None: + assert len(aop.value) == 1 + assert len(bop.value) == 1 + coords = do_box_funcs( + tIsectFuncs, aop.value[0], bop.value[0]) + relfa = aop.value[0].relflags + relfb = bop.value[0].relflags + if relfa == relfb: + res.value = [Ref3D(coords + relfa)] + else: + pass + spush(res) + if blah: print("tIsect post", stack, file=bk.logfile) + elif opcode == 0x10: # tList + if blah: print("tList pre", stack, file=bk.logfile) + assert len(stack) >= 2 + bop = stack.pop() + aop = stack.pop() + sym = ',' + rank = 80 ########## check ####### + otext = ''.join([ + '('[:aop.rank < rank], + aop.text, + ')'[:aop.rank < rank], + sym, + '('[:bop.rank < rank], + bop.text, + ')'[:bop.rank < rank], + ]) + res = Operand(oREF, None, rank, otext) + if bop.kind == oERR or aop.kind == oERR: + res.kind = oERR + elif bop.kind in (oREF, oREL) and aop.kind in (oREF, oREL): + res.kind = oREF + if aop.kind == oREL or bop.kind == oREL: + res.kind = oREL + if aop.value is not None and bop.value is not None: + assert len(aop.value) >= 1 + assert len(bop.value) == 1 + res.value = aop.value + bop.value + else: + pass + spush(res) + if blah: print("tList post", stack, file=bk.logfile) + elif opcode == 0x11: # tRange + if blah: print("tRange pre", stack, file=bk.logfile) + assert len(stack) >= 2 + bop = stack.pop() + aop = stack.pop() + sym = ':' + rank = 80 ########## check ####### + otext = ''.join([ + '('[:aop.rank < rank], + aop.text, + ')'[:aop.rank < rank], + sym, + '('[:bop.rank < rank], + bop.text, + ')'[:bop.rank < rank], + ]) + res = Operand(oREF, None, rank, otext) + if bop.kind == oERR or aop.kind == oERR: + res = oERR + elif bop.kind == oREF == aop.kind: + if aop.value is not None and bop.value is not None: + assert len(aop.value) == 1 + assert len(bop.value) == 1 + coords = do_box_funcs( + tRangeFuncs, aop.value[0], bop.value[0]) + res.value = [Ref3D(coords)] + elif bop.kind == oREL == aop.kind: + res.kind = oREL + if aop.value is not None and bop.value is not None: + assert len(aop.value) == 1 + assert len(bop.value) == 1 + coords = do_box_funcs( + tRangeFuncs, aop.value[0], bop.value[0]) + relfa = aop.value[0].relflags + relfb = bop.value[0].relflags + if relfa == relfb: + res.value = [Ref3D(coords + relfa)] + else: + pass + spush(res) + if blah: print("tRange post", stack, file=bk.logfile) + elif 0x12 <= opcode <= 0x14: # tUplus, tUminus, tPercent + do_unaryop(opcode, oNUM, stack) + elif opcode == 0x15: # tParen + # source cosmetics + pass + elif opcode == 0x16: # tMissArg + spush(Operand(oMSNG, None, LEAF_RANK, '')) + elif opcode == 0x17: # tStr + if bv <= 70: + strg, newpos = unpack_string_update_pos( + data, pos+1, bk.encoding, lenlen=1) + else: + strg, newpos = unpack_unicode_update_pos( + data, pos+1, lenlen=1) + sz = newpos - pos + if blah: print(" sz=%d strg=%r" % (sz, strg), file=bk.logfile) + text = '"' + strg.replace('"', '""') + '"' + spush(Operand(oSTRG, strg, LEAF_RANK, text)) + elif opcode == 0x18: # tExtended + # new with BIFF 8 + assert bv >= 80 + # not in OOo docs + raise FormulaError("tExtended token not implemented") + elif opcode == 0x19: # tAttr + subop, nc = unpack("= 1 + aop = stack[-1] + otext = 'SUM(%s)' % aop.text + stack[-1] = Operand(oNUM, None, FUNC_RANK, otext) + else: + sz = 4 + if blah: + print(" subop=%02xh subname=t%s sz=%d nc=%02xh" \ + % (subop, subname, sz, nc), file=bk.logfile) + elif 0x1A <= opcode <= 0x1B: # tSheet, tEndSheet + assert bv < 50 + raise FormulaError("tSheet & tEndsheet tokens not implemented") + elif 0x1C <= opcode <= 0x1F: # tErr, tBool, tInt, tNum + inx = opcode - 0x1C + nb = [1, 1, 2, 8][inx] + kind = [oERR, oBOOL, oNUM, oNUM][inx] + value, = unpack("<" + "BBHd"[inx], data[pos+1:pos+1+nb]) + if inx == 2: # tInt + value = float(value) + text = str(value) + elif inx == 3: # tNum + text = str(value) + elif inx == 1: # tBool + text = ('FALSE', 'TRUE')[value] + else: + text = '"' +error_text_from_code[value] + '"' + spush(Operand(kind, value, LEAF_RANK, text)) + else: + raise FormulaError("Unhandled opcode: 0x%02x" % opcode) + if sz <= 0: + raise FormulaError("Size not set for opcode 0x%02x" % opcode) + pos += sz + continue + if opcode == 0x00: # tArray + spush(unk_opnd) + elif opcode == 0x01: # tFunc + nb = 1 + int(bv >= 40) + funcx = unpack("<" + " BH"[nb], data[pos+1:pos+1+nb])[0] + func_attrs = func_defs.get(funcx, None) + if not func_attrs: + print("*** formula/tFunc unknown FuncID:%d" \ + % funcx, file=bk.logfile) + spush(unk_opnd) + else: + func_name, nargs = func_attrs[:2] + if blah: + print(" FuncID=%d name=%s nargs=%d" \ + % (funcx, func_name, nargs), file=bk.logfile) + assert len(stack) >= nargs + if nargs: + argtext = listsep.join([arg.text for arg in stack[-nargs:]]) + otext = "%s(%s)" % (func_name, argtext) + del stack[-nargs:] + else: + otext = func_name + "()" + res = Operand(oUNK, None, FUNC_RANK, otext) + spush(res) + elif opcode == 0x02: #tFuncVar + nb = 1 + int(bv >= 40) + nargs, funcx = unpack("= nargs + assert len(stack) >= nargs + argtext = listsep.join([arg.text for arg in stack[-nargs:]]) + otext = "%s(%s)" % (func_name, argtext) + res = Operand(oUNK, None, FUNC_RANK, otext) + if funcx == 1: # IF + testarg = stack[-nargs] + if testarg.kind not in (oNUM, oBOOL): + if blah and testarg.kind != oUNK: + print("IF testarg kind?", file=bk.logfile) + elif testarg.value not in (0, 1): + if blah and testarg.value is not None: + print("IF testarg value?", file=bk.logfile) + else: + if nargs == 2 and not testarg.value: + # IF(FALSE, tv) => FALSE + res.kind, res.value = oBOOL, 0 + else: + respos = -nargs + 2 - int(testarg.value) + chosen = stack[respos] + if chosen.kind == oMSNG: + res.kind, res.value = oNUM, 0 + else: + res.kind, res.value = chosen.kind, chosen.value + if blah: + print("$$$$$$ IF => constant", file=bk.logfile) + elif funcx == 100: # CHOOSE + testarg = stack[-nargs] + if testarg.kind == oNUM: + if 1 <= testarg.value < nargs: + chosen = stack[-nargs + int(testarg.value)] + if chosen.kind == oMSNG: + res.kind, res.value = oNUM, 0 + else: + res.kind, res.value = chosen.kind, chosen.value + del stack[-nargs:] + spush(res) + elif opcode == 0x03: #tName + tgtnamex = unpack("> bk.logfile, " ", res + # spush(res) + elif opcode == 0x0D: #tAreaN + not_in_name_formula(op, oname) + # res = get_cell_range_addr(data, pos+1, bv, reldelta=1) + # # note *ALL* tAreaN usage has signed offset for relative addresses + # any_rel = 1 + # if blah: print >> bk.logfile, " ", res + elif opcode == 0x1A: # tRef3d + if bv >= 80: + res = get_cell_addr(data, pos+3, bv, reldelta) + refx = unpack("= 80: + res1, res2 = get_cell_range_addr(data, pos+3, bv, reldelta) + refx = unpack("= 80: + refx, tgtnamex = unpack(" 0: + refx -= 1 + elif refx < 0: + refx = -refx - 1 + else: + dodgy = 1 + if blah: + print(" origrefx=%d refx=%d tgtnamex=%d dodgy=%d" \ + % (origrefx, refx, tgtnamex, dodgy), file=bk.logfile) + if tgtnamex == namex: + if blah: print("!!!! Self-referential !!!!", file=bk.logfile) + dodgy = any_err = 1 + if not dodgy: + if bv >= 80: + shx1, shx2 = get_externsheet_local_range(bk, refx, blah) + elif origrefx > 0: + shx1, shx2 = (-4, -4) # external ref + else: + exty = bk._externsheet_type_b57[refx] + if exty == 4: # non-specific sheet in own doc't + shx1, shx2 = (-1, -1) # internal, any sheet + else: + shx1, shx2 = (-666, -666) + if dodgy or shx1 < -1: + otext = "<>" \ + % (tgtnamex, origrefx) + res = Operand(oUNK, None, LEAF_RANK, otext) + else: + tgtobj = bk.name_obj_list[tgtnamex] + if not tgtobj.evaluated: + ### recursive ### + evaluate_name_formula(bk, tgtobj, tgtnamex, blah, level+1) + if tgtobj.macro or tgtobj.binary \ + or tgtobj.any_err: + if blah: + tgtobj.dump( + bk.logfile, + header="!!! bad tgtobj !!!", + footer="------------------", + ) + res = Operand(oUNK, None) + any_err = any_err or tgtobj.macro or tgtobj.binary or tgtobj.any_err + any_rel = any_rel or tgtobj.any_rel + else: + assert len(tgtobj.stack) == 1 + res = copy.deepcopy(tgtobj.stack[0]) + res.rank = LEAF_RANK + if tgtobj.scope == -1: + res.text = tgtobj.name + else: + res.text = "%s!%s" \ + % (bk._sheet_names[tgtobj.scope], tgtobj.name) + if blah: + print(" tNameX: setting text to", repr(res.text), file=bk.logfile) + spush(res) + elif opcode in error_opcodes: + any_err = 1 + spush(error_opnd) + else: + if blah: + print("FORMULA: /// Not handled yet: t" + oname, file=bk.logfile) + any_err = 1 + if sz <= 0: + raise FormulaError("Fatal: token size is not positive") + pos += sz + any_rel = not not any_rel + if blah: + fprintf(bk.logfile, "End of formula. level=%d any_rel=%d any_err=%d stack=%r\n", + level, not not any_rel, any_err, stack) + if len(stack) >= 2: + print("*** Stack has unprocessed args", file=bk.logfile) + print(file=bk.logfile) + nobj.stack = stack + if len(stack) != 1: + nobj.result = None + else: + nobj.result = stack[0] + nobj.any_rel = any_rel + nobj.any_err = any_err + nobj.any_external = any_external + nobj.evaluated = 1 + +#### under construction ############################################################################# +def decompile_formula(bk, fmla, fmlalen, + fmlatype=None, browx=None, bcolx=None, + blah=0, level=0, r1c1=0): + if level > STACK_ALARM_LEVEL: + blah = 1 + reldelta = fmlatype in (FMLA_TYPE_SHARED, FMLA_TYPE_NAME, FMLA_TYPE_COND_FMT, FMLA_TYPE_DATA_VAL) + data = fmla + bv = bk.biff_version + if blah: + print("::: decompile_formula len=%d fmlatype=%r browx=%r bcolx=%r reldelta=%d %r level=%d" \ + % (fmlalen, fmlatype, browx, bcolx, reldelta, data, level), file=bk.logfile) + hex_char_dump(data, 0, fmlalen, fout=bk.logfile) + if level > STACK_PANIC_LEVEL: + raise XLRDError("Excessive indirect references in formula") + sztab = szdict[bv] + pos = 0 + stack = [] + any_rel = 0 + any_err = 0 + any_external = 0 + unk_opnd = Operand(oUNK, None) + error_opnd = Operand(oERR, None) + spush = stack.append + + def do_binop(opcd, stk): + assert len(stk) >= 2 + bop = stk.pop() + aop = stk.pop() + argdict, result_kind, func, rank, sym = binop_rules[opcd] + otext = ''.join([ + '('[:aop.rank < rank], + aop.text, + ')'[:aop.rank < rank], + sym, + '('[:bop.rank < rank], + bop.text, + ')'[:bop.rank < rank], + ]) + resop = Operand(result_kind, None, rank, otext) + stk.append(resop) + + def do_unaryop(opcode, result_kind, stk): + assert len(stk) >= 1 + aop = stk.pop() + func, rank, sym1, sym2 = unop_rules[opcode] + otext = ''.join([ + sym1, + '('[:aop.rank < rank], + aop.text, + ')'[:aop.rank < rank], + sym2, + ]) + stk.append(Operand(result_kind, None, rank, otext)) + + def unexpected_opcode(op_arg, oname_arg): + msg = "ERROR *** Unexpected token 0x%02x (%s) found in formula type %s" \ + % (op_arg, oname_arg, FMLA_TYPEDESCR_MAP[fmlatype]) + print(msg, file=bk.logfile) + # raise FormulaError(msg) + + if fmlalen == 0: + stack = [unk_opnd] + + while 0 <= pos < fmlalen: + op = BYTES_ORD(data[pos]) + opcode = op & 0x1f + optype = (op & 0x60) >> 5 + if optype: + opx = opcode + 32 + else: + opx = opcode + oname = onames[opx] # + [" RVA"][optype] + sz = sztab[opx] + if blah: + print("Pos:%d Op:0x%02x opname:t%s Sz:%d opcode:%02xh optype:%02xh" \ + % (pos, op, oname, sz, opcode, optype), file=bk.logfile) + print("Stack =", stack, file=bk.logfile) + if sz == -2: + msg = 'ERROR *** Unexpected token 0x%02x ("%s"); biff_version=%d' \ + % (op, oname, bv) + raise FormulaError(msg) + if _TOKEN_NOT_ALLOWED(opx, 0) & fmlatype: + unexpected_opcode(op, oname) + if not optype: + if opcode <= 0x01: # tExp + if bv >= 30: + fmt = '= 2 + bop = stack.pop() + aop = stack.pop() + sym = ' ' + rank = 80 ########## check ####### + otext = ''.join([ + '('[:aop.rank < rank], + aop.text, + ')'[:aop.rank < rank], + sym, + '('[:bop.rank < rank], + bop.text, + ')'[:bop.rank < rank], + ]) + res = Operand(oREF) + res.text = otext + if bop.kind == oERR or aop.kind == oERR: + res.kind = oERR + elif bop.kind == oUNK or aop.kind == oUNK: + # This can happen with undefined + # (go search in the current sheet) labels. + # For example =Bob Sales + # Each label gets a NAME record with an empty formula (!) + # Evaluation of the tName token classifies it as oUNK + # res.kind = oREF + pass + elif bop.kind == oREF == aop.kind: + pass + elif bop.kind == oREL == aop.kind: + res.kind = oREL + else: + pass + spush(res) + if blah: print("tIsect post", stack, file=bk.logfile) + elif opcode == 0x10: # tList + if blah: print("tList pre", stack, file=bk.logfile) + assert len(stack) >= 2 + bop = stack.pop() + aop = stack.pop() + sym = ',' + rank = 80 ########## check ####### + otext = ''.join([ + '('[:aop.rank < rank], + aop.text, + ')'[:aop.rank < rank], + sym, + '('[:bop.rank < rank], + bop.text, + ')'[:bop.rank < rank], + ]) + res = Operand(oREF, None, rank, otext) + if bop.kind == oERR or aop.kind == oERR: + res.kind = oERR + elif bop.kind in (oREF, oREL) and aop.kind in (oREF, oREL): + res.kind = oREF + if aop.kind == oREL or bop.kind == oREL: + res.kind = oREL + else: + pass + spush(res) + if blah: print("tList post", stack, file=bk.logfile) + elif opcode == 0x11: # tRange + if blah: print("tRange pre", stack, file=bk.logfile) + assert len(stack) >= 2 + bop = stack.pop() + aop = stack.pop() + sym = ':' + rank = 80 ########## check ####### + otext = ''.join([ + '('[:aop.rank < rank], + aop.text, + ')'[:aop.rank < rank], + sym, + '('[:bop.rank < rank], + bop.text, + ')'[:bop.rank < rank], + ]) + res = Operand(oREF, None, rank, otext) + if bop.kind == oERR or aop.kind == oERR: + res = oERR + elif bop.kind == oREF == aop.kind: + pass + else: + pass + spush(res) + if blah: print("tRange post", stack, file=bk.logfile) + elif 0x12 <= opcode <= 0x14: # tUplus, tUminus, tPercent + do_unaryop(opcode, oNUM, stack) + elif opcode == 0x15: # tParen + # source cosmetics + pass + elif opcode == 0x16: # tMissArg + spush(Operand(oMSNG, None, LEAF_RANK, '')) + elif opcode == 0x17: # tStr + if bv <= 70: + strg, newpos = unpack_string_update_pos( + data, pos+1, bk.encoding, lenlen=1) + else: + strg, newpos = unpack_unicode_update_pos( + data, pos+1, lenlen=1) + sz = newpos - pos + if blah: print(" sz=%d strg=%r" % (sz, strg), file=bk.logfile) + text = '"' + strg.replace('"', '""') + '"' + spush(Operand(oSTRG, None, LEAF_RANK, text)) + elif opcode == 0x18: # tExtended + # new with BIFF 8 + assert bv >= 80 + # not in OOo docs, don't even know how to determine its length + raise FormulaError("tExtended token not implemented") + elif opcode == 0x19: # tAttr + subop, nc = unpack("= 1 + aop = stack[-1] + otext = 'SUM(%s)' % aop.text + stack[-1] = Operand(oNUM, None, FUNC_RANK, otext) + else: + sz = 4 + if blah: + print(" subop=%02xh subname=t%s sz=%d nc=%02xh" \ + % (subop, subname, sz, nc), file=bk.logfile) + elif 0x1A <= opcode <= 0x1B: # tSheet, tEndSheet + assert bv < 50 + raise FormulaError("tSheet & tEndsheet tokens not implemented") + elif 0x1C <= opcode <= 0x1F: # tErr, tBool, tInt, tNum + inx = opcode - 0x1C + nb = [1, 1, 2, 8][inx] + kind = [oERR, oBOOL, oNUM, oNUM][inx] + value, = unpack("<" + "BBHd"[inx], data[pos+1:pos+1+nb]) + if inx == 2: # tInt + value = float(value) + text = str(value) + elif inx == 3: # tNum + text = str(value) + elif inx == 1: # tBool + text = ('FALSE', 'TRUE')[value] + else: + text = '"' +error_text_from_code[value] + '"' + spush(Operand(kind, None, LEAF_RANK, text)) + else: + raise FormulaError("Unhandled opcode: 0x%02x" % opcode) + if sz <= 0: + raise FormulaError("Size not set for opcode 0x%02x" % opcode) + pos += sz + continue + if opcode == 0x00: # tArray + spush(unk_opnd) + elif opcode == 0x01: # tFunc + nb = 1 + int(bv >= 40) + funcx = unpack("<" + " BH"[nb], data[pos+1:pos+1+nb])[0] + func_attrs = func_defs.get(funcx, None) + if not func_attrs: + print("*** formula/tFunc unknown FuncID:%d" % funcx, file=bk.logfile) + spush(unk_opnd) + else: + func_name, nargs = func_attrs[:2] + if blah: + print(" FuncID=%d name=%s nargs=%d" \ + % (funcx, func_name, nargs), file=bk.logfile) + assert len(stack) >= nargs + if nargs: + argtext = listsep.join([arg.text for arg in stack[-nargs:]]) + otext = "%s(%s)" % (func_name, argtext) + del stack[-nargs:] + else: + otext = func_name + "()" + res = Operand(oUNK, None, FUNC_RANK, otext) + spush(res) + elif opcode == 0x02: #tFuncVar + nb = 1 + int(bv >= 40) + nargs, funcx = unpack("= nargs + assert len(stack) >= nargs + argtext = listsep.join([arg.text for arg in stack[-nargs:]]) + otext = "%s(%s)" % (func_name, argtext) + res = Operand(oUNK, None, FUNC_RANK, otext) + del stack[-nargs:] + spush(res) + elif opcode == 0x03: #tName + tgtnamex = unpack("> bk.logfile, " ", res + res1, res2 = get_cell_range_addr( + data, pos+1, bv, reldelta, browx, bcolx) + if blah: print(" ", res1, res2, file=bk.logfile) + rowx1, colx1, row_rel1, col_rel1 = res1 + rowx2, colx2, row_rel2, col_rel2 = res2 + coords = (rowx1, rowx2+1, colx1, colx2+1) + relflags = (row_rel1, row_rel2, col_rel1, col_rel2) + if sum(relflags): # relative + okind = oREL + else: + okind = oREF + if blah: print(" ", coords, relflags, file=bk.logfile) + otext = rangename2drel(coords, relflags, browx, bcolx, r1c1) + res = Operand(okind, None, LEAF_RANK, otext) + spush(res) + elif opcode == 0x1A: # tRef3d + if bv >= 80: + res = get_cell_addr(data, pos+3, bv, reldelta, browx, bcolx) + refx = unpack("= 80: + res1, res2 = get_cell_range_addr(data, pos+3, bv, reldelta) + refx = unpack("= 80: + refx, tgtnamex = unpack(" 0: + refx -= 1 + elif refx < 0: + refx = -refx - 1 + else: + dodgy = 1 + if blah: + print(" origrefx=%d refx=%d tgtnamex=%d dodgy=%d" \ + % (origrefx, refx, tgtnamex, dodgy), file=bk.logfile) + # if tgtnamex == namex: + # if blah: print >> bk.logfile, "!!!! Self-referential !!!!" + # dodgy = any_err = 1 + if not dodgy: + if bv >= 80: + shx1, shx2 = get_externsheet_local_range(bk, refx, blah) + elif origrefx > 0: + shx1, shx2 = (-4, -4) # external ref + else: + exty = bk._externsheet_type_b57[refx] + if exty == 4: # non-specific sheet in own doc't + shx1, shx2 = (-1, -1) # internal, any sheet + else: + shx1, shx2 = (-666, -666) + okind = oUNK + ovalue = None + if shx1 == -5: # addin func name + okind = oSTRG + ovalue = bk.addin_func_names[tgtnamex] + otext = '"' + ovalue.replace('"', '""') + '"' + elif dodgy or shx1 < -1: + otext = "<>" \ + % (tgtnamex, origrefx) + else: + tgtobj = bk.name_obj_list[tgtnamex] + if tgtobj.scope == -1: + otext = tgtobj.name + else: + otext = "%s!%s" \ + % (bk._sheet_names[tgtobj.scope], tgtobj.name) + if blah: + print(" tNameX: setting text to", repr(res.text), file=bk.logfile) + res = Operand(okind, ovalue, LEAF_RANK, otext) + spush(res) + elif opcode in error_opcodes: + any_err = 1 + spush(error_opnd) + else: + if blah: + print("FORMULA: /// Not handled yet: t" + oname, file=bk.logfile) + any_err = 1 + if sz <= 0: + raise FormulaError("Fatal: token size is not positive") + pos += sz + any_rel = not not any_rel + if blah: + print("End of formula. level=%d any_rel=%d any_err=%d stack=%r" % \ + (level, not not any_rel, any_err, stack), file=bk.logfile) + if len(stack) >= 2: + print("*** Stack has unprocessed args", file=bk.logfile) + print(file=bk.logfile) + + if len(stack) != 1: + result = None + else: + result = stack[0].text + return result + +#### under deconstruction ### +def dump_formula(bk, data, fmlalen, bv, reldelta, blah=0, isname=0): + if blah: + print("dump_formula", fmlalen, bv, len(data), file=bk.logfile) + hex_char_dump(data, 0, fmlalen, fout=bk.logfile) + assert bv >= 80 #### this function needs updating #### + sztab = szdict[bv] + pos = 0 + stack = [] + any_rel = 0 + any_err = 0 + spush = stack.append + while 0 <= pos < fmlalen: + op = BYTES_ORD(data[pos]) + opcode = op & 0x1f + optype = (op & 0x60) >> 5 + if optype: + opx = opcode + 32 + else: + opx = opcode + oname = onames[opx] # + [" RVA"][optype] + + sz = sztab[opx] + if blah: + print("Pos:%d Op:0x%02x Name:t%s Sz:%d opcode:%02xh optype:%02xh" \ + % (pos, op, oname, sz, opcode, optype), file=bk.logfile) + if not optype: + if 0x01 <= opcode <= 0x02: # tExp, tTbl + # reference to a shared formula or table record + rowx, colx = unpack("= 2 + bop = stack.pop() + aop = stack.pop() + spush(aop + bop) + if blah: print("tlist post", stack, file=bk.logfile) + elif opcode == 0x11: # tRange + if blah: print("tRange pre", stack, file=bk.logfile) + assert len(stack) >= 2 + bop = stack.pop() + aop = stack.pop() + assert len(aop) == 1 + assert len(bop) == 1 + result = do_box_funcs(tRangeFuncs, aop[0], bop[0]) + spush(result) + if blah: print("tRange post", stack, file=bk.logfile) + elif opcode == 0x0F: # tIsect + if blah: print("tIsect pre", stack, file=bk.logfile) + assert len(stack) >= 2 + bop = stack.pop() + aop = stack.pop() + assert len(aop) == 1 + assert len(bop) == 1 + result = do_box_funcs(tIsectFuncs, aop[0], bop[0]) + spush(result) + if blah: print("tIsect post", stack, file=bk.logfile) + elif opcode == 0x19: # tAttr + subop, nc = unpack("= 40) + funcx = unpack("<" + " BH"[nb], data[pos+1:pos+1+nb]) + if blah: print(" FuncID=%d" % funcx, file=bk.logfile) + elif opcode == 0x02: #tFuncVar + nb = 1 + int(bv >= 40) + nargs, funcx = unpack("= 2: + print("*** Stack has unprocessed args", file=bk.logfile) + +# === Some helper functions for displaying cell references === + +# I'm aware of only one possibility of a sheet-relative component in +# a reference: a 2D reference located in the "current sheet". +# xlrd stores this internally with bounds of (0, 1, ...) and +# relative flags of (1, 1, ...). These functions display the +# sheet component as empty, just like Excel etc. + +def rownamerel(rowx, rowxrel, browx=None, r1c1=0): + # if no base rowx is provided, we have to return r1c1 + if browx is None: + r1c1 = True + if not rowxrel: + if r1c1: + return "R%d" % (rowx+1) + return "$%d" % (rowx+1) + if r1c1: + if rowx: + return "R[%d]" % rowx + return "R" + return "%d" % ((browx + rowx) % 65536 + 1) + +def colnamerel(colx, colxrel, bcolx=None, r1c1=0): + # if no base colx is provided, we have to return r1c1 + if bcolx is None: + r1c1 = True + if not colxrel: + if r1c1: + return "C%d" % (colx + 1) + return "$" + colname(colx) + if r1c1: + if colx: + return "C[%d]" % colx + return "C" + return colname((bcolx + colx) % 256) + +## +# Utility function: (5, 7) => 'H6' +def cellname(rowx, colx): + """ (5, 7) => 'H6' """ + return "%s%d" % (colname(colx), rowx+1) + +## +# Utility function: (5, 7) => '$H$6' +def cellnameabs(rowx, colx, r1c1=0): + """ (5, 7) => '$H$6' or 'R8C6'""" + if r1c1: + return "R%dC%d" % (rowx+1, colx+1) + return "$%s$%d" % (colname(colx), rowx+1) + +def cellnamerel(rowx, colx, rowxrel, colxrel, browx=None, bcolx=None, r1c1=0): + if not rowxrel and not colxrel: + return cellnameabs(rowx, colx, r1c1) + if (rowxrel and browx is None) or (colxrel and bcolx is None): + # must flip the whole cell into R1C1 mode + r1c1 = True + c = colnamerel(colx, colxrel, bcolx, r1c1) + r = rownamerel(rowx, rowxrel, browx, r1c1) + if r1c1: + return r + c + return c + r + +## +# Utility function: 7 => 'H', 27 => 'AB' +def colname(colx): + """ 7 => 'H', 27 => 'AB' """ + alphabet = "ABCDEFGHIJKLMNOPQRSTUVWXYZ" + if colx <= 25: + return alphabet[colx] + else: + xdiv26, xmod26 = divmod(colx, 26) + return alphabet[xdiv26 - 1] + alphabet[xmod26] + +def rangename2d(rlo, rhi, clo, chi, r1c1=0): + """ (5, 20, 7, 10) => '$H$6:$J$20' """ + if r1c1: + return + if rhi == rlo+1 and chi == clo+1: + return cellnameabs(rlo, clo, r1c1) + return "%s:%s" % (cellnameabs(rlo, clo, r1c1), cellnameabs(rhi-1, chi-1, r1c1)) + +def rangename2drel(rlo_rhi_clo_chi, rlorel_rhirel_clorel_chirel, browx=None, bcolx=None, r1c1=0): + rlo, rhi, clo, chi = rlo_rhi_clo_chi + rlorel, rhirel, clorel, chirel = rlorel_rhirel_clorel_chirel + if (rlorel or rhirel) and browx is None: + r1c1 = True + if (clorel or chirel) and bcolx is None: + r1c1 = True + return "%s:%s" % ( + cellnamerel(rlo, clo, rlorel, clorel, browx, bcolx, r1c1), + cellnamerel(rhi-1, chi-1, rhirel, chirel, browx, bcolx, r1c1) + ) +## +# Utility function: +#
Ref3D((1, 4, 5, 20, 7, 10)) => 'Sheet2:Sheet3!$H$6:$J$20' +def rangename3d(book, ref3d): + """ Ref3D(1, 4, 5, 20, 7, 10) => 'Sheet2:Sheet3!$H$6:$J$20' + (assuming Excel's default sheetnames) """ + coords = ref3d.coords + return "%s!%s" % ( + sheetrange(book, *coords[:2]), + rangename2d(*coords[2:6])) + +## +# Utility function: +#
Ref3D(coords=(0, 1, -32, -22, -13, 13), relflags=(0, 0, 1, 1, 1, 1)) +# R1C1 mode => 'Sheet1!R[-32]C[-13]:R[-23]C[12]' +# A1 mode => depends on base cell (browx, bcolx) +def rangename3drel(book, ref3d, browx=None, bcolx=None, r1c1=0): + coords = ref3d.coords + relflags = ref3d.relflags + shdesc = sheetrangerel(book, coords[:2], relflags[:2]) + rngdesc = rangename2drel(coords[2:6], relflags[2:6], browx, bcolx, r1c1) + if not shdesc: + return rngdesc + return "%s!%s" % (shdesc, rngdesc) + +def quotedsheetname(shnames, shx): + if shx >= 0: + shname = shnames[shx] + else: + shname = { + -1: "?internal; any sheet?", + -2: "internal; deleted sheet", + -3: "internal; macro sheet", + -4: "<>", + }.get(shx, "?error %d?" % shx) + if "'" in shname: + return "'" + shname.replace("'", "''") + "'" + if " " in shname: + return "'" + shname + "'" + return shname + +def sheetrange(book, slo, shi): + shnames = book.sheet_names() + shdesc = quotedsheetname(shnames, slo) + if slo != shi-1: + shdesc += ":" + quotedsheetname(shnames, shi-1) + return shdesc + +def sheetrangerel(book, srange, srangerel): + slo, shi = srange + slorel, shirel = srangerel + if not slorel and not shirel: + return sheetrange(book, slo, shi) + assert (slo == 0 == shi-1) and slorel and shirel + return "" + +# ============================================================== diff --git a/Assignment 3/packages/xlrd/info.py b/Assignment 3/packages/xlrd/info.py new file mode 100644 index 0000000..a57a784 --- /dev/null +++ b/Assignment 3/packages/xlrd/info.py @@ -0,0 +1 @@ +__VERSION__ = "0.9.4" diff --git a/Assignment 3/packages/xlrd/licences.py b/Assignment 3/packages/xlrd/licences.py new file mode 100644 index 0000000..1e262a9 --- /dev/null +++ b/Assignment 3/packages/xlrd/licences.py @@ -0,0 +1,77 @@ +# -*- coding: cp1252 -*- + +""" +Portions copyright © 2005-2009, Stephen John Machin, Lingfo Pty Ltd +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + +1. Redistributions of source code must retain the above copyright notice, +this list of conditions and the following disclaimer. + +2. Redistributions in binary form must reproduce the above copyright notice, +this list of conditions and the following disclaimer in the documentation +and/or other materials provided with the distribution. + +3. None of the names of Stephen John Machin, Lingfo Pty Ltd and any +contributors may be used to endorse or promote products derived from this +software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" +AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, +THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR +PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS +BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR +CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF +SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN +CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) +ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF +THE POSSIBILITY OF SUCH DAMAGE. +""" + +""" +/*- + * Copyright (c) 2001 David Giffin. + * All rights reserved. + * + * Based on the the Java version: Andrew Khan Copyright (c) 2000. + * + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * 3. All advertising materials mentioning features or use of this + * software must display the following acknowledgment: + * "This product includes software developed by + * David Giffin ." + * + * 4. Redistributions of any form whatsoever must retain the following + * acknowledgment: + * "This product includes software developed by + * David Giffin ." + * + * THIS SOFTWARE IS PROVIDED BY DAVID GIFFIN ``AS IS'' AND ANY + * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL DAVID GIFFIN OR + * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT + * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; + * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, + * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) + * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED + * OF THE POSSIBILITY OF SUCH DAMAGE. + */ +""" diff --git a/Assignment 3/packages/xlrd/sheet.py b/Assignment 3/packages/xlrd/sheet.py new file mode 100644 index 0000000..721d17c --- /dev/null +++ b/Assignment 3/packages/xlrd/sheet.py @@ -0,0 +1,2424 @@ +# -*- coding: cp1252 -*- + +## +#

Portions copyright © 2005-2013 Stephen John Machin, Lingfo Pty Ltd

+#

This module is part of the xlrd package, which is released under a BSD-style licence.

+## + +# 2010-04-25 SJM fix zoom factors cooking logic +# 2010-04-15 CW r4253 fix zoom factors cooking logic +# 2010-04-09 CW r4248 add a flag so xlutils knows whether or not to write a PANE record +# 2010-03-29 SJM Fixed bug in adding new empty rows in put_cell_ragged +# 2010-03-28 SJM Tailored put_cell method for each of ragged_rows=False (fixed speed regression) and =True (faster) +# 2010-03-25 CW r4236 Slight refactoring to remove method calls +# 2010-03-25 CW r4235 Collapse expand_cells into put_cell and enhance the raggedness. This should save even more memory! +# 2010-03-25 CW r4234 remove duplicate chunks for extend_cells; refactor to remove put_number_cell and put_blank_cell which essentially duplicated the code of put_cell +# 2010-03-10 SJM r4222 Added reading of the PANE record. +# 2010-03-10 SJM r4221 Preliminary work on "cooked" mag factors; use at own peril +# 2010-03-01 SJM Reading SCL record +# 2010-03-01 SJM Added ragged_rows functionality +# 2009-08-23 SJM Reduced CPU time taken by parsing MULBLANK records. +# 2009-08-18 SJM Used __slots__ and sharing to reduce memory consumed by Rowinfo instances +# 2009-05-31 SJM Fixed problem with no CODEPAGE record on extremely minimal BIFF2.x 3rd-party file +# 2009-04-27 SJM Integrated on_demand patch by Armando Serrano Lombillo +# 2008-02-09 SJM Excel 2.0: build XFs on the fly from cell attributes +# 2007-12-04 SJM Added support for Excel 2.x (BIFF2) files. +# 2007-10-11 SJM Added missing entry for blank cell type to ctype_text +# 2007-07-11 SJM Allow for BIFF2/3-style FORMAT record in BIFF4/8 file +# 2007-04-22 SJM Remove experimental "trimming" facility. + +from __future__ import print_function + +from array import array +from struct import unpack, calcsize +from .biffh import * +from .timemachine import * +from .formula import dump_formula, decompile_formula, rangename2d, FMLA_TYPE_CELL, FMLA_TYPE_SHARED +from .formatting import nearest_colour_index, Format + +DEBUG = 0 +OBJ_MSO_DEBUG = 0 + +_WINDOW2_options = ( + # Attribute names and initial values to use in case + # a WINDOW2 record is not written. + ("show_formulas", 0), + ("show_grid_lines", 1), + ("show_sheet_headers", 1), + ("panes_are_frozen", 0), + ("show_zero_values", 1), + ("automatic_grid_line_colour", 1), + ("columns_from_right_to_left", 0), + ("show_outline_symbols", 1), + ("remove_splits_if_pane_freeze_is_removed", 0), + # Multiple sheets can be selected, but only one can be active + # (hold down Ctrl and click multiple tabs in the file in OOo) + ("sheet_selected", 0), + # "sheet_visible" should really be called "sheet_active" + # and is 1 when this sheet is the sheet displayed when the file + # is open. More than likely only one sheet should ever be set as + # visible. + # This would correspond to the Book's sheet_active attribute, but + # that doesn't exist as WINDOW1 records aren't currently processed. + # The real thing is the visibility attribute from the BOUNDSHEET record. + ("sheet_visible", 0), + ("show_in_page_break_preview", 0), + ) + +## +#

Contains the data for one worksheet.

+# +#

In the cell access functions, "rowx" is a row index, counting from zero, and "colx" is a +# column index, counting from zero. +# Negative values for row/column indexes and slice positions are supported in the expected fashion.

+# +#

For information about cell types and cell values, refer to the documentation of the {@link #Cell} class.

+# +#

WARNING: You don't call this class yourself. You access Sheet objects via the Book object that +# was returned when you called xlrd.open_workbook("myfile.xls").

+ + +class Sheet(BaseObject): + ## + # Name of sheet. + name = '' + + ## + # A reference to the Book object to which this sheet belongs. + # Example usage: some_sheet.book.datemode + book = None + + ## + # Number of rows in sheet. A row index is in range(thesheet.nrows). + nrows = 0 + + ## + # Nominal number of columns in sheet. It is 1 + the maximum column index + # found, ignoring trailing empty cells. See also open_workbook(ragged_rows=?) + # and Sheet.{@link #Sheet.row_len}(row_index). + ncols = 0 + + ## + # The map from a column index to a {@link #Colinfo} object. Often there is an entry + # in COLINFO records for all column indexes in range(257). + # Note that xlrd ignores the entry for the non-existent + # 257th column. On the other hand, there may be no entry for unused columns. + #
-- New in version 0.6.1. Populated only if open_workbook(formatting_info=True). + colinfo_map = {} + + ## + # The map from a row index to a {@link #Rowinfo} object. Note that it is possible + # to have missing entries -- at least one source of XLS files doesn't + # bother writing ROW records. + #
-- New in version 0.6.1. Populated only if open_workbook(formatting_info=True). + rowinfo_map = {} + + ## + # List of address ranges of cells containing column labels. + # These are set up in Excel by Insert > Name > Labels > Columns. + #
-- New in version 0.6.0 + #
How to deconstruct the list: + # 
+    # for crange in thesheet.col_label_ranges:
+    #     rlo, rhi, clo, chi = crange
+    #     for rx in xrange(rlo, rhi):
+    #         for cx in xrange(clo, chi):
+    #             print "Column label at (rowx=%d, colx=%d) is %r" \
+    #                 (rx, cx, thesheet.cell_value(rx, cx))
+    #
+ col_label_ranges = [] + + ## + # List of address ranges of cells containing row labels. + # For more details, see col_label_ranges above. + #
-- New in version 0.6.0 + row_label_ranges = [] + + ## + # List of address ranges of cells which have been merged. + # These are set up in Excel by Format > Cells > Alignment, then ticking + # the "Merge cells" box. + #
-- New in version 0.6.1. Extracted only if open_workbook(formatting_info=True). + #
How to deconstruct the list: + # 
+    # for crange in thesheet.merged_cells:
+    #     rlo, rhi, clo, chi = crange
+    #     for rowx in xrange(rlo, rhi):
+    #         for colx in xrange(clo, chi):
+    #             # cell (rlo, clo) (the top left one) will carry the data
+    #             # and formatting info; the remainder will be recorded as
+    #             # blank cells, but a renderer will apply the formatting info
+    #             # for the top left cell (e.g. border, pattern) to all cells in
+    #             # the range.
+    #
+ merged_cells = [] + + ## + # Mapping of (rowx, colx) to list of (offset, font_index) tuples. The offset + # defines where in the string the font begins to be used. + # Offsets are expected to be in ascending order. + # If the first offset is not zero, the meaning is that the cell's XF's font should + # be used from offset 0. + #
This is a sparse mapping. There is no entry for cells that are not formatted with + # rich text. + #
How to use: + # 
+    # runlist = thesheet.rich_text_runlist_map.get((rowx, colx))
+    # if runlist:
+    #     for offset, font_index in runlist:
+    #         # do work here.
+    #         pass
+    #
+ # Populated only if open_workbook(formatting_info=True). + #
-- New in version 0.7.2. + #
  + rich_text_runlist_map = {} + + ## + # Default column width from DEFCOLWIDTH record, else None. + # From the OOo docs:
+ # """Column width in characters, using the width of the zero character + # from default font (first FONT record in the file). Excel adds some + # extra space to the default width, depending on the default font and + # default font size. The algorithm how to exactly calculate the resulting + # column width is not known.
+ # Example: The default width of 8 set in this record results in a column + # width of 8.43 using Arial font with a size of 10 points."""
+ # For the default hierarchy, refer to the {@link #Colinfo} class. + #
-- New in version 0.6.1 + defcolwidth = None + + ## + # Default column width from STANDARDWIDTH record, else None. + # From the OOo docs:
+ # """Default width of the columns in 1/256 of the width of the zero + # character, using default font (first FONT record in the file)."""
+ # For the default hierarchy, refer to the {@link #Colinfo} class. + #
-- New in version 0.6.1 + standardwidth = None + + ## + # Default value to be used for a row if there is + # no ROW record for that row. + # From the optional DEFAULTROWHEIGHT record. + default_row_height = None + + ## + # Default value to be used for a row if there is + # no ROW record for that row. + # From the optional DEFAULTROWHEIGHT record. + default_row_height_mismatch = None + + ## + # Default value to be used for a row if there is + # no ROW record for that row. + # From the optional DEFAULTROWHEIGHT record. + default_row_hidden = None + + ## + # Default value to be used for a row if there is + # no ROW record for that row. + # From the optional DEFAULTROWHEIGHT record. + default_additional_space_above = None + + ## + # Default value to be used for a row if there is + # no ROW record for that row. + # From the optional DEFAULTROWHEIGHT record. + default_additional_space_below = None + + ## + # Visibility of the sheet. 0 = visible, 1 = hidden (can be unhidden + # by user -- Format/Sheet/Unhide), 2 = "very hidden" (can be unhidden + # only by VBA macro). + visibility = 0 + + ## + # A 256-element tuple corresponding to the contents of the GCW record for this sheet. + # If no such record, treat as all bits zero. + # Applies to BIFF4-7 only. See docs of the {@link #Colinfo} class for discussion. + gcw = (0, ) * 256 + + ## + #

A list of {@link #Hyperlink} objects corresponding to HLINK records found + # in the worksheet.
-- New in version 0.7.2

+ hyperlink_list = [] + + ## + #

A sparse mapping from (rowx, colx) to an item in {@link #Sheet.hyperlink_list}. + # Cells not covered by a hyperlink are not mapped. + # It is possible using the Excel UI to set up a hyperlink that + # covers a larger-than-1x1 rectangle of cells. + # Hyperlink rectangles may overlap (Excel doesn't check). + # When a multiply-covered cell is clicked on, the hyperlink that is activated + # (and the one that is mapped here) is the last in hyperlink_list. + #
-- New in version 0.7.2

+ hyperlink_map = {} + + ## + #

A sparse mapping from (rowx, colx) to a {@link #Note} object. + # Cells not containing a note ("comment") are not mapped. + #
-- New in version 0.7.2

+ cell_note_map = {} + + ## + # Number of columns in left pane (frozen panes; for split panes, see comments below in code) + vert_split_pos = 0 + + ## + # Number of rows in top pane (frozen panes; for split panes, see comments below in code) + horz_split_pos = 0 + + ## + # Index of first visible row in bottom frozen/split pane + horz_split_first_visible = 0 + + ## + # Index of first visible column in right frozen/split pane + vert_split_first_visible = 0 + + ## + # Frozen panes: ignore it. Split panes: explanation and diagrams in OOo docs. + split_active_pane = 0 + + ## + # Boolean specifying if a PANE record was present, ignore unless you're xlutils.copy + has_pane_record = 0 + + ## + # A list of the horizontal page breaks in this sheet. + # Breaks are tuples in the form (index of row after break, start col index, end col index). + # Populated only if open_workbook(formatting_info=True). + #
-- New in version 0.7.2 + horizontal_page_breaks = [] + + ## + # A list of the vertical page breaks in this sheet. + # Breaks are tuples in the form (index of col after break, start row index, end row index). + # Populated only if open_workbook(formatting_info=True). + #
-- New in version 0.7.2 + vertical_page_breaks = [] + + + def __init__(self, book, position, name, number): + self.book = book + self.biff_version = book.biff_version + self._position = position + self.logfile = book.logfile + self.bt = array('B', [XL_CELL_EMPTY]) + self.bf = array('h', [-1]) + self.name = name + self.number = number + self.verbosity = book.verbosity + self.formatting_info = book.formatting_info + self.ragged_rows = book.ragged_rows + if self.ragged_rows: + self.put_cell = self.put_cell_ragged + else: + self.put_cell = self.put_cell_unragged + self._xf_index_to_xl_type_map = book._xf_index_to_xl_type_map + self.nrows = 0 # actual, including possibly empty cells + self.ncols = 0 + self._maxdatarowx = -1 # highest rowx containing a non-empty cell + self._maxdatacolx = -1 # highest colx containing a non-empty cell + self._dimnrows = 0 # as per DIMENSIONS record + self._dimncols = 0 + self._cell_values = [] + self._cell_types = [] + self._cell_xf_indexes = [] + self.defcolwidth = None + self.standardwidth = None + self.default_row_height = None + self.default_row_height_mismatch = 0 + self.default_row_hidden = 0 + self.default_additional_space_above = 0 + self.default_additional_space_below = 0 + self.colinfo_map = {} + self.rowinfo_map = {} + self.col_label_ranges = [] + self.row_label_ranges = [] + self.merged_cells = [] + self.rich_text_runlist_map = {} + self.horizontal_page_breaks = [] + self.vertical_page_breaks = [] + self._xf_index_stats = [0, 0, 0, 0] + self.visibility = book._sheet_visibility[number] # from BOUNDSHEET record + for attr, defval in _WINDOW2_options: + setattr(self, attr, defval) + self.first_visible_rowx = 0 + self.first_visible_colx = 0 + self.gridline_colour_index = 0x40 + self.gridline_colour_rgb = None # pre-BIFF8 + self.hyperlink_list = [] + self.hyperlink_map = {} + self.cell_note_map = {} + + # Values calculated by xlrd to predict the mag factors that + # will actually be used by Excel to display your worksheet. + # Pass these values to xlwt when writing XLS files. + # Warning 1: Behaviour of OOo Calc and Gnumeric has been observed to differ from Excel's. + # Warning 2: A value of zero means almost exactly what it says. Your sheet will be + # displayed as a very tiny speck on the screen. xlwt will reject attempts to set + # a mag_factor that is not (10 <= mag_factor <= 400). + self.cooked_page_break_preview_mag_factor = 60 + self.cooked_normal_view_mag_factor = 100 + + # Values (if any) actually stored on the XLS file + self.cached_page_break_preview_mag_factor = None # from WINDOW2 record + self.cached_normal_view_mag_factor = None # from WINDOW2 record + self.scl_mag_factor = None # from SCL record + + self._ixfe = None # BIFF2 only + self._cell_attr_to_xfx = {} # BIFF2.0 only + + #### Don't initialise this here, use class attribute initialisation. + #### self.gcw = (0, ) * 256 #### + + if self.biff_version >= 80: + self.utter_max_rows = 65536 + else: + self.utter_max_rows = 16384 + self.utter_max_cols = 256 + + self._first_full_rowx = -1 + + # self._put_cell_exceptions = 0 + # self._put_cell_row_widenings = 0 + # self._put_cell_rows_appended = 0 + # self._put_cell_cells_appended = 0 + + + ## + # {@link #Cell} object in the given row and column. + def cell(self, rowx, colx): + if self.formatting_info: + xfx = self.cell_xf_index(rowx, colx) + else: + xfx = None + return Cell( + self._cell_types[rowx][colx], + self._cell_values[rowx][colx], + xfx, + ) + + ## + # Value of the cell in the given row and column. + def cell_value(self, rowx, colx): + return self._cell_values[rowx][colx] + + ## + # Type of the cell in the given row and column. + # Refer to the documentation of the {@link #Cell} class. + def cell_type(self, rowx, colx): + return self._cell_types[rowx][colx] + + ## + # XF index of the cell in the given row and column. + # This is an index into Book.{@link #Book.xf_list}. + #
-- New in version 0.6.1 + def cell_xf_index(self, rowx, colx): + self.req_fmt_info() + xfx = self._cell_xf_indexes[rowx][colx] + if xfx > -1: + self._xf_index_stats[0] += 1 + return xfx + # Check for a row xf_index + try: + xfx = self.rowinfo_map[rowx].xf_index + if xfx > -1: + self._xf_index_stats[1] += 1 + return xfx + except KeyError: + pass + # Check for a column xf_index + try: + xfx = self.colinfo_map[colx].xf_index + if xfx == -1: xfx = 15 + self._xf_index_stats[2] += 1 + return xfx + except KeyError: + # If all else fails, 15 is used as hardwired global default xf_index. + self._xf_index_stats[3] += 1 + return 15 + + ## + # Returns the effective number of cells in the given row. For use with + # open_workbook(ragged_rows=True) which is likely to produce rows + # with fewer than {@link #Sheet.ncols} cells. + #
-- New in version 0.7.2 + def row_len(self, rowx): + return len(self._cell_values[rowx]) + + ## + # Returns a sequence of the {@link #Cell} objects in the given row. + def row(self, rowx): + return [ + self.cell(rowx, colx) + for colx in xrange(len(self._cell_values[rowx])) + ] + + ## + # Returns a generator for iterating through each row. + def get_rows(self): + return (self.row(index) for index in range(self.nrows)) + + ## + # Returns a slice of the types + # of the cells in the given row. + def row_types(self, rowx, start_colx=0, end_colx=None): + if end_colx is None: + return self._cell_types[rowx][start_colx:] + return self._cell_types[rowx][start_colx:end_colx] + + ## + # Returns a slice of the values + # of the cells in the given row. + def row_values(self, rowx, start_colx=0, end_colx=None): + if end_colx is None: + return self._cell_values[rowx][start_colx:] + return self._cell_values[rowx][start_colx:end_colx] + + ## + # Returns a slice of the {@link #Cell} objects in the given row. + def row_slice(self, rowx, start_colx=0, end_colx=None): + nc = len(self._cell_values[rowx]) + if start_colx < 0: + start_colx += nc + if start_colx < 0: + start_colx = 0 + if end_colx is None or end_colx > nc: + end_colx = nc + elif end_colx < 0: + end_colx += nc + return [ + self.cell(rowx, colx) + for colx in xrange(start_colx, end_colx) + ] + + ## + # Returns a slice of the {@link #Cell} objects in the given column. + def col_slice(self, colx, start_rowx=0, end_rowx=None): + nr = self.nrows + if start_rowx < 0: + start_rowx += nr + if start_rowx < 0: + start_rowx = 0 + if end_rowx is None or end_rowx > nr: + end_rowx = nr + elif end_rowx < 0: + end_rowx += nr + return [ + self.cell(rowx, colx) + for rowx in xrange(start_rowx, end_rowx) + ] + + ## + # Returns a slice of the values of the cells in the given column. + def col_values(self, colx, start_rowx=0, end_rowx=None): + nr = self.nrows + if start_rowx < 0: + start_rowx += nr + if start_rowx < 0: + start_rowx = 0 + if end_rowx is None or end_rowx > nr: + end_rowx = nr + elif end_rowx < 0: + end_rowx += nr + return [ + self._cell_values[rowx][colx] + for rowx in xrange(start_rowx, end_rowx) + ] + + ## + # Returns a slice of the types of the cells in the given column. + def col_types(self, colx, start_rowx=0, end_rowx=None): + nr = self.nrows + if start_rowx < 0: + start_rowx += nr + if start_rowx < 0: + start_rowx = 0 + if end_rowx is None or end_rowx > nr: + end_rowx = nr + elif end_rowx < 0: + end_rowx += nr + return [ + self._cell_types[rowx][colx] + for rowx in xrange(start_rowx, end_rowx) + ] + + ## + # Returns a sequence of the {@link #Cell} objects in the given column. + def col(self, colx): + return self.col_slice(colx) + # Above two lines just for the docs. Here's the real McCoy: + col = col_slice + + # === Following methods are used in building the worksheet. + # === They are not part of the API. + + def tidy_dimensions(self): + if self.verbosity >= 3: + fprintf(self.logfile, + "tidy_dimensions: nrows=%d ncols=%d \n", + self.nrows, self.ncols, + ) + if 1 and self.merged_cells: + nr = nc = 0 + umaxrows = self.utter_max_rows + umaxcols = self.utter_max_cols + for crange in self.merged_cells: + rlo, rhi, clo, chi = crange + if not (0 <= rlo < rhi <= umaxrows) \ + or not (0 <= clo < chi <= umaxcols): + fprintf(self.logfile, + "*** WARNING: sheet #%d (%r), MERGEDCELLS bad range %r\n", + self.number, self.name, crange) + if rhi > nr: nr = rhi + if chi > nc: nc = chi + if nc > self.ncols: + self.ncols = nc + if nr > self.nrows: + # we put one empty cell at (nr-1,0) to make sure + # we have the right number of rows. The ragged rows + # will sort out the rest if needed. + self.put_cell(nr-1, 0, XL_CELL_EMPTY, '', -1) + if self.verbosity >= 1 \ + and (self.nrows != self._dimnrows or self.ncols != self._dimncols): + fprintf(self.logfile, + "NOTE *** sheet %d (%r): DIMENSIONS R,C = %d,%d should be %d,%d\n", + self.number, + self.name, + self._dimnrows, + self._dimncols, + self.nrows, + self.ncols, + ) + if not self.ragged_rows: + # fix ragged rows + ncols = self.ncols + s_cell_types = self._cell_types + s_cell_values = self._cell_values + s_cell_xf_indexes = self._cell_xf_indexes + s_fmt_info = self.formatting_info + # for rowx in xrange(self.nrows): + if self._first_full_rowx == -2: + ubound = self.nrows + else: + ubound = self._first_full_rowx + for rowx in xrange(ubound): + trow = s_cell_types[rowx] + rlen = len(trow) + nextra = ncols - rlen + if nextra > 0: + s_cell_values[rowx][rlen:] = [''] * nextra + trow[rlen:] = self.bt * nextra + if s_fmt_info: + s_cell_xf_indexes[rowx][rlen:] = self.bf * nextra + + def put_cell_ragged(self, rowx, colx, ctype, value, xf_index): + if ctype is None: + # we have a number, so look up the cell type + ctype = self._xf_index_to_xl_type_map[xf_index] + assert 0 <= colx < self.utter_max_cols + assert 0 <= rowx < self.utter_max_rows + fmt_info = self.formatting_info + + try: + nr = rowx + 1 + if self.nrows < nr: + + scta = self._cell_types.append + scva = self._cell_values.append + scxa = self._cell_xf_indexes.append + bt = self.bt + bf = self.bf + for _unused in xrange(self.nrows, nr): + scta(bt * 0) + scva([]) + if fmt_info: + scxa(bf * 0) + self.nrows = nr + + types_row = self._cell_types[rowx] + values_row = self._cell_values[rowx] + if fmt_info: + fmt_row = self._cell_xf_indexes[rowx] + ltr = len(types_row) + if colx >= self.ncols: + self.ncols = colx + 1 + num_empty = colx - ltr + if not num_empty: + # most common case: colx == previous colx + 1 + # self._put_cell_cells_appended += 1 + types_row.append(ctype) + values_row.append(value) + if fmt_info: + fmt_row.append(xf_index) + return + if num_empty > 0: + num_empty += 1 + # self._put_cell_row_widenings += 1 + # types_row.extend(self.bt * num_empty) + # values_row.extend([''] * num_empty) + # if fmt_info: + # fmt_row.extend(self.bf * num_empty) + types_row[ltr:] = self.bt * num_empty + values_row[ltr:] = [''] * num_empty + if fmt_info: + fmt_row[ltr:] = self.bf * num_empty + types_row[colx] = ctype + values_row[colx] = value + if fmt_info: + fmt_row[colx] = xf_index + except: + print("put_cell", rowx, colx, file=self.logfile) + raise + + def put_cell_unragged(self, rowx, colx, ctype, value, xf_index): + if ctype is None: + # we have a number, so look up the cell type + ctype = self._xf_index_to_xl_type_map[xf_index] + # assert 0 <= colx < self.utter_max_cols + # assert 0 <= rowx < self.utter_max_rows + try: + self._cell_types[rowx][colx] = ctype + self._cell_values[rowx][colx] = value + if self.formatting_info: + self._cell_xf_indexes[rowx][colx] = xf_index + except IndexError: + # print >> self.logfile, "put_cell extending", rowx, colx + # self.extend_cells(rowx+1, colx+1) + # self._put_cell_exceptions += 1 + nr = rowx + 1 + nc = colx + 1 + assert 1 <= nc <= self.utter_max_cols + assert 1 <= nr <= self.utter_max_rows + if nc > self.ncols: + self.ncols = nc + # The row self._first_full_rowx and all subsequent rows + # are guaranteed to have length == self.ncols. Thus the + # "fix ragged rows" section of the tidy_dimensions method + # doesn't need to examine them. + if nr < self.nrows: + # cell data is not in non-descending row order *AND* + # self.ncols has been bumped up. + # This very rare case ruins this optmisation. + self._first_full_rowx = -2 + elif rowx > self._first_full_rowx > -2: + self._first_full_rowx = rowx + if nr <= self.nrows: + # New cell is in an existing row, so extend that row (if necessary). + # Note that nr < self.nrows means that the cell data + # is not in ascending row order!! + trow = self._cell_types[rowx] + nextra = self.ncols - len(trow) + if nextra > 0: + # self._put_cell_row_widenings += 1 + trow.extend(self.bt * nextra) + if self.formatting_info: + self._cell_xf_indexes[rowx].extend(self.bf * nextra) + self._cell_values[rowx].extend([''] * nextra) + else: + scta = self._cell_types.append + scva = self._cell_values.append + scxa = self._cell_xf_indexes.append + fmt_info = self.formatting_info + nc = self.ncols + bt = self.bt + bf = self.bf + for _unused in xrange(self.nrows, nr): + # self._put_cell_rows_appended += 1 + scta(bt * nc) + scva([''] * nc) + if fmt_info: + scxa(bf * nc) + self.nrows = nr + # === end of code from extend_cells() + try: + self._cell_types[rowx][colx] = ctype + self._cell_values[rowx][colx] = value + if self.formatting_info: + self._cell_xf_indexes[rowx][colx] = xf_index + except: + print("put_cell", rowx, colx, file=self.logfile) + raise + except: + print("put_cell", rowx, colx, file=self.logfile) + raise + + + # === Methods after this line neither know nor care about how cells are stored. + + def read(self, bk): + global rc_stats + DEBUG = 0 + blah = DEBUG or self.verbosity >= 2 + blah_rows = DEBUG or self.verbosity >= 4 + blah_formulas = 0 and blah + r1c1 = 0 + oldpos = bk._position + bk._position = self._position + XL_SHRFMLA_ETC_ETC = ( + XL_SHRFMLA, XL_ARRAY, XL_TABLEOP, XL_TABLEOP2, + XL_ARRAY2, XL_TABLEOP_B2, + ) + self_put_cell = self.put_cell + local_unpack = unpack + bk_get_record_parts = bk.get_record_parts + bv = self.biff_version + fmt_info = self.formatting_info + do_sst_rich_text = fmt_info and bk._rich_text_runlist_map + rowinfo_sharing_dict = {} + txos = {} + eof_found = 0 + while 1: + # if DEBUG: print "SHEET.READ: about to read from position %d" % bk._position + rc, data_len, data = bk_get_record_parts() + # if rc in rc_stats: + # rc_stats[rc] += 1 + # else: + # rc_stats[rc] = 1 + # if DEBUG: print "SHEET.READ: op 0x%04x, %d bytes %r" % (rc, data_len, data) + if rc == XL_NUMBER: + # [:14] in following stmt ignores extraneous rubbish at end of record. + # Sample file testEON-8.xls supplied by Jan Kraus. + rowx, colx, xf_index, d = local_unpack('> 15) & 1 + r.outline_level = bits2 & 7 + r.outline_group_starts_ends = (bits2 >> 4) & 1 + r.hidden = (bits2 >> 5) & 1 + r.height_mismatch = (bits2 >> 6) & 1 + r.has_default_xf_index = (bits2 >> 7) & 1 + r.xf_index = (bits2 >> 16) & 0xfff + r.additional_space_above = (bits2 >> 28) & 1 + r.additional_space_below = (bits2 >> 29) & 1 + if not r.has_default_xf_index: + r.xf_index = -1 + self.rowinfo_map[rowx] = r + if 0 and r.xf_index > -1: + fprintf(self.logfile, + "**ROW %d %d %d\n", + self.number, rowx, r.xf_index) + if blah_rows: + print('ROW', rowx, bits1, bits2, file=self.logfile) + r.dump(self.logfile, + header="--- sh #%d, rowx=%d ---" % (self.number, rowx)) + elif rc in XL_FORMULA_OPCODES: # 06, 0206, 0406 + # DEBUG = 1 + # if DEBUG: print "FORMULA: rc: 0x%04x data: %r" % (rc, data) + if bv >= 50: + rowx, colx, xf_index, result_str, flags = local_unpack('= 30: + rowx, colx, xf_index, result_str, flags = local_unpack(' 255: break # Excel does 0 to 256 inclusive + self.colinfo_map[colx] = c + if 0: + fprintf(self.logfile, + "**COL %d %d %d\n", + self.number, colx, c.xf_index) + if blah: + fprintf( + self.logfile, + "COLINFO sheet #%d cols %d-%d: wid=%d xf_index=%d flags=0x%04x\n", + self.number, first_colx, last_colx, c.width, c.xf_index, flags, + ) + c.dump(self.logfile, header='===') + elif rc == XL_DEFCOLWIDTH: + self.defcolwidth, = local_unpack(">= 1 + self.gcw = tuple(gcw) + if 0: + showgcw = "".join(map(lambda x: "F "[x], gcw)).rstrip().replace(' ', '.') + print("GCW:", showgcw, file=self.logfile) + elif rc == XL_BLANK: + if not fmt_info: continue + rowx, colx, xf_index = local_unpack('> self.logfile, "BLANK", rowx, colx, xf_index + self_put_cell(rowx, colx, XL_CELL_BLANK, '', xf_index) + elif rc == XL_MULBLANK: # 00BE + if not fmt_info: continue + nitems = data_len >> 1 + result = local_unpack("<%dH" % nitems, data) + rowx, mul_first = result[:2] + mul_last = result[-1] + # print >> self.logfile, "MULBLANK", rowx, mul_first, mul_last, data_len, nitems, mul_last + 4 - mul_first + assert nitems == mul_last + 4 - mul_first + pos = 2 + for colx in xrange(mul_first, mul_last + 1): + self_put_cell(rowx, colx, XL_CELL_BLANK, '', result[pos]) + pos += 1 + elif rc == XL_DIMENSION or rc == XL_DIMENSION2: + if data_len == 0: + # Four zero bytes after some other record. See github issue 64. + continue + # if data_len == 10: + # Was crashing on BIFF 4.0 file w/o the two trailing unused bytes. + # Reported by Ralph Heimburger. + if bv < 80: + dim_tuple = local_unpack(' found EOF", file=self.logfile) + elif rc == XL_COUNTRY: + bk.handle_country(data) + elif rc == XL_LABELRANGES: + pos = 0 + pos = unpack_cell_range_address_list_update_pos( + self.row_label_ranges, data, pos, bv, addr_size=8, + ) + pos = unpack_cell_range_address_list_update_pos( + self.col_label_ranges, data, pos, bv, addr_size=8, + ) + assert pos == data_len + elif rc == XL_ARRAY: + row1x, rownx, col1x, colnx, array_flags, tokslen = \ + local_unpack("= 80 + num_CFs, needs_recalc, browx1, browx2, bcolx1, bcolx2 = \ + unpack("<6H", data[0:12]) + if self.verbosity >= 1: + fprintf(self.logfile, + "\n*** WARNING: Ignoring CONDFMT (conditional formatting) record\n" \ + "*** in Sheet %d (%r).\n" \ + "*** %d CF record(s); needs_recalc_or_redraw = %d\n" \ + "*** Bounding box is %s\n", + self.number, self.name, num_CFs, needs_recalc, + rangename2d(browx1, browx2+1, bcolx1, bcolx2+1), + ) + olist = [] # updated by the function + pos = unpack_cell_range_address_list_update_pos( + olist, data, 12, bv, addr_size=8) + # print >> self.logfile, repr(result), len(result) + if self.verbosity >= 1: + fprintf(self.logfile, + "*** %d individual range(s):\n" \ + "*** %s\n", + len(olist), + ", ".join([rangename2d(*coords) for coords in olist]), + ) + elif rc == XL_CF: + if not fmt_info: continue + cf_type, cmp_op, sz1, sz2, flags = unpack("> 26) & 1 + bord_block = (flags >> 28) & 1 + patt_block = (flags >> 29) & 1 + if self.verbosity >= 1: + fprintf(self.logfile, + "\n*** WARNING: Ignoring CF (conditional formatting) sub-record.\n" \ + "*** cf_type=%d, cmp_op=%d, sz1=%d, sz2=%d, flags=0x%08x\n" \ + "*** optional data blocks: font=%d, border=%d, pattern=%d\n", + cf_type, cmp_op, sz1, sz2, flags, + font_block, bord_block, patt_block, + ) + # hex_char_dump(data, 0, data_len, fout=self.logfile) + pos = 12 + if font_block: + (font_height, font_options, weight, escapement, underline, + font_colour_index, two_bits, font_esc, font_underl) = \ + unpack("<64x i i H H B 3x i 4x i i i 18x", data[pos:pos+118]) + font_style = (two_bits > 1) & 1 + posture = (font_options > 1) & 1 + font_canc = (two_bits > 7) & 1 + cancellation = (font_options > 7) & 1 + if self.verbosity >= 1: + fprintf(self.logfile, + "*** Font info: height=%d, weight=%d, escapement=%d,\n" \ + "*** underline=%d, colour_index=%d, esc=%d, underl=%d,\n" \ + "*** style=%d, posture=%d, canc=%d, cancellation=%d\n", + font_height, weight, escapement, underline, + font_colour_index, font_esc, font_underl, + font_style, posture, font_canc, cancellation, + ) + pos += 118 + if bord_block: + pos += 8 + if patt_block: + pos += 4 + fmla1 = data[pos:pos+sz1] + pos += sz1 + if blah and sz1: + fprintf(self.logfile, + "*** formula 1:\n", + ) + dump_formula(bk, fmla1, sz1, bv, reldelta=0, blah=1) + fmla2 = data[pos:pos+sz2] + pos += sz2 + assert pos == data_len + if blah and sz2: + fprintf(self.logfile, + "*** formula 2:\n", + ) + dump_formula(bk, fmla2, sz2, bv, reldelta=0, blah=1) + elif rc == XL_DEFAULTROWHEIGHT: + if data_len == 4: + bits, self.default_row_height = unpack("> 1) & 1 + self.default_additional_space_above = (bits >> 2) & 1 + self.default_additional_space_below = (bits >> 3) & 1 + elif rc == XL_MERGEDCELLS: + if not fmt_info: continue + pos = unpack_cell_range_address_list_update_pos( + self.merged_cells, data, 0, bv, addr_size=8) + if blah: + fprintf(self.logfile, + "MERGEDCELLS: %d ranges\n", (pos - 2) // 8) + assert pos == data_len, \ + "MERGEDCELLS: pos=%d data_len=%d" % (pos, data_len) + elif rc == XL_WINDOW2: + if bv >= 80 and data_len >= 14: + (options, + self.first_visible_rowx, self.first_visible_colx, + self.gridline_colour_index, + self.cached_page_break_preview_mag_factor, + self.cached_normal_view_mag_factor + ) = unpack("= 30 # BIFF3-7 + (options, + self.first_visible_rowx, self.first_visible_colx, + ) = unpack(">= 1 + elif rc == XL_SCL: + num, den = unpack("= 0: + print(( + "WARNING *** SCL rcd sheet %d: should have 0.1 <= num/den <= 4; got %d/%d" + % (self.number, num, den) + ), file=self.logfile) + result = 100 + self.scl_mag_factor = result + elif rc == XL_PANE: + ( + self.vert_split_pos, + self.horz_split_pos, + self.horz_split_first_visible, + self.vert_split_first_visible, + self.split_active_pane, + ) = unpack("= 80)) + 2 == data_len + pos = 2 + if bv < 80: + while pos < data_len: + self.horizontal_page_breaks.append((local_unpack("= 80)) + 2 == data_len + pos = 2 + if bv < 80: + while pos < data_len: + self.vertical_page_breaks.append((local_unpack("> 15) & 1 + r.has_default_xf_index = bits2 & 1 + r.xf_index = xf_index + # r.outline_level = 0 # set in __init__ + # r.outline_group_starts_ends = 0 # set in __init__ + # r.hidden = 0 # set in __init__ + # r.height_mismatch = 0 # set in __init__ + # r.additional_space_above = 0 # set in __init__ + # r.additional_space_below = 0 # set in __init__ + self.rowinfo_map[rowx] = r + if 0 and r.xf_index > -1: + fprintf(self.logfile, + "**ROW %d %d %d\n", + self.number, rowx, r.xf_index) + if blah_rows: + print('ROW_B2', rowx, bits1, has_defaults, file=self.logfile) + r.dump(self.logfile, + header="--- sh #%d, rowx=%d ---" % (self.number, rowx)) + elif rc == XL_COLWIDTH: # BIFF2 only + if not fmt_info: continue + first_colx, last_colx, width\ + = local_unpack("= 30) + 1 + nchars_expected = unpack("<" + "BH"[lenlen - 1], data[:lenlen])[0] + offset = lenlen + if bv < 80: + enc = bk.encoding or bk.derive_encoding() + nchars_found = 0 + result = UNICODE_LITERAL("") + while 1: + if bv >= 80: + flag = BYTES_ORD(data[offset]) & 1 + enc = ("latin_1", "utf_16_le")[flag] + offset += 1 + chunk = unicode(data[offset:], enc) + result += chunk + nchars_found += len(chunk) + if nchars_found == nchars_expected: + return result + if nchars_found > nchars_expected: + msg = ("STRING/CONTINUE: expected %d chars, found %d" + % (nchars_expected, nchars_found)) + raise XLRDError(msg) + rc, _unused_len, data = bk.get_record_parts() + if rc != XL_CONTINUE: + raise XLRDError( + "Expected CONTINUE record; found record-type 0x%04X" % rc) + offset = 0 + + def update_cooked_mag_factors(self): + # Cached values are used ONLY for the non-active view mode. + # When the user switches to the non-active view mode, + # if the cached value for that mode is not valid, + # Excel pops up a window which says: + # "The number must be between 10 and 400. Try again by entering a number in this range." + # When the user hits OK, it drops into the non-active view mode + # but uses the magn from the active mode. + # NOTE: definition of "valid" depends on mode ... see below + blah = DEBUG or self.verbosity > 0 + if self.show_in_page_break_preview: + if self.scl_mag_factor is None: # no SCL record + self.cooked_page_break_preview_mag_factor = 100 # Yes, 100, not 60, NOT a typo + else: + self.cooked_page_break_preview_mag_factor = self.scl_mag_factor + zoom = self.cached_normal_view_mag_factor + if not (10 <= zoom <=400): + if blah: + print(( + "WARNING *** WINDOW2 rcd sheet %d: Bad cached_normal_view_mag_factor: %d" + % (self.number, self.cached_normal_view_mag_factor) + ), file=self.logfile) + zoom = self.cooked_page_break_preview_mag_factor + self.cooked_normal_view_mag_factor = zoom + else: + # normal view mode + if self.scl_mag_factor is None: # no SCL record + self.cooked_normal_view_mag_factor = 100 + else: + self.cooked_normal_view_mag_factor = self.scl_mag_factor + zoom = self.cached_page_break_preview_mag_factor + if zoom == 0: + # VALID, defaults to 60 + zoom = 60 + elif not (10 <= zoom <= 400): + if blah: + print(( + "WARNING *** WINDOW2 rcd sheet %r: Bad cached_page_break_preview_mag_factor: %r" + % (self.number, self.cached_page_break_preview_mag_factor) + ), file=self.logfile) + zoom = self.cooked_normal_view_mag_factor + self.cooked_page_break_preview_mag_factor = zoom + + def fixed_BIFF2_xfindex(self, cell_attr, rowx, colx, true_xfx=None): + DEBUG = 0 + blah = DEBUG or self.verbosity >= 2 + if self.biff_version == 21: + if self.book.xf_list: + if true_xfx is not None: + xfx = true_xfx + else: + xfx = BYTES_ORD(cell_attr[0]) & 0x3F + if xfx == 0x3F: + if self._ixfe is None: + raise XLRDError("BIFF2 cell record has XF index 63 but no preceding IXFE record.") + xfx = self._ixfe + # OOo docs are capable of interpretation that each + # cell record is preceded immediately by its own IXFE record. + # Empirical evidence is that (sensibly) an IXFE record applies to all + # following cell records until another IXFE comes along. + return xfx + # Have either Excel 2.0, or broken 2.1 w/o XF records -- same effect. + self.biff_version = self.book.biff_version = 20 + #### check that XF slot in cell_attr is zero + xfx_slot = BYTES_ORD(cell_attr[0]) & 0x3F + assert xfx_slot == 0 + xfx = self._cell_attr_to_xfx.get(cell_attr) + if xfx is not None: + return xfx + if blah: + fprintf(self.logfile, "New cell_attr %r at (%r, %r)\n", cell_attr, rowx, colx) + if not self.book.xf_list: + for xfx in xrange(16): + self.insert_new_BIFF20_xf(cell_attr=b"\x40\x00\x00", style=xfx < 15) + xfx = self.insert_new_BIFF20_xf(cell_attr=cell_attr) + return xfx + + def insert_new_BIFF20_xf(self, cell_attr, style=0): + DEBUG = 0 + blah = DEBUG or self.verbosity >= 2 + book = self.book + xfx = len(book.xf_list) + xf = self.fake_XF_from_BIFF20_cell_attr(cell_attr, style) + xf.xf_index = xfx + book.xf_list.append(xf) + if blah: + xf.dump(self.logfile, header="=== Faked XF %d ===" % xfx, footer="======") + if xf.format_key not in book.format_map: + if xf.format_key: + msg = "ERROR *** XF[%d] unknown format key (%d, 0x%04x)\n" + fprintf(self.logfile, msg, + xf.xf_index, xf.format_key, xf.format_key) + fmt = Format(xf.format_key, FUN, UNICODE_LITERAL("General")) + book.format_map[xf.format_key] = fmt + book.format_list.append(fmt) + cellty_from_fmtty = { + FNU: XL_CELL_NUMBER, + FUN: XL_CELL_NUMBER, + FGE: XL_CELL_NUMBER, + FDT: XL_CELL_DATE, + FTX: XL_CELL_NUMBER, # Yes, a number can be formatted as text. + } + fmt = book.format_map[xf.format_key] + cellty = cellty_from_fmtty[fmt.type] + self._xf_index_to_xl_type_map[xf.xf_index] = cellty + self._cell_attr_to_xfx[cell_attr] = xfx + return xfx + + def fake_XF_from_BIFF20_cell_attr(self, cell_attr, style=0): + from .formatting import XF, XFAlignment, XFBorder, XFBackground, XFProtection + xf = XF() + xf.alignment = XFAlignment() + xf.alignment.indent_level = 0 + xf.alignment.shrink_to_fit = 0 + xf.alignment.text_direction = 0 + xf.border = XFBorder() + xf.border.diag_up = 0 + xf.border.diag_down = 0 + xf.border.diag_colour_index = 0 + xf.border.diag_line_style = 0 # no line + xf.background = XFBackground() + xf.protection = XFProtection() + (prot_bits, font_and_format, halign_etc) = unpack('> 6 + upkbits(xf.protection, prot_bits, ( + (6, 0x40, 'cell_locked'), + (7, 0x80, 'formula_hidden'), + )) + xf.alignment.hor_align = halign_etc & 0x07 + for mask, side in ((0x08, 'left'), (0x10, 'right'), (0x20, 'top'), (0x40, 'bottom')): + if halign_etc & mask: + colour_index, line_style = 8, 1 # black, thin + else: + colour_index, line_style = 0, 0 # none, none + setattr(xf.border, side + '_colour_index', colour_index) + setattr(xf.border, side + '_line_style', line_style) + bg = xf.background + if halign_etc & 0x80: + bg.fill_pattern = 17 + else: + bg.fill_pattern = 0 + bg.background_colour_index = 9 # white + bg.pattern_colour_index = 8 # black + xf.parent_style_index = (0x0FFF, 0)[style] + xf.alignment.vert_align = 2 # bottom + xf.alignment.rotation = 0 + for attr_stem in \ + "format font alignment border background protection".split(): + attr = "_" + attr_stem + "_flag" + setattr(xf, attr, 1) + return xf + + def req_fmt_info(self): + if not self.formatting_info: + raise XLRDError("Feature requires open_workbook(..., formatting_info=True)") + + ## + # Determine column display width. + #
-- New in version 0.6.1 + #
+ # @param colx Index of the queried column, range 0 to 255. + # Note that it is possible to find out the width that will be used to display + # columns with no cell information e.g. column IV (colx=255). + # @return The column width that will be used for displaying + # the given column by Excel, in units of 1/256th of the width of a + # standard character (the digit zero in the first font). + + def computed_column_width(self, colx): + self.req_fmt_info() + if self.biff_version >= 80: + colinfo = self.colinfo_map.get(colx, None) + if colinfo is not None: + return colinfo.width + if self.standardwidth is not None: + return self.standardwidth + elif self.biff_version >= 40: + if self.gcw[colx]: + if self.standardwidth is not None: + return self.standardwidth + else: + colinfo = self.colinfo_map.get(colx, None) + if colinfo is not None: + return colinfo.width + elif self.biff_version == 30: + colinfo = self.colinfo_map.get(colx, None) + if colinfo is not None: + return colinfo.width + # All roads lead to Rome and the DEFCOLWIDTH ... + if self.defcolwidth is not None: + return self.defcolwidth * 256 + return 8 * 256 # 8 is what Excel puts in a DEFCOLWIDTH record + + def handle_hlink(self, data): + # DEBUG = 1 + if DEBUG: print("\n=== hyperlink ===", file=self.logfile) + record_size = len(data) + h = Hyperlink() + h.frowx, h.lrowx, h.fcolx, h.lcolx, guid0, dummy, options = unpack(' 0: + fprintf( + self.logfile, + "*** WARNING: hyperlink at r=%d c=%d has %d extra data bytes: %s\n", + h.frowx, + h.fcolx, + extra_nbytes, + REPR(data[-extra_nbytes:]) + ) + # Seen: b"\x00\x00" also b"A\x00", b"V\x00" + elif extra_nbytes < 0: + raise XLRDError("Bug or corrupt file, send copy of input file for debugging") + + self.hyperlink_list.append(h) + for rowx in xrange(h.frowx, h.lrowx+1): + for colx in xrange(h.fcolx, h.lcolx+1): + self.hyperlink_map[rowx, colx] = h + + def handle_quicktip(self, data): + rcx, frowx, lrowx, fcolx, lcolx = unpack('<5H', data[:10]) + assert rcx == XL_QUICKTIP + assert self.hyperlink_list + h = self.hyperlink_list[-1] + assert (frowx, lrowx, fcolx, lcolx) == (h.frowx, h.lrowx, h.fcolx, h.lcolx) + assert data[-2:] == b'\x00\x00' + h.quicktip = unicode(data[10:-2], 'utf_16_le') + + def handle_msodrawingetc(self, recid, data_len, data): + if not OBJ_MSO_DEBUG: + return + DEBUG = 1 + if self.biff_version < 80: + return + o = MSODrawing() + pos = 0 + while pos < data_len: + tmp, fbt, cb = unpack('> 4) & 0xFFF + if ver == 0xF: + ndb = 0 # container + else: + ndb = cb + if DEBUG: + hex_char_dump(data, pos, ndb + 8, base=0, fout=self.logfile) + fprintf(self.logfile, + "fbt:0x%04X inst:%d ver:0x%X cb:%d (0x%04X)\n", + fbt, inst, ver, cb, cb) + if fbt == 0xF010: # Client Anchor + assert ndb == 18 + (o.anchor_unk, + o.anchor_colx_lo, o.anchor_rowx_lo, + o.anchor_colx_hi, o.anchor_rowx_hi) = unpack(' 0: + rc2, data2_len, data2 = self.book.get_record_parts() + assert rc2 == XL_NOTE + dummy_rowx, nb = unpack('> 1) & 1 + o.row_hidden = (option_flags >> 7) & 1 + o.col_hidden = (option_flags >> 8) & 1 + # XL97 dev kit book says NULL [sic] bytes padding between string count and string data + # to ensure that string is word-aligned. Appears to be nonsense. + o.author, endpos = unpack_unicode_update_pos(data, 8, lenlen=2) + # There is a random/undefined byte after the author string (not counted in the + # string length). + # Issue 4 on github: Google Spreadsheet doesn't write the undefined byte. + assert (data_len - endpos) in (0, 1) + if OBJ_MSO_DEBUG: + o.dump(self.logfile, header="=== Note ===", footer= " ") + txo = txos.get(o._object_id) + if txo: + o.text = txo.text + o.rich_text_runlist = txo.rich_text_runlist + self.cell_note_map[o.rowx, o.colx] = o + + def handle_txo(self, data): + if self.biff_version < 80: + return + o = MSTxo() + data_len = len(data) + fmt = ' Represents a user "comment" or "note". +# Note objects are accessible through Sheet.{@link #Sheet.cell_note_map}. +#
-- New in version 0.7.2 +#

+class Note(BaseObject): + ## + # Author of note + author = UNICODE_LITERAL('') + ## + # True if the containing column is hidden + col_hidden = 0 + ## + # Column index + colx = 0 + ## + # List of (offset_in_string, font_index) tuples. + # Unlike Sheet.{@link #Sheet.rich_text_runlist_map}, the first offset should always be 0. + rich_text_runlist = None + ## + # True if the containing row is hidden + row_hidden = 0 + ## + # Row index + rowx = 0 + ## + # True if note is always shown + show = 0 + ## + # Text of the note + text = UNICODE_LITERAL('') + +## +#

Contains the attributes of a hyperlink. +# Hyperlink objects are accessible through Sheet.{@link #Sheet.hyperlink_list} +# and Sheet.{@link #Sheet.hyperlink_map}. +#
-- New in version 0.7.2 +#

+class Hyperlink(BaseObject): + ## + # Index of first row + frowx = None + ## + # Index of last row + lrowx = None + ## + # Index of first column + fcolx = None + ## + # Index of last column + lcolx = None + ## + # Type of hyperlink. Unicode string, one of 'url', 'unc', + # 'local file', 'workbook', 'unknown' + type = None + ## + # The URL or file-path, depending in the type. Unicode string, except + # in the rare case of a local but non-existent file with non-ASCII + # characters in the name, in which case only the "8.3" filename is available, + # as a bytes (3.x) or str (2.x) string, with unknown encoding. + url_or_path = None + ## + # Description ... this is displayed in the cell, + # and should be identical to the cell value. Unicode string, or None. It seems + # impossible NOT to have a description created by the Excel UI. + desc = None + ## + # Target frame. Unicode string. Note: I have not seen a case of this. + # It seems impossible to create one in the Excel UI. + target = None + ## + # "Textmark": the piece after the "#" in + # "http://docs.python.org/library#struct_module", or the Sheet1!A1:Z99 + # part when type is "workbook". + textmark = None + ## + # The text of the "quick tip" displayed when the cursor + # hovers over the hyperlink. + quicktip = None + +# === helpers === + +def unpack_RK(rk_str): + flags = BYTES_ORD(rk_str[0]) + if flags & 2: + # There's a SIGNED 30-bit integer in there! + i, = unpack('>= 2 # div by 4 to drop the 2 flag bits + if flags & 1: + return i / 100.0 + return float(i) + else: + # It's the most significant 30 bits of an IEEE 754 64-bit FP number + d, = unpack('Contains the data for one cell.

+# +#

WARNING: You don't call this class yourself. You access Cell objects +# via methods of the {@link #Sheet} object(s) that you found in the {@link #Book} object that +# was returned when you called xlrd.open_workbook("myfile.xls").

+#

Cell objects have three attributes: ctype is an int, value +# (which depends on ctype) and xf_index. +# If "formatting_info" is not enabled when the workbook is opened, xf_index will be None. +# The following table describes the types of cells and how their values +# are represented in Python.

+# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +# +#
Type symbolType numberPython value
XL_CELL_EMPTY0empty string u''
XL_CELL_TEXT1a Unicode string
XL_CELL_NUMBER2float
XL_CELL_DATE3float
XL_CELL_BOOLEAN4int; 1 means TRUE, 0 means FALSE
XL_CELL_ERROR5int representing internal Excel codes; for a text representation, +# refer to the supplied dictionary error_text_from_code
XL_CELL_BLANK6empty string u''. Note: this type will appear only when +# open_workbook(..., formatting_info=True) is used.
+#

+ +class Cell(BaseObject): + + __slots__ = ['ctype', 'value', 'xf_index'] + + def __init__(self, ctype, value, xf_index=None): + self.ctype = ctype + self.value = value + self.xf_index = xf_index + + def __repr__(self): + if self.xf_index is None: + return "%s:%r" % (ctype_text[self.ctype], self.value) + else: + return "%s:%r (XF:%r)" % (ctype_text[self.ctype], self.value, self.xf_index) + +## +# There is one and only one instance of an empty cell -- it's a singleton. This is it. +# You may use a test like "acell is empty_cell". +empty_cell = Cell(XL_CELL_EMPTY, '') + +##### =============== Colinfo and Rowinfo ============================== ##### + +## +# Width and default formatting information that applies to one or +# more columns in a sheet. Derived from COLINFO records. +# +#

Here is the default hierarchy for width, according to the OOo docs: +# +#
"""In BIFF3, if a COLINFO record is missing for a column, +# the width specified in the record DEFCOLWIDTH is used instead. +# +#
In BIFF4-BIFF7, the width set in this [COLINFO] record is only used, +# if the corresponding bit for this column is cleared in the GCW +# record, otherwise the column width set in the DEFCOLWIDTH record +# is used (the STANDARDWIDTH record is always ignored in this case [see footnote!]). +# +#
In BIFF8, if a COLINFO record is missing for a column, +# the width specified in the record STANDARDWIDTH is used. +# If this [STANDARDWIDTH] record is also missing, +# the column width of the record DEFCOLWIDTH is used instead.""" +#
+# +# Footnote: The docs on the GCW record say this: +# """
+# If a bit is set, the corresponding column uses the width set in the STANDARDWIDTH +# record. If a bit is cleared, the corresponding column uses the width set in the +# COLINFO record for this column. +#
If a bit is set, and the worksheet does not contain the STANDARDWIDTH record, or if +# the bit is cleared, and the worksheet does not contain the COLINFO record, the DEFCOLWIDTH +# record of the worksheet will be used instead. +#
"""
+# At the moment (2007-01-17) xlrd is going with the GCW version of the story. +# Reference to the source may be useful: see the computed_column_width(colx) method +# of the Sheet class. +#
-- New in version 0.6.1 +#

+ +class Colinfo(BaseObject): + ## + # Width of the column in 1/256 of the width of the zero character, + # using default font (first FONT record in the file). + width = 0 + ## + # XF index to be used for formatting empty cells. + xf_index = -1 + ## + # 1 = column is hidden + hidden = 0 + ## + # Value of a 1-bit flag whose purpose is unknown + # but is often seen set to 1 + bit1_flag = 0 + ## + # Outline level of the column, in range(7). + # (0 = no outline) + outline_level = 0 + ## + # 1 = column is collapsed + collapsed = 0 + +_USE_SLOTS = 1 + +## +#

Height and default formatting information that applies to a row in a sheet. +# Derived from ROW records. +#
-- New in version 0.6.1

+# +#

height: Height of the row, in twips. One twip == 1/20 of a point.

+# +#

has_default_height: 0 = Row has custom height; 1 = Row has default height.

+# +#

outline_level: Outline level of the row (0 to 7)

+# +#

outline_group_starts_ends: 1 = Outline group starts or ends here (depending on where the +# outline buttons are located, see WSBOOL record [TODO ??]), +# and is collapsed

+# +#

hidden: 1 = Row is hidden (manually, or by a filter or outline group)

+# +#

height_mismatch: 1 = Row height and default font height do not match

+# +#

has_default_xf_index: 1 = the xf_index attribute is usable; 0 = ignore it

+# +#

xf_index: Index to default XF record for empty cells in this row. +# Don't use this if has_default_xf_index == 0.

+# +#

additional_space_above: This flag is set, if the upper border of at least one cell in this row +# or if the lower border of at least one cell in the row above is +# formatted with a thick line style. Thin and medium line styles are not +# taken into account.

+# +#

additional_space_below: This flag is set, if the lower border of at least one cell in this row +# or if the upper border of at least one cell in the row below is +# formatted with a medium or thick line style. Thin line styles are not +# taken into account.

+ +class Rowinfo(BaseObject): + + if _USE_SLOTS: + __slots__ = ( + "height", + "has_default_height", + "outline_level", + "outline_group_starts_ends", + "hidden", + "height_mismatch", + "has_default_xf_index", + "xf_index", + "additional_space_above", + "additional_space_below", + ) + + def __init__(self): + self.height = None + self.has_default_height = None + self.outline_level = None + self.outline_group_starts_ends = None + self.hidden = None + self.height_mismatch = None + self.has_default_xf_index = None + self.xf_index = None + self.additional_space_above = None + self.additional_space_below = None + + def __getstate__(self): + return ( + self.height, + self.has_default_height, + self.outline_level, + self.outline_group_starts_ends, + self.hidden, + self.height_mismatch, + self.has_default_xf_index, + self.xf_index, + self.additional_space_above, + self.additional_space_below, + ) + + def __setstate__(self, state): + ( + self.height, + self.has_default_height, + self.outline_level, + self.outline_group_starts_ends, + self.hidden, + self.height_mismatch, + self.has_default_xf_index, + self.xf_index, + self.additional_space_above, + self.additional_space_below, + ) = state diff --git a/Assignment 3/packages/xlrd/timemachine.py b/Assignment 3/packages/xlrd/timemachine.py new file mode 100644 index 0000000..a068db3 --- /dev/null +++ b/Assignment 3/packages/xlrd/timemachine.py @@ -0,0 +1,52 @@ +## +#

Copyright (c) 2006-2012 Stephen John Machin, Lingfo Pty Ltd

+#

This module is part of the xlrd package, which is released under a BSD-style licence.

+## + +# timemachine.py -- adaptation for single codebase. +# Currently supported: 2.6 to 2.7, 3.2+ +# usage: from timemachine import * + +from __future__ import print_function +import sys + +python_version = sys.version_info[:2] # e.g. version 2.6 -> (2, 6) + +if python_version >= (3, 0): + # Python 3 + BYTES_LITERAL = lambda x: x.encode('latin1') + UNICODE_LITERAL = lambda x: x + BYTES_ORD = lambda byte: byte + from io import BytesIO as BYTES_IO + def fprintf(f, fmt, *vargs): + fmt = fmt.replace("%r", "%a") + if fmt.endswith('\n'): + print(fmt[:-1] % vargs, file=f) + else: + print(fmt % vargs, end=' ', file=f) + EXCEL_TEXT_TYPES = (str, bytes, bytearray) # xlwt: isinstance(obj, EXCEL_TEXT_TYPES) + REPR = ascii + xrange = range + unicode = lambda b, enc: b.decode(enc) + ensure_unicode = lambda s: s + unichr = chr +else: + # Python 2 + BYTES_LITERAL = lambda x: x + UNICODE_LITERAL = lambda x: x.decode('latin1') + BYTES_ORD = ord + from cStringIO import StringIO as BYTES_IO + def fprintf(f, fmt, *vargs): + if fmt.endswith('\n'): + print(fmt[:-1] % vargs, file=f) + else: + print(fmt % vargs, end=' ', file=f) + try: + EXCEL_TEXT_TYPES = basestring # xlwt: isinstance(obj, EXCEL_TEXT_TYPES) + except NameError: + EXCEL_TEXT_TYPES = (str, unicode) + REPR = repr + xrange = xrange + # following used only to overcome 2.x ElementTree gimmick which + # returns text as `str` if it's ascii, otherwise `unicode` + ensure_unicode = unicode # used only in xlsx.py diff --git a/Assignment 3/packages/xlrd/xldate.py b/Assignment 3/packages/xlrd/xldate.py new file mode 100644 index 0000000..dc7b9c8 --- /dev/null +++ b/Assignment 3/packages/xlrd/xldate.py @@ -0,0 +1,213 @@ +# -*- coding: cp1252 -*- + +# No part of the content of this file was derived from the works of David Giffin. + +## +#

Copyright © 2005-2008 Stephen John Machin, Lingfo Pty Ltd

+#

This module is part of the xlrd package, which is released under a BSD-style licence.

+# +#

Provides function(s) for dealing with Microsoft Excel ™ dates.

+## + +# 2008-10-18 SJM Fix bug in xldate_from_date_tuple (affected some years after 2099) + +# The conversion from days to (year, month, day) starts with +# an integral "julian day number" aka JDN. +# FWIW, JDN 0 corresponds to noon on Monday November 24 in Gregorian year -4713. +# More importantly: +# Noon on Gregorian 1900-03-01 (day 61 in the 1900-based system) is JDN 2415080.0 +# Noon on Gregorian 1904-01-02 (day 1 in the 1904-based system) is JDN 2416482.0 +import datetime + +_JDN_delta = (2415080 - 61, 2416482 - 1) +assert _JDN_delta[1] - _JDN_delta[0] == 1462 + +# Pre-calculate the datetime epochs for efficiency. +epoch_1904 = datetime.datetime(1904, 1, 1) +epoch_1900 = datetime.datetime(1899, 12, 31) +epoch_1900_minus_1 = datetime.datetime(1899, 12, 30) + +class XLDateError(ValueError): pass + +class XLDateNegative(XLDateError): pass +class XLDateAmbiguous(XLDateError): pass +class XLDateTooLarge(XLDateError): pass +class XLDateBadDatemode(XLDateError): pass +class XLDateBadTuple(XLDateError): pass + +_XLDAYS_TOO_LARGE = (2958466, 2958466 - 1462) # This is equivalent to 10000-01-01 + +## +# Convert an Excel number (presumed to represent a date, a datetime or a time) into +# a tuple suitable for feeding to datetime or mx.DateTime constructors. +# @param xldate The Excel number +# @param datemode 0: 1900-based, 1: 1904-based. +#
WARNING: when using this function to +# interpret the contents of a workbook, you should pass in the Book.datemode +# attribute of that workbook. Whether +# the workbook has ever been anywhere near a Macintosh is irrelevant. +# @return Gregorian (year, month, day, hour, minute, nearest_second). +#
Special case: if 0.0 <= xldate < 1.0, it is assumed to represent a time; +# (0, 0, 0, hour, minute, second) will be returned. +#
Note: 1904-01-01 is not regarded as a valid date in the datemode 1 system; its "serial number" +# is zero. +# @throws XLDateNegative xldate < 0.00 +# @throws XLDateAmbiguous The 1900 leap-year problem (datemode == 0 and 1.0 <= xldate < 61.0) +# @throws XLDateTooLarge Gregorian year 10000 or later +# @throws XLDateBadDatemode datemode arg is neither 0 nor 1 +# @throws XLDateError Covers the 4 specific errors + +def xldate_as_tuple(xldate, datemode): + if datemode not in (0, 1): + raise XLDateBadDatemode(datemode) + if xldate == 0.00: + return (0, 0, 0, 0, 0, 0) + if xldate < 0.00: + raise XLDateNegative(xldate) + xldays = int(xldate) + frac = xldate - xldays + seconds = int(round(frac * 86400.0)) + assert 0 <= seconds <= 86400 + if seconds == 86400: + hour = minute = second = 0 + xldays += 1 + else: + # second = seconds % 60; minutes = seconds // 60 + minutes, second = divmod(seconds, 60) + # minute = minutes % 60; hour = minutes // 60 + hour, minute = divmod(minutes, 60) + if xldays >= _XLDAYS_TOO_LARGE[datemode]: + raise XLDateTooLarge(xldate) + + if xldays == 0: + return (0, 0, 0, hour, minute, second) + + if xldays < 61 and datemode == 0: + raise XLDateAmbiguous(xldate) + + jdn = xldays + _JDN_delta[datemode] + yreg = ((((jdn * 4 + 274277) // 146097) * 3 // 4) + jdn + 1363) * 4 + 3 + mp = ((yreg % 1461) // 4) * 535 + 333 + d = ((mp % 16384) // 535) + 1 + # mp /= 16384 + mp >>= 14 + if mp >= 10: + return ((yreg // 1461) - 4715, mp - 9, d, hour, minute, second) + else: + return ((yreg // 1461) - 4716, mp + 3, d, hour, minute, second) + + +## +# Convert an Excel date/time number into a datetime.datetime object. +# +# @param xldate The Excel number +# @param datemode 0: 1900-based, 1: 1904-based. +# +# @return a datetime.datetime() object. +# +def xldate_as_datetime(xldate, datemode): + """Convert an Excel date/time number into a datetime.datetime object.""" + + # Set the epoch based on the 1900/1904 datemode. + if datemode: + epoch = epoch_1904 + else: + if xldate < 60: + epoch = epoch_1900 + else: + # Workaround Excel 1900 leap year bug by adjusting the epoch. + epoch = epoch_1900_minus_1 + + # The integer part of the Excel date stores the number of days since + # the epoch and the fractional part stores the percentage of the day. + days = int(xldate) + fraction = xldate - days + + # Get the the integer and decimal seconds in Excel's millisecond resolution. + seconds = int(round(fraction * 86400000.0)) + seconds, milliseconds = divmod(seconds, 1000) + + return epoch + datetime.timedelta(days, seconds, 0, milliseconds) + + +# === conversions from date/time to xl numbers + +def _leap(y): + if y % 4: return 0 + if y % 100: return 1 + if y % 400: return 0 + return 1 + +_days_in_month = (None, 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31) + +## +# Convert a date tuple (year, month, day) to an Excel date. +# @param year Gregorian year. +# @param month 1 <= month <= 12 +# @param day 1 <= day <= last day of that (year, month) +# @param datemode 0: 1900-based, 1: 1904-based. +# @throws XLDateAmbiguous The 1900 leap-year problem (datemode == 0 and 1.0 <= xldate < 61.0) +# @throws XLDateBadDatemode datemode arg is neither 0 nor 1 +# @throws XLDateBadTuple (year, month, day) is too early/late or has invalid component(s) +# @throws XLDateError Covers the specific errors + +def xldate_from_date_tuple(date_tuple, datemode): + """Create an excel date from a tuple of (year, month, day)""" + year, month, day = date_tuple + + if datemode not in (0, 1): + raise XLDateBadDatemode(datemode) + + if year == 0 and month == 0 and day == 0: + return 0.00 + + if not (1900 <= year <= 9999): + raise XLDateBadTuple("Invalid year: %r" % ((year, month, day),)) + if not (1 <= month <= 12): + raise XLDateBadTuple("Invalid month: %r" % ((year, month, day),)) + if day < 1 \ + or (day > _days_in_month[month] and not(day == 29 and month == 2 and _leap(year))): + raise XLDateBadTuple("Invalid day: %r" % ((year, month, day),)) + + Yp = year + 4716 + M = month + if M <= 2: + Yp = Yp - 1 + Mp = M + 9 + else: + Mp = M - 3 + jdn = (1461 * Yp // 4) + ((979 * Mp + 16) // 32) + \ + day - 1364 - (((Yp + 184) // 100) * 3 // 4) + xldays = jdn - _JDN_delta[datemode] + if xldays <= 0: + raise XLDateBadTuple("Invalid (year, month, day): %r" % ((year, month, day),)) + if xldays < 61 and datemode == 0: + raise XLDateAmbiguous("Before 1900-03-01: %r" % ((year, month, day),)) + return float(xldays) + +## +# Convert a time tuple (hour, minute, second) to an Excel "date" value (fraction of a day). +# @param hour 0 <= hour < 24 +# @param minute 0 <= minute < 60 +# @param second 0 <= second < 60 +# @throws XLDateBadTuple Out-of-range hour, minute, or second + +def xldate_from_time_tuple(time_tuple): + """Create an excel date from a tuple of (hour, minute, second)""" + hour, minute, second = time_tuple + if 0 <= hour < 24 and 0 <= minute < 60 and 0 <= second < 60: + return ((second / 60.0 + minute) / 60.0 + hour) / 24.0 + raise XLDateBadTuple("Invalid (hour, minute, second): %r" % ((hour, minute, second),)) + +## +# Convert a datetime tuple (year, month, day, hour, minute, second) to an Excel date value. +# For more details, refer to other xldate_from_*_tuple functions. +# @param datetime_tuple (year, month, day, hour, minute, second) +# @param datemode 0: 1900-based, 1: 1904-based. + +def xldate_from_datetime_tuple(datetime_tuple, datemode): + return ( + xldate_from_date_tuple(datetime_tuple[:3], datemode) + + + xldate_from_time_tuple(datetime_tuple[3:]) + ) diff --git a/Assignment 3/packages/xlrd/xlsx.py b/Assignment 3/packages/xlrd/xlsx.py new file mode 100644 index 0000000..89f9753 --- /dev/null +++ b/Assignment 3/packages/xlrd/xlsx.py @@ -0,0 +1,804 @@ +## +# Portions copyright (c) 2008-2012 Stephen John Machin, Lingfo Pty Ltd +# This module is part of the xlrd package, which is released under a BSD-style licence. +## + +from __future__ import print_function, unicode_literals + +DEBUG = 0 + +import sys +import re +from .timemachine import * +from .book import Book, Name +from .biffh import error_text_from_code, XLRDError, XL_CELL_BLANK, XL_CELL_TEXT, XL_CELL_BOOLEAN, XL_CELL_ERROR +from .formatting import is_date_format_string, Format, XF +from .sheet import Sheet + +DLF = sys.stdout # Default Log File + +ET = None +ET_has_iterparse = False +Element_has_iter = False + +def ensure_elementtree_imported(verbosity, logfile): + global ET, ET_has_iterparse, Element_has_iter + if ET is not None: + return + if "IronPython" in sys.version: + import xml.etree.ElementTree as ET + #### 2.7.2.1: fails later with + #### NotImplementedError: iterparse is not supported on IronPython. (CP #31923) + else: + try: import xml.etree.cElementTree as ET + except ImportError: + try: import cElementTree as ET + except ImportError: + try: import lxml.etree as ET + except ImportError: + try: import xml.etree.ElementTree as ET + except ImportError: + try: import elementtree.ElementTree as ET + except ImportError: + raise Exception("Failed to import an ElementTree implementation") + if hasattr(ET, 'iterparse'): + _dummy_stream = BYTES_IO(b'') + try: + ET.iterparse(_dummy_stream) + ET_has_iterparse = True + except NotImplementedError: + pass + Element_has_iter = hasattr(ET.ElementTree, 'iter') + if verbosity: + etree_version = repr([ + (item, getattr(ET, item)) + for item in ET.__dict__.keys() + if item.lower().replace('_', '') == 'version' + ]) + print(ET.__file__, ET.__name__, etree_version, ET_has_iterparse, file=logfile) + +def split_tag(tag): + pos = tag.rfind('}') + 1 + if pos >= 2: + return tag[:pos], tag[pos:] + return '', tag + +def augment_keys(adict, uri): + # uri must already be enclosed in {} + for x in list(adict.keys()): + adict[uri + x] = adict[x] + +_UPPERCASE_1_REL_INDEX = {} # Used in fast conversion of column names (e.g. "XFD") to indices (16383) +for _x in xrange(26): + _UPPERCASE_1_REL_INDEX["ABCDEFGHIJKLMNOPQRSTUVWXYZ"[_x]] = _x + 1 +for _x in "123456789": + _UPPERCASE_1_REL_INDEX[_x] = 0 +del _x + +def cell_name_to_rowx_colx(cell_name, letter_value=_UPPERCASE_1_REL_INDEX): + # Extract column index from cell name + # A => 0, Z =>25, AA => 26, XFD => 16383 + colx = 0 + charx = -1 + try: + for c in cell_name: + charx += 1 + lv = letter_value[c] + if lv: + colx = colx * 26 + lv + else: # start of row number; can't be '0' + colx = colx - 1 + assert 0 <= colx < X12_MAX_COLS + break + except KeyError: + raise Exception('Unexpected character %r in cell name %r' % (c, cell_name)) + rowx = int(cell_name[charx:]) - 1 + return rowx, colx + +error_code_from_text = {} +for _code, _text in error_text_from_code.items(): + error_code_from_text[_text] = _code + +# === X12 === Excel 2007 .xlsx =============================================== + +U_SSML12 = "{http://schemas.openxmlformats.org/spreadsheetml/2006/main}" +U_ODREL = "{http://schemas.openxmlformats.org/officeDocument/2006/relationships}" +U_PKGREL = "{http://schemas.openxmlformats.org/package/2006/relationships}" +U_CP = "{http://schemas.openxmlformats.org/package/2006/metadata/core-properties}" +U_DC = "{http://purl.org/dc/elements/1.1/}" +U_DCTERMS = "{http://purl.org/dc/terms/}" +XML_SPACE_ATTR = "{http://www.w3.org/XML/1998/namespace}space" +XML_WHITESPACE = "\t\n \r" +X12_MAX_ROWS = 2 ** 20 +X12_MAX_COLS = 2 ** 14 +V_TAG = U_SSML12 + 'v' # cell child: value +F_TAG = U_SSML12 + 'f' # cell child: formula +IS_TAG = U_SSML12 + 'is' # cell child: inline string + +def unescape(s, + subber=re.compile(r'_x[0-9A-Fa-f]{4,4}_', re.UNICODE).sub, + repl=lambda mobj: unichr(int(mobj.group(0)[2:6], 16)), + ): + if "_" in s: + return subber(repl, s) + return s + +def cooked_text(self, elem): + t = elem.text + if t is None: + return '' + if elem.get(XML_SPACE_ATTR) != 'preserve': + t = t.strip(XML_WHITESPACE) + return ensure_unicode(unescape(t)) + +def get_text_from_si_or_is(self, elem, r_tag=U_SSML12+'r', t_tag=U_SSML12 +'t'): + "Returns unescaped unicode" + accum = [] + for child in elem: + # self.dump_elem(child) + tag = child.tag + if tag == t_tag: + t = cooked_text(self, child) + if t: # note: .text attribute can be None + accum.append(t) + elif tag == r_tag: + for tnode in child: + if tnode.tag == t_tag: + t = cooked_text(self, tnode) + if t: + accum.append(t) + return ''.join(accum) + +def map_attributes(amap, elem, obj): + for xml_attr, obj_attr, cnv_func_or_const in amap: + if not xml_attr: + setattr(obj, obj_attr, cnv_func_or_const) + continue + if not obj_attr: continue #### FIX ME #### + raw_value = elem.get(xml_attr) + cooked_value = cnv_func_or_const(raw_value) + setattr(obj, obj_attr, cooked_value) + +def cnv_ST_Xstring(s): + if s is None: return "" + return ensure_unicode(s) + +def cnv_xsd_unsignedInt(s): + if not s: + return None + value = int(s) + assert value >= 0 + return value + +def cnv_xsd_boolean(s): + if not s: + return 0 + if s in ("1", "true", "on"): + return 1 + if s in ("0", "false", "off"): + return 0 + raise ValueError("unexpected xsd:boolean value: %r" % s) + + +_defined_name_attribute_map = ( + ("name", "name", cnv_ST_Xstring, ), + ("comment", "", cnv_ST_Xstring, ), + ("customMenu", "", cnv_ST_Xstring, ), + ("description", "", cnv_ST_Xstring, ), + ("help", "", cnv_ST_Xstring, ), + ("statusBar", "", cnv_ST_Xstring, ), + ("localSheetId", "scope", cnv_xsd_unsignedInt, ), + ("hidden", "hidden", cnv_xsd_boolean, ), + ("function", "func", cnv_xsd_boolean, ), + ("vbProcedure", "vbasic", cnv_xsd_boolean, ), + ("xlm", "macro", cnv_xsd_boolean, ), + ("functionGroupId", "funcgroup", cnv_xsd_unsignedInt, ), + ("shortcutKey", "", cnv_ST_Xstring, ), + ("publishToServer", "", cnv_xsd_boolean, ), + ("workbookParameter", "", cnv_xsd_boolean, ), + ("", "any_err", 0, ), + ("", "any_external", 0, ), + ("", "any_rel", 0, ), + ("", "basic_formula_len", 0, ), + ("", "binary", 0, ), + ("", "builtin", 0, ), + ("", "complex", 0, ), + ("", "evaluated", 0, ), + ("", "excel_sheet_index", 0, ), + ("", "excel_sheet_num", 0, ), + ("", "option_flags", 0, ), + ("", "result", None, ), + ("", "stack", None, ), + ) + +def make_name_access_maps(bk): + name_and_scope_map = {} # (name.lower(), scope): Name_object + name_map = {} # name.lower() : list of Name_objects (sorted in scope order) + num_names = len(bk.name_obj_list) + for namex in xrange(num_names): + nobj = bk.name_obj_list[namex] + name_lcase = nobj.name.lower() + key = (name_lcase, nobj.scope) + if key in name_and_scope_map: + msg = 'Duplicate entry %r in name_and_scope_map' % (key, ) + if 0: + raise XLRDError(msg) + else: + if bk.verbosity: + print(msg, file=bk.logfile) + name_and_scope_map[key] = nobj + sort_data = (nobj.scope, namex, nobj) + if name_lcase in name_map: + name_map[name_lcase].append(sort_data) + else: + name_map[name_lcase] = [sort_data] + for key in name_map.keys(): + alist = name_map[key] + alist.sort() + name_map[key] = [x[2] for x in alist] + bk.name_and_scope_map = name_and_scope_map + bk.name_map = name_map + +class X12General(object): + + def process_stream(self, stream, heading=None): + if self.verbosity >= 2 and heading is not None: + fprintf(self.logfile, "\n=== %s ===\n", heading) + self.tree = ET.parse(stream) + getmethod = self.tag2meth.get + for elem in self.tree.iter() if Element_has_iter else self.tree.getiterator(): + if self.verbosity >= 3: + self.dump_elem(elem) + meth = getmethod(elem.tag) + if meth: + meth(self, elem) + self.finish_off() + + def finish_off(self): + pass + + def dump_elem(self, elem): + fprintf(self.logfile, + "===\ntag=%r len=%d attrib=%r text=%r tail=%r\n", + split_tag(elem.tag)[1], len(elem), elem.attrib, elem.text, elem.tail) + + def dumpout(self, fmt, *vargs): + text = (12 * ' ' + fmt + '\n') % vargs + self.logfile.write(text) + +class X12Book(X12General): + + def __init__(self, bk, logfile=DLF, verbosity=False): + self.bk = bk + self.logfile = logfile + self.verbosity = verbosity + self.bk.nsheets = 0 + self.bk.props = {} + self.relid2path = {} + self.relid2reltype = {} + self.sheet_targets = [] # indexed by sheetx + self.sheetIds = [] # indexed by sheetx + + core_props_menu = { + U_CP+"lastModifiedBy": ("last_modified_by", cnv_ST_Xstring), + U_DC+"creator": ("creator", cnv_ST_Xstring), + U_DCTERMS+"modified": ("modified", cnv_ST_Xstring), + U_DCTERMS+"created": ("created", cnv_ST_Xstring), + } + + def process_coreprops(self, stream): + if self.verbosity >= 2: + fprintf(self.logfile, "\n=== coreProps ===\n") + self.tree = ET.parse(stream) + getmenu = self.core_props_menu.get + props = {} + for elem in self.tree.iter() if Element_has_iter else self.tree.getiterator(): + if self.verbosity >= 3: + self.dump_elem(elem) + menu = getmenu(elem.tag) + if menu: + attr, func = menu + value = func(elem.text) + props[attr] = value + self.bk.user_name = props.get('last_modified_by') or props.get('creator') + self.bk.props = props + if self.verbosity >= 2: + fprintf(self.logfile, "props: %r\n", props) + self.finish_off() + + def process_rels(self, stream): + if self.verbosity >= 2: + fprintf(self.logfile, "\n=== Relationships ===\n") + tree = ET.parse(stream) + r_tag = U_PKGREL + 'Relationship' + for elem in tree.findall(r_tag): + rid = elem.get('Id') + target = elem.get('Target') + reltype = elem.get('Type').split('/')[-1] + if self.verbosity >= 2: + self.dumpout('Id=%r Type=%r Target=%r', rid, reltype, target) + self.relid2reltype[rid] = reltype + # self.relid2path[rid] = 'xl/' + target + if target.startswith('/'): + self.relid2path[rid] = target[1:] # drop the / + else: + self.relid2path[rid] = 'xl/' + target + + def do_defined_name(self, elem): + #### UNDER CONSTRUCTION #### + if 0 and self.verbosity >= 3: + self.dump_elem(elem) + nobj = Name() + bk = self.bk + nobj.bk = bk + nobj.name_index = len(bk.name_obj_list) + bk.name_obj_list.append(nobj) + nobj.name = elem.get('name') + nobj.raw_formula = None # compiled bytecode formula -- not in XLSX + nobj.formula_text = cooked_text(self, elem) + map_attributes(_defined_name_attribute_map, elem, nobj) + if nobj.scope is None: + nobj.scope = -1 # global + if nobj.name.startswith("_xlnm."): + nobj.builtin = 1 + if self.verbosity >= 2: + nobj.dump(header='=== Name object ===') + + def do_defined_names(self, elem): + for child in elem: + self.do_defined_name(child) + make_name_access_maps(self.bk) + + def do_sheet(self, elem): + bk = self.bk + sheetx = bk.nsheets + # print elem.attrib + rid = elem.get(U_ODREL + 'id') + sheetId = int(elem.get('sheetId')) + name = unescape(ensure_unicode(elem.get('name'))) + reltype = self.relid2reltype[rid] + target = self.relid2path[rid] + if self.verbosity >= 2: + self.dumpout( + 'sheetx=%d sheetId=%r rid=%r type=%r name=%r', + sheetx, sheetId, rid, reltype, name) + if reltype != 'worksheet': + if self.verbosity >= 2: + self.dumpout('Ignoring sheet of type %r (name=%r)', reltype, name) + return + state = elem.get('state') + visibility_map = { + None: 0, + 'visible': 0, + 'hidden': 1, + 'veryHidden': 2 + } + bk._sheet_visibility.append(visibility_map[state]) + sheet = Sheet(bk, position=None, name=name, number=sheetx) + sheet.utter_max_rows = X12_MAX_ROWS + sheet.utter_max_cols = X12_MAX_COLS + bk._sheet_list.append(sheet) + bk._sheet_names.append(name) + bk.nsheets += 1 + self.sheet_targets.append(target) + self.sheetIds.append(sheetId) + + + def do_workbookpr(self, elem): + datemode = cnv_xsd_boolean(elem.get('date1904')) + if self.verbosity >= 2: + self.dumpout('datemode=%r', datemode) + self.bk.datemode = datemode + + tag2meth = { + 'definedNames': do_defined_names, + 'workbookPr': do_workbookpr, + 'sheet': do_sheet, + } + augment_keys(tag2meth, U_SSML12) + +class X12SST(X12General): + + def __init__(self, bk, logfile=DLF, verbosity=0): + self.bk = bk + self.logfile = logfile + self.verbosity = verbosity + if ET_has_iterparse: + self.process_stream = self.process_stream_iterparse + else: + self.process_stream = self.process_stream_findall + + def process_stream_iterparse(self, stream, heading=None): + if self.verbosity >= 2 and heading is not None: + fprintf(self.logfile, "\n=== %s ===\n", heading) + si_tag = U_SSML12 + 'si' + elemno = -1 + sst = self.bk._sharedstrings + for event, elem in ET.iterparse(stream): + if elem.tag != si_tag: continue + elemno = elemno + 1 + if self.verbosity >= 3: + fprintf(self.logfile, "element #%d\n", elemno) + self.dump_elem(elem) + result = get_text_from_si_or_is(self, elem) + sst.append(result) + elem.clear() # destroy all child elements + if self.verbosity >= 2: + self.dumpout('Entries in SST: %d', len(sst)) + if self.verbosity >= 3: + for x, s in enumerate(sst): + fprintf(self.logfile, "SST x=%d s=%r\n", x, s) + + def process_stream_findall(self, stream, heading=None): + if self.verbosity >= 2 and heading is not None: + fprintf(self.logfile, "\n=== %s ===\n", heading) + self.tree = ET.parse(stream) + si_tag = U_SSML12 + 'si' + elemno = -1 + sst = self.bk._sharedstrings + for elem in self.tree.findall(si_tag): + elemno = elemno + 1 + if self.verbosity >= 3: + fprintf(self.logfile, "element #%d\n", elemno) + self.dump_elem(elem) + result = get_text_from_si_or_is(self, elem) + sst.append(result) + if self.verbosity >= 2: + self.dumpout('Entries in SST: %d', len(sst)) + +class X12Styles(X12General): + + def __init__(self, bk, logfile=DLF, verbosity=0): + self.bk = bk + self.logfile = logfile + self.verbosity = verbosity + self.xf_counts = [0, 0] + self.xf_type = None + self.fmt_is_date = {} + for x in list(range(14, 23)) + list(range(45, 48)): #### hard-coding FIX ME #### + self.fmt_is_date[x] = 1 + # dummy entry for XF 0 in case no Styles section + self.bk._xf_index_to_xl_type_map[0] = 2 + # fill_in_standard_formats(bk) #### pre-integration kludge + + def do_cellstylexfs(self, elem): + self.xf_type = 0 + + def do_cellxfs(self, elem): + self.xf_type = 1 + + def do_numfmt(self, elem): + formatCode = ensure_unicode(elem.get('formatCode')) + numFmtId = int(elem.get('numFmtId')) + is_date = is_date_format_string(self.bk, formatCode) + self.fmt_is_date[numFmtId] = is_date + fmt_obj = Format(numFmtId, is_date + 2, formatCode) + self.bk.format_map[numFmtId] = fmt_obj + if self.verbosity >= 3: + self.dumpout('numFmtId=%d formatCode=%r is_date=%d', numFmtId, formatCode, is_date) + + def do_xf(self, elem): + if self.xf_type != 1: + #### ignoring style XFs for the moment + return + xfx = self.xf_counts[self.xf_type] + self.xf_counts[self.xf_type] = xfx + 1 + xf = XF() + self.bk.xf_list.append(xf) + self.bk.xfcount += 1 + numFmtId = int(elem.get('numFmtId', '0')) + xf.format_key = numFmtId + is_date = self.fmt_is_date.get(numFmtId, 0) + self.bk._xf_index_to_xl_type_map[xfx] = is_date + 2 + if self.verbosity >= 3: + self.dumpout( + 'xfx=%d numFmtId=%d', + xfx, numFmtId, + ) + self.dumpout(repr(self.bk._xf_index_to_xl_type_map)) + + tag2meth = { + 'cellStyleXfs': do_cellstylexfs, + 'cellXfs': do_cellxfs, + 'numFmt': do_numfmt, + 'xf': do_xf, + } + augment_keys(tag2meth, U_SSML12) + +class X12Sheet(X12General): + + def __init__(self, sheet, logfile=DLF, verbosity=0): + self.sheet = sheet + self.logfile = logfile + self.verbosity = verbosity + self.rowx = -1 # We may need to count them. + self.bk = sheet.book + self.sst = self.bk._sharedstrings + self.merged_cells = sheet.merged_cells + self.warned_no_cell_name = 0 + self.warned_no_row_num = 0 + if ET_has_iterparse: + self.process_stream = self.own_process_stream + + def own_process_stream(self, stream, heading=None): + if self.verbosity >= 2 and heading is not None: + fprintf(self.logfile, "\n=== %s ===\n", heading) + getmethod = self.tag2meth.get + row_tag = U_SSML12 + "row" + self_do_row = self.do_row + for event, elem in ET.iterparse(stream): + if elem.tag == row_tag: + self_do_row(elem) + elem.clear() # destroy all child elements (cells) + elif elem.tag == U_SSML12 + "dimension": + self.do_dimension(elem) + elif elem.tag == U_SSML12 + "mergeCell": + self.do_merge_cell(elem) + self.finish_off() + + def process_comments_stream(self, stream): + root = ET.parse(stream).getroot() + author_list = root[0] + assert author_list.tag == U_SSML12 + 'authors' + authors = [elem.text for elem in author_list] + comment_list = root[1] + assert comment_list.tag == U_SSML12 + 'commentList' + cell_note_map = self.sheet.cell_note_map + from .sheet import Note + text_tag = U_SSML12 + 'text' + r_tag = U_SSML12 + 'r' + t_tag = U_SSML12 + 't' + for elem in comment_list.findall(U_SSML12 + 'comment'): + ts = elem.findall('./' + text_tag + '/' + t_tag) + ts += elem.findall('./' + text_tag + '/' + r_tag + '/' + t_tag) + ref = elem.get('ref') + note = Note() + note.author = authors[int(elem.get('authorId'))] + note.rowx, note.colx = coords = cell_name_to_rowx_colx(ref) + note.text = '' + for t in ts: + note.text += cooked_text(self, t) + cell_note_map[coords] = note + + def do_dimension(self, elem): + ref = elem.get('ref') # example: "A1:Z99" or just "A1" + if ref: + # print >> self.logfile, "dimension: ref=%r" % ref + last_cell_ref = ref.split(':')[-1] # example: "Z99" + rowx, colx = cell_name_to_rowx_colx(last_cell_ref) + self.sheet._dimnrows = rowx + 1 + self.sheet._dimncols = colx + 1 + + def do_merge_cell(self, elem): + # The ref attribute should be a cell range like "B1:D5". + ref = elem.get('ref') + if ref: + first_cell_ref, last_cell_ref = ref.split(':') + first_rowx, first_colx = cell_name_to_rowx_colx(first_cell_ref) + last_rowx, last_colx = cell_name_to_rowx_colx(last_cell_ref) + self.merged_cells.append((first_rowx, last_rowx + 1, + first_colx, last_colx + 1)) + + def do_row(self, row_elem): + + def bad_child_tag(child_tag): + raise Exception('cell type %s has unexpected child <%s> at rowx=%r colx=%r' % (cell_type, child_tag, rowx, colx)) + + row_number = row_elem.get('r') + if row_number is None: # Yes, it's optional. + self.rowx += 1 + explicit_row_number = 0 + if self.verbosity and not self.warned_no_row_num: + self.dumpout("no row number; assuming rowx=%d", self.rowx) + self.warned_no_row_num = 1 + else: + self.rowx = int(row_number) - 1 + explicit_row_number = 1 + assert 0 <= self.rowx < X12_MAX_ROWS + rowx = self.rowx + colx = -1 + if self.verbosity >= 3: + self.dumpout(" row_number=%r rowx=%d explicit=%d", + row_number, self.rowx, explicit_row_number) + letter_value = _UPPERCASE_1_REL_INDEX + for cell_elem in row_elem: + cell_name = cell_elem.get('r') + if cell_name is None: # Yes, it's optional. + colx += 1 + if self.verbosity and not self.warned_no_cell_name: + self.dumpout("no cellname; assuming rowx=%d colx=%d", rowx, colx) + self.warned_no_cell_name = 1 + else: + # Extract column index from cell name + # A => 0, Z =>25, AA => 26, XFD => 16383 + colx = 0 + charx = -1 + try: + for c in cell_name: + charx += 1 + if c == '$': + continue + lv = letter_value[c] + if lv: + colx = colx * 26 + lv + else: # start of row number; can't be '0' + colx = colx - 1 + assert 0 <= colx < X12_MAX_COLS + break + except KeyError: + raise Exception('Unexpected character %r in cell name %r' % (c, cell_name)) + if explicit_row_number and cell_name[charx:] != row_number: + raise Exception('cell name %r but row number is %r' % (cell_name, row_number)) + xf_index = int(cell_elem.get('s', '0')) + cell_type = cell_elem.get('t', 'n') + tvalue = None + formula = None + if cell_type == 'n': + # n = number. Most frequent type. + # child contains plain text which can go straight into float() + # OR there's no text in which case it's a BLANK cell + for child in cell_elem: + child_tag = child.tag + if child_tag == V_TAG: + tvalue = child.text + elif child_tag == F_TAG: + formula = cooked_text(self, child) + else: + raise Exception('unexpected tag %r' % child_tag) + if not tvalue: + if self.bk.formatting_info: + self.sheet.put_cell(rowx, colx, XL_CELL_BLANK, '', xf_index) + else: + self.sheet.put_cell(rowx, colx, None, float(tvalue), xf_index) + elif cell_type == "s": + # s = index into shared string table. 2nd most frequent type + # child contains plain text which can go straight into int() + for child in cell_elem: + child_tag = child.tag + if child_tag == V_TAG: + tvalue = child.text + elif child_tag == F_TAG: + # formula not expected here, but gnumeric does it. + formula = child.text + else: + bad_child_tag(child_tag) + if not tvalue: + # + if self.bk.formatting_info: + self.sheet.put_cell(rowx, colx, XL_CELL_BLANK, '', xf_index) + else: + value = self.sst[int(tvalue)] + self.sheet.put_cell(rowx, colx, XL_CELL_TEXT, value, xf_index) + elif cell_type == "str": + # str = string result from formula. + # Should have (formula) child; however in one file, all text cells are str with no formula. + # child can contain escapes + for child in cell_elem: + child_tag = child.tag + if child_tag == V_TAG: + tvalue = cooked_text(self, child) + elif child_tag == F_TAG: + formula = cooked_text(self, child) + else: + bad_child_tag(child_tag) + # assert tvalue is not None and formula is not None + # Yuk. Fails with file created by gnumeric -- no tvalue! + self.sheet.put_cell(rowx, colx, XL_CELL_TEXT, tvalue, xf_index) + elif cell_type == "b": + # b = boolean + # child contains "0" or "1" + # Maybe the data should be converted with cnv_xsd_boolean; + # ECMA standard is silent; Excel 2007 writes 0 or 1 + for child in cell_elem: + child_tag = child.tag + if child_tag == V_TAG: + tvalue = child.text + elif child_tag == F_TAG: + formula = cooked_text(self, child) + else: + bad_child_tag(child_tag) + self.sheet.put_cell(rowx, colx, XL_CELL_BOOLEAN, int(tvalue), xf_index) + elif cell_type == "e": + # e = error + # child contains e.g. "#REF!" + for child in cell_elem: + child_tag = child.tag + if child_tag == V_TAG: + tvalue = child.text + elif child_tag == F_TAG: + formula = cooked_text(self, child) + else: + bad_child_tag(child_tag) + value = error_code_from_text[tvalue] + self.sheet.put_cell(rowx, colx, XL_CELL_ERROR, value, xf_index) + elif cell_type == "inlineStr": + # Not expected in files produced by Excel. + # Only possible child is . + # It's a way of allowing 3rd party s/w to write text (including rich text) cells + # without having to build a shared string table + for child in cell_elem: + child_tag = child.tag + if child_tag == IS_TAG: + tvalue = get_text_from_si_or_is(self, child) + else: + bad_child_tag(child_tag) + assert tvalue is not None + self.sheet.put_cell(rowx, colx, XL_CELL_TEXT, tvalue, xf_index) + else: + raise Exception("Unknown cell type %r in rowx=%d colx=%d" % (cell_type, rowx, colx)) + + tag2meth = { + 'row': do_row, + } + augment_keys(tag2meth, U_SSML12) + +def open_workbook_2007_xml( + zf, + component_names, + logfile=sys.stdout, + verbosity=0, + use_mmap=0, + formatting_info=0, + on_demand=0, + ragged_rows=0, + ): + ensure_elementtree_imported(verbosity, logfile) + bk = Book() + bk.logfile = logfile + bk.verbosity = verbosity + bk.formatting_info = formatting_info + if formatting_info: + raise NotImplementedError("formatting_info=True not yet implemented") + bk.use_mmap = False #### Not supported initially + bk.on_demand = on_demand + if on_demand: + if verbosity: + print("WARNING *** on_demand=True not yet implemented; falling back to False", file=bk.logfile) + bk.on_demand = False + bk.ragged_rows = ragged_rows + + x12book = X12Book(bk, logfile, verbosity) + zflo = zf.open(component_names['xl/_rels/workbook.xml.rels']) + x12book.process_rels(zflo) + del zflo + zflo = zf.open(component_names['xl/workbook.xml']) + x12book.process_stream(zflo, 'Workbook') + del zflo + props_name = 'docprops/core.xml' + if props_name in component_names: + zflo = zf.open(component_names[props_name]) + x12book.process_coreprops(zflo) + + x12sty = X12Styles(bk, logfile, verbosity) + if 'xl/styles.xml' in component_names: + zflo = zf.open(component_names['xl/styles.xml']) + x12sty.process_stream(zflo, 'styles') + del zflo + else: + # seen in MS sample file MergedCells.xlsx + pass + + sst_fname = 'xl/sharedstrings.xml' + x12sst = X12SST(bk, logfile, verbosity) + if sst_fname in component_names: + zflo = zf.open(component_names[sst_fname]) + x12sst.process_stream(zflo, 'SST') + del zflo + + for sheetx in range(bk.nsheets): + fname = x12book.sheet_targets[sheetx] + zflo = zf.open(component_names[fname]) + sheet = bk._sheet_list[sheetx] + x12sheet = X12Sheet(sheet, logfile, verbosity) + heading = "Sheet %r (sheetx=%d) from %r" % (sheet.name, sheetx, fname) + x12sheet.process_stream(zflo, heading) + del zflo + comments_fname = 'xl/comments%d.xml' % (sheetx + 1) + if comments_fname in component_names: + comments_stream = zf.open(component_names[comments_fname]) + x12sheet.process_comments_stream(comments_stream) + del comments_stream + + sheet.tidy_dimensions() + + return bk -- cgit v1.2.3