gGfddlZddlmZddlmZddlmZGddeZdZe dk(reyy) N) defaultdict)reduce) CorpusReadercveZdZdZej dZedZd fd Z d dZ d dZ d dZ dZ xZS) LinThesaurusCorpusReaderzEWrapper for the LISP-formatted thesauruses distributed by Dekang Lin.z \("?([^"]+)"? \(desc [0-9.]+\).+c ttS)z6Factory for creating defaultdict of defaultdict(dict)s)rdictK/var/www/openai/venv/lib/python3.12/site-packages/nltk/corpus/reader/lin.py__defaultdict_factoryz.LinThesaurusCorpusReader.__defaultdict_factorys4  r ct ||dttj|_||_|jddD]\}}}t|5}d}|D]}|j}|r#tjjd|} d}8|dk(rd}@|jd} t| dk(s`| \} } t| |j | | jd < d d d y #1swYxYw) a Initialize the thesaurus. :param root: root directory containing thesaurus LISP files :type root: C{string} :param badscore: the score to give to words which do not appear in each other's sets of synonyms :type badscore: C{float} z sim[A-Z]\.lspT)include_encodinginclude_fileidz\1Fz)) "N)super__init__rr._LinThesaurusCorpusReader__defaultdict_factory _thesaurus _badscoreabspathsopenstrip_key_resubsplitlenfloat)selfrootbadscorepathencodingfileidlin_filefirstlinekey split_linengramscore __class__s r rz!LinThesaurusCorpusReader.__init__s /0%&>&T&TU!&*mm!$'4' "D(Fdx$D::>BB5$O % $&*ZZ%5 z?a/+5LE5MR %NDOOF3C8S9IJ%' sA$D4DD cj||k(r|ry|jDcgc]}|dfc}S|r5||j||vr|j|||S|jS|jDcgc]9}|||j||vr|j|||n |jf;c}Scc}wcc}w)a Returns the similarity score for two ngrams. :param ngram1: first ngram to compare :type ngram1: C{string} :param ngram2: second ngram to compare :type ngram2: C{string} :param fileid: thesaurus fileid to search in. If None, search all fileids. :type fileid: C{string} :return: If fileid is specified, just the score for the two ngrams; otherwise, list of tuples of fileids and scores. g?)_fileidsrr)r!ngram1ngram2r&fids r similarityz#LinThesaurusCorpusReader.similarity?s V .2mmB0c|r |j||jS|jDcgc]$}||j||jf&c}Scc}w)a Returns a list of scored synonyms (tuples of synonyms and scores) for the current ngram :param ngram: ngram to lookup :type ngram: C{string} :param fileid: thesaurus fileid to search in. If None, search all fileids. :type fileid: C{string} :return: If fileid is specified, list of tuples of scores and synonyms; otherwise, list of tuples of fileids and lists, where inner lists consist of tuples of scores and synonyms. )ritemsr0r!r,r&s r scored_synonymsz(LinThesaurusCorpusReader.scored_synonymsfsk ??6*51779 9#mm+F07==?@+ )Ac|r |j||jS|jDcgc]$}||j||jf&c}Scc}w)a Returns a list of synonyms for the current ngram. :param ngram: ngram to lookup :type ngram: C{string} :param fileid: thesaurus fileid to search in. If None, search all fileids. :type fileid: C{string} :return: If fileid is specified, list of synonyms; otherwise, list of tuples of fileids and lists, where inner lists contain synonyms. )rkeysr0r7s r synonymsz!LinThesaurusCorpusReader.synonymszsk ??6*51668 8#mm+F07<<>?+ r9c<tfdjdS)z Determines whether or not the given ngram is in the thesaurus. :param ngram: ngram to lookup :type ngram: C{string} :return: whether the given ngram is in the thesaurus. c.|xsj|vSN)r)accumr&r,r!s r z7LinThesaurusCorpusReader.__contains__..s%"MET__V5L,L"Mr F)rr0)r!r,s``r __contains__z%LinThesaurusCorpusReader.__contains__s M MM   r )gr?)__name__ __module__ __qualname____doc__recompiler staticmethodrrr4r8r<rB __classcell__)r.s@r rrsFObjj<=G!!"H%N(&  r rcddlm}d}d}td|zt|j|td|zt|j |td|zt|j|d td|zt|j|d td |d |d t|j ||y) Nr) lin_thesaurusbusiness enterprisezGetting synonyms for zGetting scored synonyms for z5Getting synonyms from simN.lsp (noun subsection) for zsimN.lsp)r&zSimilarity score for z and :) nltk.corpusrLprintr<r8r4)thesword1word2s r demorUs1 E E !E )* $--  (5 01 $  u %& AE IJ $--j- 12 AE IJ $--j- 12 !%eWA 67 $//% '(r __main__) rG collectionsr functoolsrnltk.corpus.readerrrrUrCr r r rZs; #+J |J d)* zFr