biblicit 1.0 → 2.0.3

Sign up to get free protection for your applications and to get access to all the features.
Files changed (406) hide show
  1. data/.gitmodules +3 -0
  2. data/Gemfile +1 -1
  3. data/README.md +125 -30
  4. data/Rakefile +22 -0
  5. data/biblicit.gemspec +9 -7
  6. data/lib/biblicit/cb2bib.rb +10 -11
  7. data/lib/biblicit/citeseer.rb +14 -26
  8. data/lib/biblicit/extractor.rb +40 -19
  9. data/lib/biblicit/parscit.rb +38 -0
  10. data/parscit/.gitignore +8 -0
  11. data/parscit/CHANGELOG +125 -0
  12. data/parscit/COPYING +674 -0
  13. data/parscit/COPYING.LESSER +165 -0
  14. data/parscit/INSTALL +105 -0
  15. data/parscit/README +97 -0
  16. data/{perl/ParsCit/README.TXT → parscit/USAGE} +25 -15
  17. data/parscit/bin/archtest.pl +31 -0
  18. data/parscit/bin/citeExtract.pl +562 -0
  19. data/parscit/bin/conlleval.pl +315 -0
  20. data/parscit/bin/headExtract.pl +40 -0
  21. data/parscit/bin/parsHed/convert2TokenLevel.pl +138 -0
  22. data/parscit/bin/parsHed/keywordGen.pl +308 -0
  23. data/parscit/bin/parsHed/parseXmlHeader.pl +141 -0
  24. data/parscit/bin/parsHed/redo.parsHed.pl +198 -0
  25. data/parscit/bin/parsHed/tr2crfpp_parsHed.pl +521 -0
  26. data/parscit/bin/parseRefStrings.pl +102 -0
  27. data/parscit/bin/phOutput2xml.pl +223 -0
  28. data/parscit/bin/redo.parsCit.pl +105 -0
  29. data/parscit/bin/sectExtract.pl +149 -0
  30. data/parscit/bin/sectLabel/README +110 -0
  31. data/parscit/bin/sectLabel/README.txt +110 -0
  32. data/parscit/bin/sectLabel/genericSect/crossValidation.rb +98 -0
  33. data/parscit/bin/sectLabel/genericSect/extractFeature.rb +104 -0
  34. data/parscit/bin/sectLabel/genericSectExtract.rb +53 -0
  35. data/parscit/bin/sectLabel/getStructureInfo.pl +156 -0
  36. data/parscit/bin/sectLabel/processOmniXML.pl +1427 -0
  37. data/parscit/bin/sectLabel/processOmniXML_new.pl +1025 -0
  38. data/parscit/bin/sectLabel/processOmniXMLv2.pl +1529 -0
  39. data/parscit/bin/sectLabel/processOmniXMLv3.pl +964 -0
  40. data/parscit/bin/sectLabel/redo.sectLabel.pl +219 -0
  41. data/parscit/bin/sectLabel/simplifyOmniXML.pl +382 -0
  42. data/parscit/bin/sectLabel/single2multi.pl +190 -0
  43. data/parscit/bin/sectLabel/tr2crfpp.pl +158 -0
  44. data/parscit/bin/tr2crfpp.pl +260 -0
  45. data/parscit/bin/xml2train.pl +193 -0
  46. data/parscit/lib/CSXUtil/SafeText.pm +130 -0
  47. data/parscit/lib/Omni/Config.pm +93 -0
  48. data/parscit/lib/Omni/Omnicell.pm +263 -0
  49. data/parscit/lib/Omni/Omnicol.pm +292 -0
  50. data/parscit/lib/Omni/Omnidd.pm +328 -0
  51. data/parscit/lib/Omni/Omnidoc.pm +153 -0
  52. data/parscit/lib/Omni/Omniframe.pm +223 -0
  53. data/parscit/lib/Omni/Omniline.pm +423 -0
  54. data/parscit/lib/Omni/Omnipage.pm +282 -0
  55. data/parscit/lib/Omni/Omnipara.pm +232 -0
  56. data/parscit/lib/Omni/Omnirun.pm +303 -0
  57. data/parscit/lib/Omni/Omnitable.pm +336 -0
  58. data/parscit/lib/Omni/Omniword.pm +162 -0
  59. data/parscit/lib/Omni/Traversal.pm +313 -0
  60. data/parscit/lib/ParsCit/.PostProcess.pm.swp +0 -0
  61. data/parscit/lib/ParsCit/Citation.pm +737 -0
  62. data/parscit/lib/ParsCit/CitationContext.pm +220 -0
  63. data/parscit/lib/ParsCit/Config.pm +35 -0
  64. data/parscit/lib/ParsCit/Controller.pm +653 -0
  65. data/parscit/lib/ParsCit/PostProcess.pm +505 -0
  66. data/parscit/lib/ParsCit/PreProcess.pm +1041 -0
  67. data/parscit/lib/ParsCit/Tr2crfpp.pm +1195 -0
  68. data/parscit/lib/ParsHed/Config.pm +49 -0
  69. data/parscit/lib/ParsHed/Controller.pm +143 -0
  70. data/parscit/lib/ParsHed/PostProcess.pm +322 -0
  71. data/parscit/lib/ParsHed/Tr2crfpp.pm +448 -0
  72. data/{perl/ParsCit/lib/ParsCit/Tr2crfpp.pm → parscit/lib/ParsHed/Tr2crfpp_token.pm} +22 -21
  73. data/parscit/lib/SectLabel/AAMatching.pm +1949 -0
  74. data/parscit/lib/SectLabel/Config.pm +88 -0
  75. data/parscit/lib/SectLabel/Controller.pm +332 -0
  76. data/parscit/lib/SectLabel/PostProcess.pm +425 -0
  77. data/parscit/lib/SectLabel/PreProcess.pm +116 -0
  78. data/parscit/lib/SectLabel/Tr2crfpp.pm +1246 -0
  79. data/parscit/resources/parsCit.model +0 -0
  80. data/parscit/resources/parsCit.split.model +0 -0
  81. data/{perl/ParsCit → parscit}/resources/parsCitDict.txt +205 -0
  82. data/parscit/resources/parsHed/bigram +10 -0
  83. data/parscit/resources/parsHed/keywords +10 -0
  84. data/parscit/resources/parsHed/parsHed.model +0 -0
  85. data/parscit/resources/parsHed/parsHed.template +178 -0
  86. data/parscit/resources/sectLabel/affiliation.model +0 -0
  87. data/parscit/resources/sectLabel/author.model +0 -0
  88. data/parscit/resources/sectLabel/funcWord +320 -0
  89. data/parscit/resources/sectLabel/genericSect.model +0 -0
  90. data/parscit/resources/sectLabel/sectLabel.config +42 -0
  91. data/parscit/resources/sectLabel/sectLabel.configXml +42 -0
  92. data/parscit/resources/sectLabel/sectLabel.model +0 -0
  93. data/sh/convert_to_text.sh +20 -0
  94. data/spec/biblicit/extractor_spec.rb +121 -0
  95. data/spec/fixtures/Review_of_Michael_Tyes_Consciousness_Revisited.docx +0 -0
  96. data/spec/fixtures/critical-infrastructures.ps +63951 -0
  97. data/spec/fixtures/txt/E06-1050.txt +867 -0
  98. data/spec/fixtures/txt/sample1.txt +902 -0
  99. data/spec/fixtures/txt/sample2.txt +394 -0
  100. data/spec/spec_helper.rb +3 -0
  101. data/{perl → svm-header-parse}/HeaderParseService/lib/HeaderParse/API/Function.pm +2 -20
  102. data/{perl → svm-header-parse}/HeaderParseService/lib/HeaderParse/API/MultiClassChunking.pm +0 -7
  103. data/{perl → svm-header-parse}/HeaderParseService/lib/HeaderParse/API/Parser.pm +0 -2
  104. data/{perl → svm-header-parse}/HeaderParseService/lib/HeaderParse/API/ParserMethods.pm +0 -7
  105. data/{perl → svm-header-parse}/HeaderParseService/lib/HeaderParse/Config/API_Config.pm +6 -1
  106. data/svm-header-parse/HeaderParseService/tmp/.gitignore +4 -0
  107. data/svm-header-parse/extract.pl +75 -0
  108. metadata +351 -317
  109. data/perl/DocFilter/lib/DocFilter/Config.pm +0 -35
  110. data/perl/DocFilter/lib/DocFilter/Filter.pm +0 -51
  111. data/perl/FileConversionService/README.TXT +0 -11
  112. data/perl/FileConversionService/converters/PDFBox/pdfbox-app-1.7.1.jar +0 -0
  113. data/perl/FileConversionService/lib/CSXUtil/SafeText.pm +0 -140
  114. data/perl/FileConversionService/lib/FileConverter/CheckSum.pm +0 -77
  115. data/perl/FileConversionService/lib/FileConverter/Compression.pm +0 -137
  116. data/perl/FileConversionService/lib/FileConverter/Config.pm +0 -57
  117. data/perl/FileConversionService/lib/FileConverter/Controller.pm +0 -191
  118. data/perl/FileConversionService/lib/FileConverter/JODConverter.pm +0 -61
  119. data/perl/FileConversionService/lib/FileConverter/PDFBox.pm +0 -69
  120. data/perl/FileConversionService/lib/FileConverter/PSConverter.pm +0 -69
  121. data/perl/FileConversionService/lib/FileConverter/PSToText.pm +0 -88
  122. data/perl/FileConversionService/lib/FileConverter/Prescript.pm +0 -68
  123. data/perl/FileConversionService/lib/FileConverter/TET.pm +0 -75
  124. data/perl/FileConversionService/lib/FileConverter/Utils.pm +0 -130
  125. data/perl/HeaderParseService/lib/CSXUtil/SafeText.pm +0 -140
  126. data/perl/HeaderParseService/resources/data/EbizHeaders.txt +0 -24330
  127. data/perl/HeaderParseService/resources/data/EbizHeaders.txt.parsed +0 -27506
  128. data/perl/HeaderParseService/resources/data/EbizHeaders.txt.parsed.old +0 -26495
  129. data/perl/HeaderParseService/resources/data/tagged_headers.txt +0 -40668
  130. data/perl/HeaderParseService/resources/data/test_header.txt +0 -31
  131. data/perl/HeaderParseService/resources/data/test_header.txt.parsed +0 -31
  132. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test1 +0 -23
  133. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test10 +0 -23
  134. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test11 +0 -23
  135. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test12 +0 -23
  136. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test13 +0 -23
  137. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test14 +0 -23
  138. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test15 +0 -23
  139. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test2 +0 -23
  140. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test3 +0 -23
  141. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test4 +0 -23
  142. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test5 +0 -23
  143. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test6 +0 -23
  144. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test7 +0 -23
  145. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test8 +0 -23
  146. data/perl/HeaderParseService/tmp/tmpVec_1156237246.08016_test9 +0 -23
  147. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test1 +0 -23
  148. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test10 +0 -23
  149. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test11 +0 -23
  150. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test12 +0 -23
  151. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test13 +0 -23
  152. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test14 +0 -23
  153. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test15 +0 -23
  154. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test2 +0 -23
  155. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test3 +0 -23
  156. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test4 +0 -23
  157. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test5 +0 -23
  158. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test6 +0 -23
  159. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test7 +0 -23
  160. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test8 +0 -23
  161. data/perl/HeaderParseService/tmp/tmpVec_914027525.276114_test9 +0 -23
  162. data/perl/ParsCit/crfpp/traindata/parsCit.template +0 -60
  163. data/perl/ParsCit/crfpp/traindata/parsCit.train.data +0 -12104
  164. data/perl/ParsCit/crfpp/traindata/tagged_references.txt +0 -500
  165. data/perl/ParsCit/lib/CSXUtil/SafeText.pm +0 -140
  166. data/perl/ParsCit/lib/ParsCit/Citation.pm +0 -462
  167. data/perl/ParsCit/lib/ParsCit/CitationContext.pm +0 -132
  168. data/perl/ParsCit/lib/ParsCit/Config.pm +0 -46
  169. data/perl/ParsCit/lib/ParsCit/Controller.pm +0 -306
  170. data/perl/ParsCit/lib/ParsCit/PostProcess.pm +0 -367
  171. data/perl/ParsCit/lib/ParsCit/PreProcess.pm +0 -333
  172. data/perl/ParsCit/resources/parsCit.model +0 -0
  173. data/perl/extract.pl +0 -199
  174. data/spec/biblicit/cb2bib_spec.rb +0 -48
  175. data/spec/biblicit/citeseer_spec.rb +0 -40
  176. /data/{perl → svm-header-parse}/HeaderParseService/README.TXT +0 -0
  177. /data/{perl/DocFilter → svm-header-parse/HeaderParseService}/lib/CSXUtil/SafeText.pm +0 -0
  178. /data/{perl → svm-header-parse}/HeaderParseService/lib/HeaderParse/API/AssembleXMLMetadata.pm +0 -0
  179. /data/{perl → svm-header-parse}/HeaderParseService/lib/HeaderParse/API/LoadInformation.pm +0 -0
  180. /data/{perl → svm-header-parse}/HeaderParseService/lib/HeaderParse/API/NamePatternMatch.pm +0 -0
  181. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/50states +0 -0
  182. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/AddrTopWords.txt +0 -0
  183. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/AffiTopWords.txt +0 -0
  184. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/AffiTopWordsAll.txt +0 -0
  185. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/ChineseSurNames.txt +0 -0
  186. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/Csurnames.bin +0 -0
  187. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/Csurnames_spec.bin +0 -0
  188. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/DomainSuffixes.txt +0 -0
  189. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/LabeledHeader +0 -0
  190. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/README +0 -0
  191. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/TrainMulClassLines +0 -0
  192. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/TrainMulClassLines1 +0 -0
  193. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/abstract.txt +0 -0
  194. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/abstractTopWords +0 -0
  195. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/addr.txt +0 -0
  196. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/affi.txt +0 -0
  197. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/affis.bin +0 -0
  198. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/all_namewords_spec.bin +0 -0
  199. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/allnamewords.bin +0 -0
  200. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/cities_US.txt +0 -0
  201. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/cities_world.txt +0 -0
  202. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/city.txt +0 -0
  203. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/cityname.txt +0 -0
  204. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/country_abbr.txt +0 -0
  205. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/countryname.txt +0 -0
  206. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/dateTopWords +0 -0
  207. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/degree.txt +0 -0
  208. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/email.txt +0 -0
  209. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/excludeWords.txt +0 -0
  210. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/female-names +0 -0
  211. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/firstNames.txt +0 -0
  212. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/firstnames.bin +0 -0
  213. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/firstnames_spec.bin +0 -0
  214. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/intro.txt +0 -0
  215. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/keyword.txt +0 -0
  216. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/keywordTopWords +0 -0
  217. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/male-names +0 -0
  218. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/middleNames.txt +0 -0
  219. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/month.txt +0 -0
  220. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/mul +0 -0
  221. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/mul.label +0 -0
  222. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/mul.label.old +0 -0
  223. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/mul.processed +0 -0
  224. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/mulAuthor +0 -0
  225. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/mulClassStat +0 -0
  226. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/nickname.txt +0 -0
  227. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/nicknames.bin +0 -0
  228. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/note.txt +0 -0
  229. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/page.txt +0 -0
  230. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/phone.txt +0 -0
  231. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/postcode.txt +0 -0
  232. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/pubnum.txt +0 -0
  233. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/statename.bin +0 -0
  234. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/statename.txt +0 -0
  235. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/states_and_abbreviations.txt +0 -0
  236. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/stopwords +0 -0
  237. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/stopwords.bin +0 -0
  238. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/surNames.txt +0 -0
  239. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/surnames.bin +0 -0
  240. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/surnames_spec.bin +0 -0
  241. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/A.html +0 -0
  242. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/B.html +0 -0
  243. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/C.html +0 -0
  244. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/D.html +0 -0
  245. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/E.html +0 -0
  246. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/F.html +0 -0
  247. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/G.html +0 -0
  248. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/H.html +0 -0
  249. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/I.html +0 -0
  250. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/J.html +0 -0
  251. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/K.html +0 -0
  252. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/L.html +0 -0
  253. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/M.html +0 -0
  254. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/N.html +0 -0
  255. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/O.html +0 -0
  256. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/P.html +0 -0
  257. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/Q.html +0 -0
  258. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/R.html +0 -0
  259. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/S.html +0 -0
  260. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/T.html +0 -0
  261. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/U.html +0 -0
  262. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/V.html +0 -0
  263. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/W.html +0 -0
  264. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/WCSelect.gif +0 -0
  265. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/X.html +0 -0
  266. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/Y.html +0 -0
  267. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/Z.html +0 -0
  268. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ae.html +0 -0
  269. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/am.html +0 -0
  270. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ar.html +0 -0
  271. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/at.html +0 -0
  272. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/au.html +0 -0
  273. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/bd.html +0 -0
  274. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/be.html +0 -0
  275. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/bg.html +0 -0
  276. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/bh.html +0 -0
  277. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/blueribbon.gif +0 -0
  278. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/bm.html +0 -0
  279. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/bn.html +0 -0
  280. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/br.html +0 -0
  281. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ca.html +0 -0
  282. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ch.html +0 -0
  283. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/cl.html +0 -0
  284. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/cn.html +0 -0
  285. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/co.html +0 -0
  286. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/cr.html +0 -0
  287. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/cy.html +0 -0
  288. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/cz.html +0 -0
  289. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/de.html +0 -0
  290. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/dean-mainlink.jpg +0 -0
  291. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/dk.html +0 -0
  292. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ec.html +0 -0
  293. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ee.html +0 -0
  294. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/eg.html +0 -0
  295. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/es.html +0 -0
  296. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/et.html +0 -0
  297. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/faq.html +0 -0
  298. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/fi.html +0 -0
  299. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/fj.html +0 -0
  300. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/fo.html +0 -0
  301. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/fr.html +0 -0
  302. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/geog.html +0 -0
  303. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/gr.html +0 -0
  304. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/gu.html +0 -0
  305. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/hk.html +0 -0
  306. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/hr.html +0 -0
  307. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/hu.html +0 -0
  308. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/id.html +0 -0
  309. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ie.html +0 -0
  310. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/il.html +0 -0
  311. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/in.html +0 -0
  312. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/is.html +0 -0
  313. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/it.html +0 -0
  314. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/jm.html +0 -0
  315. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/jo.html +0 -0
  316. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/jp.html +0 -0
  317. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/kaplan.gif +0 -0
  318. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/kr.html +0 -0
  319. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/kw.html +0 -0
  320. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/lb.html +0 -0
  321. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/linkbw2.gif +0 -0
  322. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/lk.html +0 -0
  323. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/lt.html +0 -0
  324. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/lu.html +0 -0
  325. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/lv.html +0 -0
  326. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ma.html +0 -0
  327. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/maczynski.gif +0 -0
  328. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/mirror.tar +0 -0
  329. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/mk.html +0 -0
  330. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/mo.html +0 -0
  331. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/mseawdm.gif +0 -0
  332. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/mt.html +0 -0
  333. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/mx.html +0 -0
  334. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/my.html +0 -0
  335. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ni.html +0 -0
  336. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/nl.html +0 -0
  337. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/no.html +0 -0
  338. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/nz.html +0 -0
  339. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/pa.html +0 -0
  340. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/pe.html +0 -0
  341. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ph.html +0 -0
  342. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/pl.html +0 -0
  343. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/pointcom.gif +0 -0
  344. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/pr.html +0 -0
  345. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ps.html +0 -0
  346. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/pt.html +0 -0
  347. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/recognition.html +0 -0
  348. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/results.html +0 -0
  349. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ro.html +0 -0
  350. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ru.html +0 -0
  351. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/sd.html +0 -0
  352. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/se.html +0 -0
  353. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/sg.html +0 -0
  354. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/si.html +0 -0
  355. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/sk.html +0 -0
  356. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/th.html +0 -0
  357. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/tr.html +0 -0
  358. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/tw.html +0 -0
  359. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ua.html +0 -0
  360. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/uk.html +0 -0
  361. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/univ-full.html +0 -0
  362. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/univ.html +0 -0
  363. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/uy.html +0 -0
  364. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/ve.html +0 -0
  365. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/yu.html +0 -0
  366. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/za.html +0 -0
  367. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list/zm.html +0 -0
  368. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/university_list.txt +0 -0
  369. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/url.txt +0 -0
  370. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/webTopWords +0 -0
  371. /data/{perl → svm-header-parse}/HeaderParseService/resources/database/words +0 -0
  372. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/10ContextModelfold1 +0 -0
  373. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/10Modelfold1 +0 -0
  374. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/11ContextModelfold1 +0 -0
  375. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/11Modelfold1 +0 -0
  376. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/12ContextModelfold1 +0 -0
  377. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/12Modelfold1 +0 -0
  378. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/13ContextModelfold1 +0 -0
  379. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/13Modelfold1 +0 -0
  380. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/14ContextModelfold1 +0 -0
  381. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/14Modelfold1 +0 -0
  382. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/15ContextModelfold1 +0 -0
  383. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/15Modelfold1 +0 -0
  384. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/1ContextModelfold1 +0 -0
  385. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/1Modelfold1 +0 -0
  386. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/2ContextModelfold1 +0 -0
  387. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/2Modelfold1 +0 -0
  388. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/3ContextModelfold1 +0 -0
  389. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/3Modelfold1 +0 -0
  390. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/4ContextModelfold1 +0 -0
  391. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/4Modelfold1 +0 -0
  392. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/5ContextModelfold1 +0 -0
  393. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/5Modelfold1 +0 -0
  394. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/6ContextModelfold1 +0 -0
  395. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/6Modelfold1 +0 -0
  396. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/7ContextModelfold1 +0 -0
  397. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/7Modelfold1 +0 -0
  398. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/8ContextModelfold1 +0 -0
  399. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/8Modelfold1 +0 -0
  400. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/9ContextModelfold1 +0 -0
  401. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/9Modelfold1 +0 -0
  402. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/NameSpaceModel +0 -0
  403. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/NameSpaceTrainF +0 -0
  404. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/WrapperBaseFeaDict +0 -0
  405. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/WrapperContextFeaDict +0 -0
  406. /data/{perl → svm-header-parse}/HeaderParseService/resources/models/WrapperSpaceAuthorFeaDict +0 -0
@@ -0,0 +1,292 @@
1
+ package Omni::Omnicol;
2
+
3
+ # Configuration
4
+ use strict;
5
+
6
+ # Local libraries
7
+ use Omni::Config;
8
+ use Omni::Omnidd;
9
+ use Omni::Omnipara;
10
+ use Omni::Omniframe;
11
+ use Omni::Omnitable;
12
+
13
+ # Extern libraries
14
+ use XML::Twig;
15
+ use XML::Parser;
16
+
17
+ # Global variables
18
+ my $tag_list = $Omni::Config::tag_list;
19
+ my $att_list = $Omni::Config::att_list;
20
+ my $obj_list = $Omni::Config::obj_list;
21
+
22
+ # Temporary variables
23
+
24
+ ###
25
+ # A column object in Omnipage xml: a column contains zero or many paragraphs
26
+ #
27
+ # Do Hoang Nhat Huy, 11 Jan 2011
28
+ ###
29
+ # Initialization
30
+ sub new
31
+ {
32
+ my ($class) = @_;
33
+
34
+ # Column: a column can have many paragraphs, dd, tables, or pictures
35
+ my @objs = ();
36
+
37
+ # Class members
38
+ my $self = { '_self' => $obj_list->{ 'OMNICOL' },
39
+ '_raw' => undef,
40
+ '_content' => undef,
41
+ '_bottom' => undef,
42
+ '_top' => undef,
43
+ '_left' => undef,
44
+ '_right' => undef,
45
+ '_objs' => \@objs };
46
+
47
+ bless $self, $class;
48
+ return $self;
49
+ }
50
+
51
+ #
52
+ sub set_raw
53
+ {
54
+ my ($self, $raw) = @_;
55
+
56
+ # Save the raw xml <column> ... </column>
57
+ $self->{ '_raw' } = $raw;
58
+
59
+ # Parse the raw string
60
+ my $twig_roots = { $tag_list->{ 'COLUMN' } => 1 };
61
+ my $twig_handlers = { $tag_list->{ 'COLUMN' } => sub { parse(@_, \$self); } };
62
+
63
+ # XML::Twig
64
+ my $twig = new XML::Twig( twig_roots => $twig_roots,
65
+ twig_handlers => $twig_handlers,
66
+ pretty_print => 'indented' );
67
+
68
+ # Start the XML parsing
69
+ $twig->parse($raw, \$self);
70
+ $twig->purge;
71
+ }
72
+
73
+ sub get_raw
74
+ {
75
+ my ($self) = @_;
76
+ return $self->{ '_raw' };
77
+ }
78
+
79
+ sub parse
80
+ {
81
+ my ($twig, $node, $self) = @_;
82
+
83
+ # At first, content is blank
84
+ my $tmp_content = "";
85
+ # because there's no object
86
+ my @tmp_objs = ();
87
+
88
+ # Get <column> node attributes
89
+ my $tmp_bottom = GetNodeAttr($node, $att_list->{ 'BOTTOM' });
90
+ my $tmp_top = GetNodeAttr($node, $att_list->{ 'TOP' });
91
+ my $tmp_left = GetNodeAttr($node, $att_list->{ 'LEFT' });
92
+ my $tmp_right = GetNodeAttr($node, $att_list->{ 'RIGHT' });
93
+
94
+ # Check if there's any paragraph, dd, table, or picture
95
+ # The large number of possible children is due to the
96
+ # ambiguous structure of the Omnipage XML
97
+ my $dd_tag = $tag_list->{ 'DD' };
98
+ my $img_tag = $tag_list->{ 'PICTURE' };
99
+ my $para_tag = $tag_list->{ 'PARA' };
100
+ my $table_tag = $tag_list->{ 'TABLE' };
101
+ my $column_tag = $tag_list->{ 'COLUMN' };
102
+ my $frame_tag = $tag_list->{ 'FRAME' };
103
+
104
+ my $child = undef;
105
+ # Get the first child in the body text
106
+ $child = $node->first_child();
107
+
108
+ while (defined $child)
109
+ {
110
+ my $xpath = $child->path();
111
+
112
+ # if this child is a <para> tag
113
+ if ($xpath =~ m/\/$para_tag$/)
114
+ {
115
+ my $para = new Omni::Omnipara();
116
+
117
+ # Set raw content
118
+ $para->set_raw($child->sprint());
119
+
120
+ # Update paragraph list
121
+ push @tmp_objs, $para;
122
+
123
+ # Update content
124
+ $tmp_content = $tmp_content . $para->get_content() . "\n";
125
+ }
126
+ # TODO: I'll handle this one later. Seriously
127
+ # if this child is a <dd> tag
128
+ elsif ($xpath =~ m/\/$dd_tag$/)
129
+ {
130
+ #my $dd = new Omni::Omnidd();
131
+
132
+ # Set raw content
133
+ #$dd->set_raw($child->sprint());
134
+
135
+ # Update paragraph list
136
+ #push @tmp_objs, $dd;
137
+
138
+ # Update content
139
+ #$tmp_content = $tmp_content . $dd->get_content() . "\n";
140
+ }
141
+ # if this child is a <table> tag
142
+ elsif ($xpath =~ m/\/$table_tag$/)
143
+ {
144
+ my $table = new Omni::Omnitable();
145
+
146
+ # Set raw content
147
+ $table->set_raw($child->sprint());
148
+
149
+ # Update paragraph list
150
+ push @tmp_objs, $table;
151
+
152
+ # Update content
153
+ $tmp_content = $tmp_content . $table->get_content() . "\n";
154
+ }
155
+ # if this child is a <picture> tag
156
+ elsif ($xpath =~ m/\/$img_tag$/)
157
+ {
158
+ #my $img = new Omni::Omniimg();
159
+
160
+ # Set raw content
161
+ #$img->set_raw($child->sprint());
162
+
163
+ # Update paragraph list
164
+ #push @tmp_objs, $img;
165
+
166
+ # Update content
167
+ #$tmp_content = $tmp_content . $img->get_content() . "\n";
168
+ }
169
+ # if this child is a <column> tag
170
+ elsif ($xpath =~ m/\/$column_tag$/)
171
+ {
172
+ my $col = new Omni::Omnicol();
173
+
174
+ # Set raw content
175
+ $col->set_raw($child->sprint());
176
+
177
+ # Nested <column> is not allowed so we copy the objects
178
+ my $objects = $col->get_objs_ref();
179
+
180
+ # Update <column> objects list
181
+ push @tmp_objs, @{ $objects };
182
+
183
+ # Update content
184
+ $tmp_content = $tmp_content . $col->get_content() . "\n";
185
+ }
186
+ # if this child is <frame>
187
+ elsif ($xpath =~ m/\/$frame_tag$/)
188
+ {
189
+ my $frame = new Omni::Omniframe();
190
+
191
+ # Set raw content
192
+ $frame->set_raw($child->sprint());
193
+
194
+ # Update column list
195
+ push @tmp_objs, $frame;
196
+
197
+ # Update content
198
+ $tmp_content = $tmp_content . $frame->get_content() . "\n";
199
+ }
200
+
201
+ # Little brother
202
+ if ($child->is_last_child)
203
+ {
204
+ last;
205
+ }
206
+ else
207
+ {
208
+ $child = $child->next_sibling();
209
+ }
210
+ }
211
+
212
+ # Copy information from temporary variables to class members
213
+ $$self->{ '_bottom' } = $tmp_bottom;
214
+ $$self->{ '_top' } = $tmp_top;
215
+ $$self->{ '_left' } = $tmp_left;
216
+ $$self->{ '_right' } = $tmp_right;
217
+
218
+ # Copy all objects
219
+ @{$$self->{ '_objs' } } = @tmp_objs;
220
+
221
+ # Copy content
222
+ $$self->{ '_content' } = $tmp_content;
223
+ }
224
+
225
+ sub get_name
226
+ {
227
+ my ($self) = @_;
228
+ return $self->{ '_self' };
229
+ }
230
+
231
+ sub get_objs_ref
232
+ {
233
+ my ($self) = @_;
234
+ return $self->{ '_objs' };
235
+ }
236
+
237
+ sub get_content
238
+ {
239
+ my ($self) = @_;
240
+ return $self->{ '_content' };
241
+ }
242
+
243
+ sub get_bottom_pos
244
+ {
245
+ my ($self) = @_;
246
+ return $self->{ '_bottom' };
247
+ }
248
+
249
+ sub get_top_pos
250
+ {
251
+ my ($self) = @_;
252
+ return $self->{ '_top' };
253
+ }
254
+
255
+ sub get_left_pos
256
+ {
257
+ my ($self) = @_;
258
+ return $self->{ '_left' };
259
+ }
260
+
261
+ sub get_right_pos
262
+ {
263
+ my ($self) = @_;
264
+ return $self->{ '_right' };
265
+ }
266
+
267
+ # Support functions
268
+ sub GetNodeAttr
269
+ {
270
+ my ($node, $attr) = @_;
271
+ return ($node->att($attr) ? $node->att($attr) : "");
272
+ }
273
+
274
+ sub SetNodeAttr
275
+ {
276
+ my ($node, $attr, $value) = @_;
277
+ $node->set_att($attr, $value);
278
+ }
279
+
280
+ sub GetNodeText
281
+ {
282
+ my ($node) = @_;
283
+ return $node->text;
284
+ }
285
+
286
+ sub SetNodeText
287
+ {
288
+ my ($node, $value) = @_;
289
+ $node->set_text($value);
290
+ }
291
+
292
+ 1;
@@ -0,0 +1,328 @@
1
+ package Omni::Omnidd;
2
+
3
+ # Configuration
4
+ use strict;
5
+
6
+ # Local libraries
7
+ use Omni::Config;
8
+ use Omni::Omnicol;
9
+ use Omni::Omnipara;
10
+ use Omni::Omniframe;
11
+ use Omni::Omnitable;
12
+
13
+ # Extern libraries
14
+ use XML::Twig;
15
+ use XML::Parser;
16
+
17
+ # Global variables
18
+ my $tag_list = $Omni::Config::tag_list;
19
+ my $att_list = $Omni::Config::att_list;
20
+ my $obj_list = $Omni::Config::obj_list;
21
+
22
+ # Temporary variables
23
+
24
+ ###
25
+ # A dd object in Omnipage xml: a dd, don't know what it is, but its structure
26
+ # is quite similar to a column
27
+ #
28
+ # Do Hoang Nhat Huy, 11 Jan 2011
29
+ ###
30
+ # Initialization
31
+ sub new
32
+ {
33
+ my ($class) = @_;
34
+
35
+ # dd: a dd can have many tables, or pictures, may be paras, and columns
36
+ my @objs = ();
37
+
38
+ # Class members
39
+ my $self = { '_self' => $obj_list->{ 'OMNIDD' },
40
+ '_raw' => undef,
41
+ '_content' => undef,
42
+ '_bottom' => undef,
43
+ '_top' => undef,
44
+ '_left' => undef,
45
+ '_right' => undef,
46
+ '_bottom_dist' => undef,
47
+ '_top_dist' => undef,
48
+ '_left_dist' => undef,
49
+ '_right_dist' => undef,
50
+ '_objs' => \@objs };
51
+
52
+ bless $self, $class;
53
+ return $self;
54
+ }
55
+
56
+ #
57
+ sub set_raw
58
+ {
59
+ my ($self, $raw) = @_;
60
+
61
+ # Save the raw xml <column> ... </column>
62
+ $self->{ '_raw' } = $raw;
63
+
64
+ # Parse the raw string
65
+ my $twig_roots = { $tag_list->{ 'DD' } => 1 };
66
+ my $twig_handlers = { $tag_list->{ 'DD' } => sub { parse(@_, \$self); } };
67
+
68
+ # XML::Twig
69
+ my $twig = new XML::Twig( twig_roots => $twig_roots,
70
+ twig_handlers => $twig_handlers,
71
+ pretty_print => 'indented' );
72
+
73
+ # Start the XML parsing
74
+ $twig->parse($raw, \$self);
75
+ $twig->purge;
76
+ }
77
+
78
+ sub get_raw
79
+ {
80
+ my ($self) = @_;
81
+ return $self->{ '_raw' };
82
+ }
83
+
84
+ sub parse
85
+ {
86
+ my ($twig, $node, $self) = @_;
87
+
88
+ # At first, content is blank
89
+ my $tmp_content = "";
90
+ # because there's no object
91
+ my @tmp_objs = ();
92
+
93
+ # Get <dd> node attributes
94
+ my $tmp_bottom = GetNodeAttr($node, $att_list->{ 'BOTTOM' });
95
+ my $tmp_top = GetNodeAttr($node, $att_list->{ 'TOP' });
96
+ my $tmp_left = GetNodeAttr($node, $att_list->{ 'LEFT' });
97
+ my $tmp_right = GetNodeAttr($node, $att_list->{ 'RIGHT' });
98
+ my $tmp_bottom_dist = GetNodeAttr($node, $att_list->{ 'BOTTOMDIST' });
99
+ my $tmp_top_dist = GetNodeAttr($node, $att_list->{ 'TOPDIST' });
100
+ my $tmp_left_dist = GetNodeAttr($node, $att_list->{ 'LEFTDIST' });
101
+ my $tmp_right_dist = GetNodeAttr($node, $att_list->{ 'RIGHTDIST' });
102
+
103
+ # Check if there's any paragraph, col, table, or picture
104
+ # The large number of possible children is due to the
105
+ # ambiguous structure of the Omnipage XML
106
+ my $dd_tag = $tag_list->{ 'DD' };
107
+ my $img_tag = $tag_list->{ 'PICTURE' };
108
+ my $para_tag = $tag_list->{ 'PARA' };
109
+ my $table_tag = $tag_list->{ 'TABLE' };
110
+ my $column_tag = $tag_list->{ 'COLUMN' };
111
+ my $frame_tag = $tag_list->{ 'FRAME' };
112
+
113
+ my $child = undef;
114
+ # Get the first child in the body text
115
+ $child = $node->first_child();
116
+
117
+ while (defined $child)
118
+ {
119
+ my $xpath = $child->path();
120
+
121
+ # if this child is a <para> tag
122
+ if ($xpath =~ m/\/$para_tag$/)
123
+ {
124
+ my $para = new Omni::Omnipara();
125
+
126
+ # Set raw content
127
+ $para->set_raw($child->sprint());
128
+
129
+ # Update paragraph list
130
+ push @tmp_objs, $para;
131
+
132
+ # Update content
133
+ $tmp_content = $tmp_content . $para->get_content() . "\n";
134
+ }
135
+ # if this child is a <dd> tag
136
+ elsif ($xpath =~ m/\/$dd_tag$/)
137
+ {
138
+ my $dd = new Omni::Omnidd();
139
+
140
+ # Set raw content
141
+ $dd->set_raw($child->sprint());
142
+
143
+ # Nested <dd> is not allowed so we copy the objects
144
+ my $objects = $dd->get_objs_ref();
145
+
146
+ # Update <dd> objects list
147
+ push @tmp_objs, @{ $objects };
148
+
149
+ # Update content
150
+ $tmp_content = $tmp_content . $dd->get_content() . "\n";
151
+ }
152
+ # if this child is a <table> tag
153
+ elsif ($xpath =~ m/\/$table_tag$/)
154
+ {
155
+ my $table = new Omni::Omnitable();
156
+
157
+ # Set raw content
158
+ $table->set_raw($child->sprint());
159
+
160
+ # Update paragraph list
161
+ push @tmp_objs, $table;
162
+
163
+ # Update content
164
+ $tmp_content = $tmp_content . $table->get_content() . "\n";
165
+ }
166
+ # if this child is a <picture> tag
167
+ elsif ($xpath =~ m/\/$img_tag$/)
168
+ {
169
+ #my $img = new Omni::Omniimg();
170
+
171
+ # Set raw content
172
+ #$img->set_raw($child->sprint());
173
+
174
+ # Update paragraph list
175
+ #push @tmp_objs, $img;
176
+
177
+ # Update content
178
+ #$tmp_content = $tmp_content . $img->get_content() . "\n";
179
+ }
180
+ # if this child is a <column> tag
181
+ elsif ($xpath =~ m/\/$column_tag$/)
182
+ {
183
+ #my $col = new Omni::Omnicol();
184
+
185
+ # Set raw content
186
+ #$col->set_raw($child->sprint());
187
+
188
+ # Update paragraph list
189
+ #push @tmp_objs, $col;
190
+
191
+ # Update content
192
+ #$tmp_content = $tmp_content . $col->get_content() . "\n";
193
+ }
194
+ # if this child is <frame>
195
+ elsif ($xpath =~ m/\/$frame_tag$/)
196
+ {
197
+ my $frame = new Omni::Omniframe();
198
+
199
+ # Set raw content
200
+ $frame->set_raw($child->sprint());
201
+
202
+ # Update column list
203
+ push @tmp_objs, $frame;
204
+
205
+ # Update content
206
+ $tmp_content = $tmp_content . $frame->get_content() . "\n";
207
+ }
208
+
209
+ # Little brother
210
+ if ($child->is_last_child)
211
+ {
212
+ last;
213
+ }
214
+ else
215
+ {
216
+ $child = $child->next_sibling();
217
+ }
218
+ }
219
+
220
+ # Copy information from temporary variables to class members
221
+ $$self->{ '_bottom' } = $tmp_bottom;
222
+ $$self->{ '_top' } = $tmp_top;
223
+ $$self->{ '_left' } = $tmp_left;
224
+ $$self->{ '_right' } = $tmp_right;
225
+ $$self->{ '_bottom_dist' } = $tmp_bottom_dist;
226
+ $$self->{ '_top_dist' } = $tmp_top_dist;
227
+ $$self->{ '_left_dist' } = $tmp_left_dist;
228
+ $$self->{ '_right_dist' } = $tmp_right_dist;
229
+
230
+ # Copy content
231
+ $$self->{ '_content' } = $tmp_content;
232
+
233
+ # Copy all objects
234
+ @{$$self->{ '_objs' } } = @tmp_objs;
235
+ }
236
+
237
+ sub get_name
238
+ {
239
+ my ($self) = @_;
240
+ return $self->{ '_self' };
241
+ }
242
+
243
+ sub get_objs_ref
244
+ {
245
+ my ($self) = @_;
246
+ return $self->{ '_objs' };
247
+ }
248
+
249
+ sub get_content
250
+ {
251
+ my ($self) = @_;
252
+ return $self->{ '_content' };
253
+ }
254
+
255
+ sub get_bottom_pos
256
+ {
257
+ my ($self) = @_;
258
+ return $self->{ '_bottom' };
259
+ }
260
+
261
+ sub get_top_pos
262
+ {
263
+ my ($self) = @_;
264
+ return $self->{ '_top' };
265
+ }
266
+
267
+ sub get_left_pos
268
+ {
269
+ my ($self) = @_;
270
+ return $self->{ '_left' };
271
+ }
272
+
273
+ sub get_right_pos
274
+ {
275
+ my ($self) = @_;
276
+ return $self->{ '_right' };
277
+ }
278
+
279
+ sub get_bottom_distance
280
+ {
281
+ my ($self) = @_;
282
+ return $self->{ '_bottom_dist' };
283
+ }
284
+
285
+ sub get_top_distance
286
+ {
287
+ my ($self) = @_;
288
+ return $self->{ '_top_dist' };
289
+ }
290
+
291
+ sub get_left_distance
292
+ {
293
+ my ($self) = @_;
294
+ return $self->{ '_left_dist' };
295
+ }
296
+
297
+ sub get_right_distance
298
+ {
299
+ my ($self) = @_;
300
+ return $self->{ '_right_dist' };
301
+ }
302
+
303
+ # Support functions
304
+ sub GetNodeAttr
305
+ {
306
+ my ($node, $attr) = @_;
307
+ return ($node->att($attr) ? $node->att($attr) : "");
308
+ }
309
+
310
+ sub SetNodeAttr
311
+ {
312
+ my ($node, $attr, $value) = @_;
313
+ $node->set_att($attr, $value);
314
+ }
315
+
316
+ sub GetNodeText
317
+ {
318
+ my ($node) = @_;
319
+ return $node->text;
320
+ }
321
+
322
+ sub SetNodeText
323
+ {
324
+ my ($node, $value) = @_;
325
+ $node->set_text($value);
326
+ }
327
+
328
+ 1;