o rZh(@sdZddlZddlZddlTddlmZddlmZddlTddl m Z Gddde Z Gd d d eZ Gd d d eZid ddddddddddddddddddd d!d"d#d$d%d&d'd(d)d*d+d,d-d.id/d0d1d2d3d4d5d6d7d8d9d:d;d<d=d>d?d@dAdBdCdDdEdFdGdHdIdJdKdLdMdNdOdNidPdQdRdSdTdUdVdWdXdYdZd[d\d]d^d_d`dadbdcdddedfdgdhdidjdkdldmdndodpdqidrdsdtdudvdwdxdydzd{d|d}d~dddddddddddddddddddddiddddddddddddddddddddddddddddddddddddddddddddddddddŜZdS)a Corpus reader for the York-Toronto-Helsinki Parsed Corpus of Old English Prose (YCOE), a 1.5 million word syntactically-annotated corpus of Old English prose texts. The corpus is distributed by the Oxford Text Archive: http://www.ota.ahds.ac.uk/ It is not included with NLTK. The YCOE corpus is divided into 100 files, each representing an Old English prose text. Tags used within each text complies to the YCOE standard: https://www-users.york.ac.uk/~lang22/YCOE/YcoeHome.htm N)*)BracketParseCorpusReader)TaggedCorpusReader)RegexpTokenizerc@s|eZdZdZdddZdddZddd Zd d Zdd d ZdddZ dddZ dddZ dddZ dddZ dddZdS)YCOECorpusReaderz Corpus reader for the York-Toronto-Helsinki Parsed Corpus of Old English Prose (YCOE), a 1.5 million word syntactically-annotated corpus of Old English prose texts. utf8cCst||g|t|jddd|d|_t|jddd|_dd|jD}d d|jD|kr:t d t d d |Dd d |D}t||||t ||_ dS)Npsdz.*.psd)encodingpos.poscSh|]}|ddqSN.0frrF/var/www/auris/lib/python3.10/site-packages/nltk/corpus/reader/ycoe.py .z,YCOECorpusReader.__init__..cSr rrrrrrr/rz5Items in "psd" and "pos" subdirectories do not match.cSg|]}d|qSz%s.psdrrdocrrr 3z-YCOECorpusReader.__init__..cSrz%s.posrrrrrr4r) CorpusReader__init__YCOEParseCorpusReaderrootjoin _psd_readerYCOETaggedCorpusReader _pos_readerfileids ValueErrorsorted _documents)selfr!r documentsr&rrrr%s  zYCOECorpusReader.__init__NcCsP|dur|jSt|tr|g}|D] }||jvrtd|qtdd|DS)z Return a list of document identifiers for all documents in this corpus, or for the documents with the given file(s) if specified. NzFile id %s not foundcSr rrrrrrrGrz-YCOECorpusReader.documents..)r) isinstancestr_fileidsKeyErrorr()r*r&rrrrr+9s   zYCOECorpusReader.documentscCsB|dur|jSt|tr|g}ttdd|Ddd|DS)z Return a list of file identifiers for the files that make up this corpus, or that store the given document(s) if specified. NcSrrrrrrrrTrz,YCOECorpusReader.fileids..cSrrrrrrrrUr)r.r,r-r(setr*r+rrrr&Is   zYCOECorpusReader.fileidscsj|dur|j}n$t|tr|g}|D]}||jvr+|dddvr%tdtd|qfdd|DS)z Helper that selects the appropriate fileids for a given set of documents from a given subcorpus (pos or psd). Nr)r r zvExpected a document identifier, not a file identifier. (Use corpus.documents() to get a list of document identifiers.z Document identifier %s not foundcsg|] }|dqS).r)rd subcorpusrrrmsz0YCOECorpusReader._getfileids..)r)r,r-r')r*r+r5Zdocumentrr4r _getfileidsYs    zYCOECorpusReader._getfileidscC|j||dSNr )r%wordsr6r1rrrr9pzYCOECorpusReader.wordscCr7r8)r%sentsr6r1rrrr;sr:zYCOECorpusReader.sentscCr7r8)r%parasr6r1rrrr<vr:zYCOECorpusReader.parascCr7r8)r% tagged_wordsr6r1rrrr=yr:zYCOECorpusReader.tagged_wordscCr7r8)r% tagged_sentsr6r1rrrr>|r:zYCOECorpusReader.tagged_sentscCr7r8)r% tagged_parasr6r1rrrr?r:zYCOECorpusReader.tagged_parascCr7)Nr)r# parsed_sentsr6r1rrrr@r:zYCOECorpusReader.parsed_sentsr)N)__name__ __module__ __qualname____doc__rr+r&r6r9r;r<r=r>r?r@rrrrrs         rc@seZdZdZddZdS)r zrSpecialized version of the standard bracket parse corpus reader that strips out (CODE ...) and (ID ...) nodes.cCs*tdd|}td|rdSt||S)Nz(?u)\((CODE|ID)[^\)]*\)z\s*\(\s*\)\s*$)resubmatchr_parse)r*trrrrJs  zYCOEParseCorpusReader._parseN)rBrCrDrErJrrrrr s r c@seZdZdddZdS)r$rcCs(d}t|dd}tj|||d|ddS)Nz+(?u)(?<=/\.)\s+|\s*\S*_CODE\s*|\s*\S*_ID\s*T)Zgaps_)sepsent_tokenizer)rrr)r*r!itemsr Zgaps_rerNrrrrs    zYCOETaggedCorpusReader.__init__NrA)rBrCrDrrrrrr$sr$z coadrian.o34zAdrian and Ritheusz coaelhom.o3uÆlfric, Supplemental Homiliesz coaelive.o3uÆlfric's Lives of SaintsZcoalcuinzAlcuin De virtutibus et vitiisz coalex.o23zAlexander's Letter to Aristotlez coapollo.o3zApollonius of TyreZcoaugustZ Augustinez cobede.o2z$Bede's History of the English Churchz cobenrul.o3zBenedictine Rulez coblick.o23zBlickling Homiliesz coboeth.o2z#Boethius' Consolation of Philosophyz cobyrhtf.o3zByrhtferth's ManualZ cocanedgDzCanons of Edgar (D)Z cocanedgXzCanons of Edgar (X)z cocathom1.o3uÆlfric's Catholic Homilies Iz cocathom2.o3uÆlfric's Catholic Homilies IIz cochad.o24z Saint ChadZcochdrulzChrodegang of Metz, RuleZ cochristophzSaint Christopherz cochronA.o23zAnglo-Saxon Chronicle AZcochronCzAnglo-Saxon Chronicle CZcochronDzAnglo-Saxon Chronicle Dz cochronE.o34zAnglo-Saxon Chronicle Ez cocura.o2zCura PastoralisZcocuraCzCura Pastoralis (Cotton)z codicts.o34z Dicts of Catoz codocu1.o1zDocuments 1 (O1)z codocu2.o12zDocuments 2 (O1/O2)z codocu2.o2zDocuments 2 (O2)z codocu3.o23zDocuments 3 (O2/O3)z codocu3.o3zDocuments 3 (O3)z codocu4.o24zDocuments 4 (O2/O4)Zcoeluc1z Honorius of Autun, Elucidarium 1Zcoeluc2z coepigen.o3uÆlfric's Epilogue to GenesisZcoeuphrzSaint EuphrosyneZcoeustz Saint Eustace and his companionsZ coexodusPz Exodus (P)Z cogenesiCz Genesis (C)z cogregdC.o24zGregory's Dialogues (C)z cogregdH.o23zGregory's Dialogues (H)ZcoherbarzPseudo-Apuleius, Herbariumz coinspolD.o34z"Wulfstan's Institute of Polity (D)Z coinspolXz"Wulfstan's Institute of Polity (X)Zcojamesz Saint Jamesz colacnu.o23ZLacnungaz colaece.o2Z Leechdomsz colaw1cn.o3z Laws, Cnut Iz colaw2cn.o3z Laws, Cnut IIz colaw5atr.o3uLaws, Æthelred Vz colaw6atr.o3uLaws, Æthelred VIz colawaf.o2z Laws, Alfredz colawafint.o2zAlfred's Introduction to Lawsz colawger.o34z Laws, Gerefaz colawine.ox2z Laws, Inezcolawnorthu.o3zNorthumbra Preosta Laguz colawwllad.o4zLaws, William I, Ladz coleofri.o4ZLeofricz colsigef.o3uÆlfric's Letter to SigefyrthZ colsigewBu!Ælfric's Letter to Sigeweard (B)z colsigewZ.o34u!Ælfric's Letter to Sigeweard (Z)ZcolwgeatuÆlfric's Letter to WulfgeatZ colwsigeTu Ælfric's Letter to Wulfsige (T)zcolwsigeXa.o34u!Ælfric's Letter to Wulfsige (Xa)z colwstan1.o3uÆlfric's Letter to Wulfstan Iz colwstan2.o3uÆlfric's Letter to Wulfstan IIz comargaC.o34zSaint Margaret (C)ZcomargaTzSaint Margaret (T)Zcomart1zMartyrology, IZcomart2zMartyrology, IIz comart3.o23zMartyrology, IIIz comarvel.o23zMarvels of the EastZcomaryz Mary of EgyptZconeotz Saint NeotZconicodAzGospel of Nicodemus (A)ZconicodCzGospel of Nicodemus (C)ZconicodDzGospel of Nicodemus (D)ZconicodEzGospel of Nicodemus (E)z coorosiu.o2ZOrosiusz cootest.o3Z Heptateuchzcoprefcath1.o3u(Ælfric's Preface to Catholic Homilies Izcoprefcath2.o3u)Ælfric's Preface to Catholic Homilies IIz coprefcura.o2zPreface to the Cura Pastoralisz coprefgen.o3uÆlfric's Preface to Genesiszcopreflives.o3u$Ælfric's Preface to Lives of Saintsz"Preface to Augustine's Soliloquiesz*Pseudo-Apuleius, Medicina de quadrupedibuszHistory of the Holy Rood-TreezSeven SleeperszSt. Augustine's SoliloquieszSolomon and Saturn IzSolomon and Saturn IIuÆlfric's De Temporibus AnnizVercelli HomilieszVercelli Homilies (E)zVercelli Homilies (L)zSaint Vincent (Bodley 343)zVindicta SalvatoriszWest-Saxon GospelszWulfstan's Homilies)Z coprefsoliloz coquadru.o23ZcoroodZ cosevenslZcosoliloz cosolsat1.o4Z cosolsat2z cotempo.o3ZcoverhomZ coverhomEZ coverhomLZcovinceBZcovinsalz cowsgosp.o3z cowulf.o34)rEosrGZnltk.corpus.reader.apiZ nltk.corpus.reader.bracket_parserZnltk.corpus.reader.taggedrZnltk.corpus.reader.utilZ nltk.tokenizerrrr r$r+rrrrs    h        !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUV