³ò B_Kc @sÔddkZddkZddklZddklZddkTddkTdefd„ƒYZ de fd„ƒYZ d e fd „ƒYZ d e fd „ƒYZ d e fd„ƒYZde fd„ƒYZdS(iÿÿÿÿN(tTree(t ElementTree(t*tNombankCorpusReadercBs_eZdZdeeeed„Zed„Zd„Zd„Zd„Zd„Z d„Z RS( s– Corpus reader for the nombank corpus, which augments the Penn Treebank with information about the predicate argument structure of every noun instance. The corpus consists of two parts: the predicate-argument annotations themselves, and a set of X{frameset files} which define the argument labels used by the annotations, on a per-noun basis. Each X{frameset file} contains one or more predicates, such as C{'turn'} or C{'turn_on'}, each of which is divided into coarse-grained word senses called X{rolesets}. For each X{roleset}, the frameset file provides descriptions of the argument roles, along with examples. tcCs€t|tƒot||ƒ}nt|ƒ}ti||||g||ƒ||_||_||_||_ ||_ dS(sÙ @param root: The root directory for this corpus. @param nomfile: The name of the file containing the predicate- argument annotations (relative to C{root}). @param framefiles: A list or regexp specifying the frameset fileids for this corpus. @param parse_fileid_xform: A transform that should be applied to the fileids in this corpus. This should be a function of one argument (a fileid) that returns a string (the new fileid). @param parse_corpus: The corpus containing the parse trees corresponding to this corpus. These parse trees are necessary to resolve the tree pointers used by nombank. N( t isinstancet basestringtfind_corpus_fileidstlistt CorpusReadert__init__t_nomfilet _framefilest _nounsfilet_parse_fileid_xformt _parse_corpus(tselftroottnomfilet framefilest nounsfiletparse_fileid_xformt parse_corpustencoding((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR s     cCsk|tjo |i}nt|tƒo |g}ntg}|D]}||i|ƒiƒqE~ƒS(sV @return: the text contents of the given fileids, as a single string. (tNonet_fileidsRRtconcattopentread(Rtfileidst_[1]tf((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pytraw?scCs.t|i|iƒ|id|i|iƒƒS(s„ @return: a corpus view that acts as a list of L{NombankInstance} objects, one for each noun in the corpus. R(tStreamBackedCorpusViewtabspathR t_read_instance_blockR(R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyt instancesGs cCs+t|i|iƒtd|i|iƒƒS(s @return: a corpus view that acts as a list of strings, one for each line in the predicate-argument annotation file. R(R!R"R tread_line_blockR(R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pytlinesPscCsµ|idƒd}d|}||ijotd|ƒ‚nti|i|ƒiƒƒiƒ}xI|idƒD]"}|i d|jo|SququWtd||fƒ‚dS( sE @return: the xml description for the given roleset. t.is frames/%s.xmlsFrameset file for %s not foundspredicate/rolesettidsRoleset %s not found in %sN( tsplitR t ValueErrorRtparseR"Rtgetroottfindalltattrib(Rt roleset_idtlemmat framefiletetreetroleset((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR3Ys $ cCs+t|i|iƒtd|i|iƒƒS(s‰ @return: a corpus view that acts as a list of all noun lemmas in this corpus (from the nombank.1.0.words file). R(R!R"R R%R(R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pytnounsmscCs`g}xStdƒD]E}|iƒiƒ}|o&|iti||i|iƒƒqqW|S(Nid(trangetreadlinetstriptappendtNombankInstanceR+RR(Rtstreamtblocktitline((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR#vs   ( t__name__t __module__t__doc__RR R R$R&R3R4R#(((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRs    R9cBseeZed„Zed„ƒZd„Zd„Zd„ZeeddƒZ e eed„ƒZ RS(c Cs[||_||_||_||_||_||_||_t|ƒ|_| |_ dS(N( tfileidtsentnumtwordnumtbaseformt sensenumbert predicatetpredidttuplet argumentsR( RRARBRCRDRERFRGRIR((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR ‰s        cCsd|i|ifS(s«The name of the roleset used by this instance's predicate. Use L{nombank.roleset() } to look up information about the roleset.s%s.%s(RDRE(R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR3±scCsd|i|i|ifS(Ns'(RARBRC(R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyt__repr__¸scCsvd|i|i|i|i|if}|i|idff}x.t|ƒD] \}}|d||f7}qNW|S(Ns%s %s %s %s %strels %s-%s(RARBRCtbasenameRERIRFtsorted(Rtstitemstargloctargid((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyt__str__¼s cCsS|idjodSn|i|iiƒjodSn|ii|iƒ|iS(N(RRRARt parsed_sentsRB(R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyt _get_treeÄs!tdocsv The parse tree corresponding to this instance, or C{None} if the corresponding tree is not available.cCs‘|iƒ}t|ƒdjotd|ƒ‚n|d \}}}}}|d} g} t| ƒD]-\} } d| jo| | i| ƒqgqg~ } t| ƒdjotd|ƒ‚n|dj o||ƒ}nt|ƒ}t|ƒ}| diddƒ\}}ti|ƒ}g}xB| D]:}|iddƒ\}}|i ti|ƒ|fƒq1Wt |||||||||ƒ S(Nis Badly formatted nombank line: %ris-reliit-( R)tlenR*t enumeratetpopRtinttNombankTreePointerR+R8R9(RNRRtpiecesRARBRCRDREtargsRR<tpRKtpredlocRGRFRItargRPRQ((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR+Ìs*  G    ( R>R?RR tpropertyR3RJRRRTttreet staticmethodR+(((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR9‡s '     tNombankPointercBseZdZd„ZRS(sf A pointer used by nombank to identify one or more constituents in a parse tree. C{NombankPointer} is an abstract base class with three concrete subclasses: - L{NombankTreePointer} is used to point to single constituents. - L{NombankSplitTreePointer} is used to point to 'split' constituents, which consist of a sequence of two or more C{NombankTreePointer}s. - L{NombankChainTreePointer} is used to point to entire trace chains in a tree. It consists of a sequence of pieces, which can be C{NombankTreePointer}s or C{NombankSplitTreePointer}s. cCs$|itjotdƒ‚ndS(Ns(NombankPointer is an abstract base class(t __class__tNombankPoitnertAssertionError(R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR s(R>R?R@R (((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRdòs tNombankChainTreePointercBs,eZd„Zd„Zd„Zd„ZRS(cCs ||_dS(N(R\(RR\((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR s cCsdid„|iDƒƒS(NRcssx|]}d|VqWdS(s%sN((t.0R^((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pys s(tjoinR\(R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRR scCsd|S(Ns((R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRJ scCsQ|djotdƒ‚ntdg}|iD]}||i|ƒq1~ƒS(NsParse tree not avaialables*CHAIN*(RR*RR\tselect(RRbRR^((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRks(R>R?R RRRJRk(((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRhs   tNombankSplitTreePointercBs,eZd„Zd„Zd„Zd„ZRS(cCs ||_dS(N(R\(RR\((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR s cCsdid„|iDƒƒS(Nt,cssx|]}d|VqWdS(s%sN((RiR^((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pys s(RjR\(R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRRscCsd|S(Ns((R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRJscCsQ|djotdƒ‚ntdg}|iD]}||i|ƒq1~ƒS(NsParse tree not avaialables*SPLIT*(RR*RR\Rk(RRbRR^((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRks(R>R?R RRRJRk(((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRls   R[cBsSeZdZd„Zed„ƒZd„Zd„Zd„Zd„Z d„Z RS(sD wordnum:height*wordnum:height*... wordnum:height, cCs||_||_dS(N(RCtheight(RRCRn((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR 's c Csÿ|idƒ}t|ƒdjo2tg}|D]}|ti|ƒq0~ƒSn|idƒ}t|ƒdjo2tg}|D]}|ti|ƒq„~ƒSn|idƒ}t|ƒdjotd|ƒ‚ntt|dƒt|dƒƒS(NRiRmt:isbad nombank pointer %ri(R)RWRhR[R+RlR*RZ(RNR\Rteltt_[2]((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR++s ( ('cCsd|i|ifS(Ns%s:%s(RCRn(R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRR>scCsd|i|ifS(NsNombankTreePointer(%d, %d)(RCRn(R((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRJAscCsx(t|ttfƒo|id}qWt|tƒptt|ƒt|ƒƒSnt|i|i f|i|i fƒS(Ni( RRhRlR\R[tcmpR(RCRn(Rtother((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyt__cmp__Ds  cCs.|djotdƒ‚n||i|ƒS(NsParse tree not avaialable(RR*ttreepos(RRb((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRkOscCs*|djotdƒ‚n|g}g}d}xñtoét|dtƒoŠt|ƒt|ƒjo|idƒn|dcd7<|dt|dƒjo|i|d|dƒq!|iƒ|iƒq5||ijo#t |t|ƒ|i d ƒSq5|d7}|iƒq5WdS(s} Convert this pointer to a standard 'tree position' pointer, given that it points to the given tree. sParse tree not avaialableiiÿÿÿÿiN( RR*tTrueRRRWR8RYRCRHRn(RRbtstackRuRC((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyRuSs"   # ( R>R?R@R RcR+RRRJRtRkRu(((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pyR[!s    (tretcodecst nltk.treeRt nltk.etreeRtnltk.corpus.reader.utiltnltk.corpus.reader.apiR RtobjectR9RdRhRlR[(((s0/p/zhu/06/nlp/nltk/nltk/corpus/reader/nombank.pys s    uk