³ò B_Kc @sþddkZddkZddklZlZddklZddkTddkTddk l Z ei dƒZ ei dƒZ ei dƒZei d ƒZei d ƒZei d ƒZei d ƒZd efd„ƒYZdee fd„ƒYZdS(iÿÿÿÿN(ttokenizettree(t deprecated(t*(tXMLCorpusReaders]*){0,1}>(.*?)

s]*){0,1}>(.*?)s#<([wc](?: [^>]*){0,1}>)(.*?)s!<[wc](?: [^>]*){0,1}>(.*?)s type="(.*?)"s ana="(.*?)"stext id="(.*?)"t TEICorpusViewcBs2eZeded„ZdZd„Zd„ZRS(icCs>||_||_||_||_ti||d|ƒdS(Ntstartpos(t_taggedt_textidst_group_by_sentt_group_by_paratStreamBackedCorpusViewt__init__(tselft corpus_filettaggedt group_by_sentt group_by_parattag_mapping_functiontheadLenttextids((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyR &s     ic Csï|i|iƒ}t|ƒ}xh|idƒ|idƒjp|idƒdjo2|iƒ}t|ƒdjoPn||7}q!W|iddƒ}ti|ƒ}|i ooxl|D]`}||i joJ|i |ƒd}||i dƒtdƒ}|| |||}q»q»Wng}x¿t i|ƒD]®} g} xwt i| ƒD]f} |i pti| ƒ} nt|iti| ƒƒ} |io| i| ƒqU| i| ƒqUW|io|i| ƒq9|i| ƒq9W|S(Nsis ti(t readlinest _pagesizetconcattcounttreadlinetlentreplacetTEXTIDtfindallRtfindtPARAtSENTRtWORDtmapt _parse_tagt TAGGEDWORDR tappendtextendR ( R tstreamtblockttmpRttidtbegtendtoutputtpara_strtparatsent_strtsent((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyt read_block4sB "  "   cCsZ|\}}|idƒoti|ƒidƒ}nti|ƒidƒ}||fS(Ntwi(t startswithtANAtsearchtgrouptTYPE(R t.1ttagtword((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyR$Zs (t__name__t __module__tNoneR RR3R$(((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyR%s   &tPl196xCorpusReadercBs×eZdZd„Zd„Zd„Zed„Zd„Zeed„Z eeed„Z eeed„Z eeed „Z eeed „Z eeed „Zeeed „Zeed „Zeed„ZRS(iÒ cOsUd|jo|d|_n d|_ti||Œti||ƒ|iƒdS(Nt textid_file(RR?RR tCategorizedCorpusReadert _init_textids(R targstkwargs((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyR fs  cCs×ttƒ|_ttƒ|_|idj o¥x¢|i|iƒiƒD]„}|iƒ}|i ddƒ\}}||i ƒjot dt |fƒ‚nx*|i |i ƒD]}|i||ƒq±WqGWndS(Nt is(In text_id mapping file %s: %s not found(t defaultdicttlistt_f2tt_t2fRR?topenRtstriptsplittfileidst ValueErrortcatfilet _delimitert _add_textids(R tlinetfile_idttext_idsttext_id((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyRCos cCs,|i|i|ƒ|i|i|ƒdS(N(RIR&RJ(R RTRV((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyRR|scs&d}|dj o'|p|df}q:tdƒ‚n|dj o0|pˆi|ƒdf}qwtdƒ‚n|dj ož|p‡t|tƒo |g}nt‡fd†|Dƒgƒ}tƒ}x/|D]'}tˆi|ƒt|ƒ@||s(NN( R?RORNt isinstancet basestringtsumtdicttsetRI(R RNt categoriesRR*tfilesttdicttf((R s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyt_resolve€s(    %cCs|S(N((R R;((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyt decode_tag˜scsvˆi||ƒ\}}|tjotˆiƒSnt|tƒo |g}ntt‡fd†|DƒgƒƒS(sM In the pl196x corpus each category is stored in single file and thus both methods provide identical functionality. In order to accommodate finer granularity, a non-standard textids() method was implemented. All the main functions can be supplied with a list of required chunks---giving much more control to the user. c3s x|]}ˆi|VqWdS(N(RI(RWtd(R (s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pys ©s(RbR?tsortedRJRYRZR[(R RNR^t_((R s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyRœs  cCs÷|i|||ƒ\}}|djo |i}nt|tƒo |g}n|oTtg}|D]8}|t|i|ƒtttd|i d||ƒqg~ƒSnGtg}|D].}|t|i|ƒtttd|i ƒq»~ƒSdS(NRR( RbR?t_fileidsRYRZRRtabspathtFalseR(R RNR^Rt_[1]tfileidt_[2]((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pytwords«s J cCs÷|i|||ƒ\}}|djo |i}nt|tƒo |g}n|oTtg}|D]8}|t|i|ƒtt td|i d||ƒqg~ƒSnGtg}|D].}|t|i|ƒtt td|i ƒq»~ƒSdS(NRR( RbR?RgRYRZRRRhRitTrueR(R RNR^RRjRkRl((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pytsents¼s J cCs÷|i|||ƒ\}}|djo |i}nt|tƒo |g}n|oTtg}|D]8}|t|i|ƒtt t d|i d||ƒqg~ƒSnGtg}|D].}|t|i|ƒtt t d|i ƒq»~ƒSdS(NRR( RbR?RgRYRZRRRhRiRnR(R RNR^RRjRkRl((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pytparasÍs J cCs÷|i|||ƒ\}}|djo |i}nt|tƒo |g}n|oTtg}|D]8}|t|i|ƒtt t d|i d||ƒqg~ƒSnGtg}|D].}|t|i|ƒtt t d|i ƒq»~ƒSdS(NRR( RbR?RgRYRZRRRhRnRiR(R RNR^RRjRkRl((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyt tagged_wordsÞs J cCs÷|i|||ƒ\}}|djo |i}nt|tƒo |g}n|oTtg}|D]8}|t|i|ƒttt d|i d||ƒqg~ƒSnGtg}|D].}|t|i|ƒttt d|i ƒq»~ƒSdS(NRR( RbR?RgRYRZRRRhRnRiR(R RNR^RRjRkRl((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyt tagged_sentsïs J cCs÷|i|||ƒ\}}|djo |i}nt|tƒo |g}n|oTtg}|D]8}|t|i|ƒtttd|i d||ƒqg~ƒSnGtg}|D].}|t|i|ƒtttd|i ƒq»~ƒSdS(NRR( RbR?RgRYRZRRRhRnR(R RNR^RRjRkRl((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyt tagged_parass J cCsS|i||ƒ\}}t|ƒdjoti||dƒSn tdƒ‚dS(NiisExpected a single file(RbRRtxmlt TypeError(R RNR^Rf((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyRts+cCsƒ|i||ƒ\}}|djo |i}nt|tƒo |g}ntg}|D]}||i|ƒiƒq]~ƒS(N(RbR?RgRYRZRRKtread(R RNR^RfRjRa((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pytraws(R=R>RR RCRRR?RbRcRRmRoRpRqRrRsRtRw(((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyR@bs   (tostretnltkRRtnltk.internalsRtutiltapitxmldocsRtcompileR R!R%R"R9R6RR RRBR@(((s//p/zhu/06/nlp/nltk/nltk/corpus/reader/pl196x.pyss    =