Source Code for Module src.inspectors.ooo_meta_inspector

  1  #!/usr/bin/python 
  2  # -*- coding: UTF-8 -*- 
  3  # 
  4  #The MIT License 
  5  # 
  6  #Copyright (c) 2011 
  7  # 
  8  #Permission is hereby granted, free of charge, to any person obtaining a copy 
  9  #of this software and associated documentation files (the "Software"), to deal 
 10  #in the Software without restriction, including without limitation the rights 
 11  #to use, copy, modify, merge, publish, distribute, sublicense, and/or sell 
 12  #copies of the Software, and to permit persons to whom the Software is 
 13  #furnished to do so, subject to the following conditions: 
 14  # 
 15  #The above copyright notice and this permission notice shall be included in 
 16  #all copies or substantial portions of the Software. 
 17  # 
 18  #THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 
 19  #IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 
 20  #FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 
 21  #AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 
 22  #LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 
 23  #OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN 
 24  #THE SOFTWARE. 
 25  # 
 26  #Authors: 
 27  #   Vili Auvinen (vili.k.auvinen@jyu.fi) 
 28  #   Olli Kauppinen (olli.kauppinen@jyu.fi) 
 29  #   Juho Tammela (juho.i.tammela@jyu.fi) 
 30   
 31  ''' 
 32  The module provides the methods for getting OpenOffice.org file formats' odt, odp, and ods meta informations. 
 33   
 34  @author: Olli Kauppinen 
 35  ''' 
 36   
 37  import sys 
 38  import zipfile 
 39  import xml.dom.minidom 
 40  import fnmatch 
 41     
 42          
 43 -def getMetaInformation(documentDict,elementTagName): 
 44      '''Gets the meta information by the given elementTagName from odt,odp and ods documents. 
 45       
 46      @return: The meta information. 
 47      ''' 
 48      element=documentDict['meta.xml'].getElementsByTagName(elementTagName) 
 49      if element: 
 50          return element[0].firstChild.nodeValue 
 51      else: 
 52          return 'not defined' 
 53       
 54 -def getDocumentStatistic(documentDict,attributeName): 
 55      '''Gets the document statistic by the given attributeName. 
 56       
 57      @return: The statistic value. 
 58      ''' 
 59      element=documentDict['meta.xml'].getElementsByTagName('meta:document-statistic') 
 60      if element: 
 61          return element[0].getAttribute (attributeName) 
 62      else: 
 63          return 'Statistic not defined' 
 64   
 65       
 66 -def getMeta(documentDict): 
 67      '''Gets all the document meta information. 
 68       
 69      @return: The meta information in the dictionary. Tag names are the key values in the dictionary. 
 70      ''' 
 71      meta = {'dc:title': None, 
 72      'meta:creation-date': None, 
 73      'meta:initial-creator': None, 
 74      'meta:editing-duration': None, 
 75      'dc:date': None, 
 76      'dc:creator': None, 
 77      'meta:editing-cycles': None} 
 78       
 79      for metaName in meta.keys(): 
 80          meta[metaName] = getMetaInformation(documentDict,metaName) 
 81          
 82      return meta  
 83       
 84   
 85 -def getOdtStatistic (documentDict): 
 86      '''Gets all the odt document statistic information. 
 87       
 88      @return: The statistic information in the dictionary. Attribute names are the key values in the dictionary.  
 89      ''' 
 90      odtStatistic = {'meta:table-count': None, 
 91      'meta:image-count': None, 
 92      'meta:object-count': None, 
 93      'meta:page-count': None, 
 94      'meta:paragraph-count': None, 
 95      'meta:paragraph-count': None, 
 96      'meta:word-count': None, 
 97      'meta:character-count': None} 
 98       
 99      for odtStatisticName in odtStatistic.keys(): 
100          odtStatistic[odtStatisticName] = getDocumentStatistic(documentDict,odtStatisticName) 
101       
102      return odtStatistic 
103       
104 -def getOdsStatistic (documentDict): 
105      '''Gets all the ods document statistic information. 
106       
107      @return: The statistic information in the dictionary. Attribute names are the key values in the dictionary. 
108      ''' 
109      odsStatistic = {'meta:table-count': None, 
110      'meta:cell-count': None, 
111      'meta:object-count': None} 
112                 
113      for odsStatisticName in odtStatistic.keys(): 
114          odsStatistic[odtStatisticName] = getDocumentStatistic(documentDict,odtStatisticName) 
115       
116      return odsStatistic 
117   
118 -def printOdpStatistic (): 
119      '''Get all the odp document statistic information. 
120       
121      @return: The statistic information in the dictionary. Attribute names are the key values in the dictionary. 
122      ''' 
123      odpStatistic = {'meta:object-count': None} 
124      odpStatistic['meta:object-count'] = getDocumentStatistic(documentDict,odtStatisticName) 
125      return odpStatistic 
126