g`FdZddlZddlZddlmZddlmZGddeZy)z An NLTK interface to the VerbNet verb lexicon For details about VerbNet see: https://verbs.colorado.edu/~mpalmer/projects/verbnet.html N) defaultdict)XMLCorpusReadercFeZdZdZd"dZej dZ ej dZ ej dZ d#dZ d#dZ d$d Z d Z d#d Zd Zd ZdZdZdZdZdZdZdZdZdZdZdZd%dZd%dZd%dZd%dZd%dZ d%dZ!d%dZ"d%d Z#d%d!Z$y)&VerbnetCorpusReadera An NLTK interface to the VerbNet verb lexicon. From the VerbNet site: "VerbNet (VN) (Kipper-Schuler 2006) is the largest on-line verb lexicon currently available for English. It is a hierarchical domain-independent, broad-coverage verb lexicon with mappings to other lexical resources such as WordNet (Miller, 1990; Fellbaum, 1998), XTAG (XTAG Research Group, 2001), and FrameNet (Baker et al., 1998)." For details about VerbNet see: https://verbs.colorado.edu/~mpalmer/projects/verbnet.html ctj||||tt|_ tt|_ i|_ i|_|jyN) r__init__rlist_lemma_to_class_wordnet_to_class_class_to_fileid_shortid_to_longid _quick_index)selfrootfileids wrap_etrees O/var/www/openai/venv/lib/python3.12/site-packages/nltk/corpus/reader/verbnet.pyr zVerbnetCorpusReader.__init__%sd  tWjA*40 &"-T!2 /!# B#% z([^\-\.]*)-([\d+.\-]+)$z [\d+.\-]+$zH]+>|Nc|#t|jjSt|tr|j |}|j dDcgc]}|jdc}Scc}w)zx Return a list of all verb lemmas that appear in any class, or in the ``classid`` if specified. MEMBERS/MEMBERname)sortedr keys isinstancestrvnclassfindallget)rrmembers rlemmaszVerbnetCorpusReader.lemmasHsj ?$..3356 6'3',,w/5<__EU5VW5V6FJJv&5VW WWsA5c|#t|jjSt|tr|j |}t d|jdDgS)z| Return a list of all wordnet identifiers that appear in any class, or in ``classid`` if specified. c3\K|]$}|jddj&yw)wnN)rsplit).0r s r z1VerbnetCorpusReader.wordnetids..as,"CJJtR(..0"Cs*,r)rr rrrrsumr)rrs r wordnetidszVerbnetCorpusReader.wordnetidsUsf ?$005578 8'3',,w/")//2B"C  rc|3|jjDcgc] \}}||k(s |c}}S||j|S||j|S|?|j |}|j dDcgc]}|j dc}St|jjScc}}wcc}w)a Return a list of the VerbNet class identifiers. If a file identifier is specified, then return only the VerbNet class identifiers for classes (and subclasses) defined by that file. If a lemma is specified, then return only VerbNet class identifiers for classes that contain that lemma as a member. If a wordnetid is specified, then return only identifiers for classes that contain that wordnetid as a member. If a classid is specified, then return only identifiers for subclasses of the specified VerbNet class. If nothing is specified, return all classids within VerbNet SUBCLASSES/VNSUBCLASSID) r itemsr r rrrrr) rlemma wordnetidfileidclassidcfxmltreesubclasss rclassidszVerbnetCorpusReader.classidshs  $($9$9$?$?$AQ$A&1aQ&[A$AQ Q  ''. .  ")))4 4  ll7+G!(0G H HH T" H  $//4467 7Rs B<B<>Cc||jvr|j|S|j|}||jvru|j|j|}|j|}||j dk(r|S|j dD]}||j dk(s|cSJt d|)aReturns VerbNet class ElementTree Return an ElementTree containing the xml for the specified VerbNet class. :param fileid_or_classid: An identifier specifying which class should be returned. Can be a file identifier (such as ``'put-9.1.xml'``), or a VerbNet class identifier (such as ``'put-9.1'``) or a short VerbNet class identifier (such as ``'9.1'``). r-z .//VNSUBCLASSzUnknown identifier )_fileidsxmllongidr rr ValueError)rfileid_or_classidr2r1treer6s rrzVerbnetCorpusReader.vnclasss  -88-. .++/0 d++ +**4;;w+?@F88F#D$((4.( $ _ =H(,,t"44'!>!523D2EFG Grc| |jSt|tr|j|j |gS|Dcgc] }|j|j |"c}Scc}w)z Return a list of fileids that make up this corpus. If ``vnclass_ids`` is specified, then return the fileids that make up the specified VerbNet class(es). )r9rrr r;)r vnclass_ids vnclass_ids rrzVerbnetCorpusReader.fileidssv  ==  S )))$++k*BCD D#."-J%%dkk*&=>"- s%A*c  t|tr|j|}g}|jd}|D]T}|j |j ||j ||j||j|dV|S)apGiven a VerbNet class, this method returns VerbNet frames The members returned are: 1) Example 2) Description 3) Syntax 4) Semantics :param vnclass: A VerbNet class identifier; or an ElementTree containing the xml contents of a VerbNet class. :return: frames - a list of frame dictionaries z FRAMES/FRAME)example descriptionsyntax semantics) rrrrappend_get_example_within_frame_get_description_within_frame _get_syntactic_list_within_frame_get_semantics_within_frame)rrframesvnframesvnframes rrLzVerbnetCorpusReader.framess gs #ll7+G??>2G MM#==gF#'#E#Eg#N"CCGL!%!A!A'!J    rct|tr|j|}|jdDcgc]}|j d}}|Scc}w)aAReturns subclass ids, if any exist Given a VerbNet class, this method returns subclass ids (if they exist) in a list of strings. :param vnclass: A VerbNet class identifier; or an ElementTree containing the xml contents of a VerbNet class. :return: list of subclasses r,r-)rrrrr)rrr6 subclassess rrPzVerbnetCorpusReader.subclassess[ gs #ll7+G07?V/W /W8HLL /W   sAc Dt|tr|j|}g}|jdD]c}|j |j d|jdDcgc]%}|j d|j dd'c}de|Scc}w)abReturns thematic roles participating in a VerbNet class Members returned as part of roles are- 1) Type 2) Modifiers :param vnclass: A VerbNet class identifier; or an ElementTree containing the xml contents of a VerbNet class. :return: themroles: A list of thematic roles in the VerbNet class zTHEMROLES/THEMROLEtypeSELRESTRS/SELRESTRValuevaluerR)rR modifiers)rrrrrGr)rr themrolestrolerestrs rrXzVerbnetCorpusReader.themroless gs #ll7+G __%9:E   !IIf-&+]]3G%H"%HE#())G"4eii>OP%H" ; "s&*B ch|jD]#}|j|j||%y)aC Initialize the indexes ``_lemma_to_class``, ``_wordnet_to_class``, and ``_class_to_fileid`` by scanning through the corpus fileids. This is fast if ElementTree uses the C implementation (<0.1 secs), but quite slow (>10 secs) if only the python implementation is available. N)r9 _index_helperr:)rr1s r_indexzVerbnetCorpusReader._indexs+mmF   txx/ 8$rc|jd}||j|<||j|j|<|j dD]r}|j |jdj ||jddjD] }|j|j |"t|j dD]}|j||y)zHelper for ``_index()``r-rrr$r%r,N) rr rshortidrr rGr&r r\)rr5r1rr r$r6s rr\z!VerbnetCorpusReader._index_helper s++d#)/g&9@ W 56oo&67F  F!3 4 ; ;G Djjr*002&&r*11':38 (?@H   x 0Arc|jD],}|dd}||j|<||j|j|<|j |5}|j j |jD]}|j}|dX|j|dj||djD] }|j|j|"p|d6||j|d<|d}||j|j|<Jd ddd/y#1swY;xYw)a Initialize the indexes ``_lemma_to_class``, ``_wordnet_to_class``, and ``_class_to_fileid`` by scanning through the corpus fileids. This doesn't do proper xml parsing, but is good enough to find everything in the standard VerbNet corpus -- and it runs about 30 times faster than xml parsing (with the python ElementTree; only 2-3 times faster if ElementTree uses the C implementation). Nrzunexpected match condition) r9r rr_open _INDEX_REfinditerreadgroupsr rGr&r )rr1rfpmrhr$s rrz VerbnetCorpusReader._quick_indexs/mmFSbkG-3D ! !' *=DD # #DLL$9 :6"b00;AXXZFay,,,VAY7>>wG"()//"3B 2226==gF#4.;A--fQi8"()IP// W0EFB&BBu<#" $#"s CD>>E c|jj|r|S|jj|std|z |j|S#t $r}td|z|d}~wwxYw)zReturns longid of a VerbNet class Given a short VerbNet class identifier (eg '37.10'), map it to a long id (eg 'confess-37.10'). If ``shortid`` is already a long id, then return it as-isvnclass identifier %r not foundN) _LONGID_REmatch _SHORTID_REr<rKeyError)rr_es rr;zVerbnetCorpusReader.longid;s| ??  )N!!''0>HI I Q**73 3 Q>HIq P QsA A4 A//A4c|jj|r|S|jj|}|r|jdSt d|z)zReturns shortid of a VerbNet class Given a long VerbNet class identifier (eg 'confess-37.10'), map it to a short id (eg '37.10'). If ``longid`` is already a short id, then return it as-is.rcrl)rornrmgroupr<)rr;rjs rr_zVerbnetCorpusReader.shortidJsS    ! !& )M OO ! !& ) 771: >GH Hrc*g}|jdD]w}|jdDcgc]%}|jd|jdd'}}|j|jd||jddk(dy|Scc}w) anReturns semantics within a single frame A utility function to retrieve semantics within a frame in VerbNet Members of the semantics dictionary: 1) Predicate value 2) Arguments :param vnframe: An ElementTree containing the xml contents of a VerbNet frame. :return: semantics: semantics dictionary zSEMANTICS/PREDzARGS/ARGrRrV)rRrVbool!)predicate_value argumentsnegated)rrrG)rrNsemantics_within_single_framepredargrxs rrKz/VerbnetCorpusReader._get_semantics_within_frame\s)+%OO$45D << 33C37773CD3  * 0 0'+xx'8!*#xx/36  6-,s*BcL|jd}||j}|Sd}|S)a'Returns example within a frame A utility function to retrieve an example within a frame in VerbNet. :param vnframe: An ElementTree containing the xml contents of a VerbNet frame. :return: example_text: The example sentence for this particular frame zEXAMPLES/EXAMPLEr%)findtext)rrNexample_element example_texts rrHz-VerbnetCorpusReader._get_example_within_framews9",,'9:  &*//LLrch|jd}|jd|jdddS)abReturns member description within frame A utility function to retrieve a description of participating members within a frame in VerbNet. :param vnframe: An ElementTree containing the xml contents of a VerbNet frame. :return: description: a description dictionary with members - primary and secondary DESCRIPTIONprimary secondaryr%)rr)r~attribr)rrNdescription_elements rrIz1VerbnetCorpusReader._get_description_within_frames;&ll=9*11)<,00bA  rcg}|jdD]}|j}t}d|jvr|j dnd|d<|j dDcgc]%}|j d|j dd'c}|d<|j d Dcgc]%}|j d|j dd'c}|d <|j ||d |Scc}wcc}w) a[Returns semantics within a frame A utility function to retrieve semantics within a frame in VerbNet. Members of the syntactic dictionary: 1) POS Tag 2) Modifiers :param vnframe: An ElementTree containing the xml contents of a VerbNet frame. :return: syntax_within_single_frame SYNTAXrVr%rSrTrRrU selrestrszSYNRESTRS/SYNRESTR synrestrs)pos_tagrW)r~tagdictrrrrG)rrNsyntax_within_single_frameeltrrWrZs rrJz4VerbnetCorpusReader._get_syntactic_list_within_frames&("<<)CggGI5< 5J!1PRIg ![[)=>&>E ))G,eii6GH>&Ik " ![[)=>&>E ))G,eii6GH>&Ik " ' - -#)< **)&&s $*C/'*C4cFt|tr|j|}|jddz}||j |ddzz }||j |ddzz }|dz }||j |ddzz }|dz }||j|dz }|S)a%Returns pretty printed version of a VerbNet class Return a string containing a pretty-printed representation of the given VerbNet class. :param vnclass: A VerbNet class identifier; or an ElementTree containing the xml contents of a VerbNet class. r-  )indentz Thematic roles:  z Frames: )rrrrpprint_subclassespprint_memberspprint_themroles pprint_frames)rrss rpprintzVerbnetCorpusReader.pprints gs #ll7+G KK  $ T # #GD # 9D @@ T  6 == "" T " "76 " :T AA ] T    77rct|tr|j|}|j|}|sdg}ddj |z}t j |d||dzS)a>Returns pretty printed version of subclasses of VerbNet class Return a string containing a pretty-printed representation of the given VerbNet class's subclasses. :param vnclass: A VerbNet class identifier; or an ElementTree containing the xml contents of a VerbNet class. (none)z Subclasses:  Frinitial_indentsubsequent_indent)rrrrPjointextwrapfill)rrrrPrs rrz%VerbnetCorpusReader.pprint_subclassesse gs #ll7+G__W- "J SXXj1 1}} r&FTM  rct|tr|j|}|j|}|sdg}ddj |z}t j |d||dzS)a?Returns pretty printed version of members in a VerbNet class Return a string containing a pretty-printed representation of the given VerbNet class's member verbs. :param vnclass: A VerbNet class identifier; or an ElementTree containing the xml contents of a VerbNet class. rz Members: rrrr)rrrr!rrr)rrrmembersrs rrz"VerbnetCorpusReader.pprint_membersse gs #ll7+G++g&jG #((7+ +}} r&FTM  rcpt|tr|j|}g}|j|D]j}|dz|j dz}|dDcgc] }|d|dz}}|r#|dj dj |z }|j|ldj |Scc}w)aHReturns pretty printed version of thematic roles in a VerbNet class Return a string containing a pretty-printed representation of the given VerbNet class's thematic roles. :param vnclass: A VerbNet class identifier; or an ElementTree containing the xml contents of a VerbNet class. * rRrWrV[{}]rr)rrrrXrformatrrG)rrrpiecesthemrolepiecemodifierrWs rrz$VerbnetCorpusReader.pprint_themroless gs #ll7+Gw/HTMHLL$88E!) 5 5H!HV$44 5 sxx ':;; MM% 0yy  sB3ct|tr|j|}g}|j|D]#}|j |j ||%dj |S)a?Returns pretty version of all frames in a VerbNet class Return a string containing a pretty-printed representation of the list of frames within the VerbNet class. :param vnclass: A VerbNet class identifier; or an ElementTree containing the xml contents of a VerbNet class. r)rrrrLrG_pprint_single_framer)rrrrrNs rrz!VerbnetCorpusReader.pprint_framess\ gs #ll7+G{{7+G MM$33GVD E,yy  rc|j||dz}||j||dzdzz }||j||dzdzz }||dzz }||j||dzz }|S)aReturns pretty printed version of a single frame in a VerbNet class Returns a string containing a pretty-printed representation of the given frame. :param vnframe: An ElementTree containing the xml contents of a VerbNet frame. rrz Syntax: z Semantics: r) _pprint_description_within_frame_pprint_example_within_frame_pprint_syntax_within_frame_pprint_semantics_within_frame)rrNr frame_strings rrz(VerbnetCorpusReader._pprint_single_frames<>'-*@*MN NKrc g}|dD]}|d}g}d|dvr|ddr|j|dd||dd|ddzDcgc]}dj|d|dc}z }|r#|d jd j|z }|j||d j|zScc}w) a&Returns pretty printed version of syntax within a frame in a VerbNet class Return a string containing a pretty-printed representation of the given VerbNet frame syntax. :param vnframe: An ElementTree containing the xml contents of a VerbNet frame. rErrVrWrrz{}{}rRrr)rGrr)rrNrrelementr modifier_listrZs rrz/VerbnetCorpusReader._pprint_syntax_within_frameJsx(GI&EM'+..7;3G3P$$W[%9'%BC K(5k*;788 eGneFm<8 Msxx '>?? MM% ) (((s B<c g}|dD]I}|dDcgc]}|d }}|j|drdnd|ddd j|d Kd jfd |DScc}w) a,Returns a pretty printed version of semantics within frame in a VerbNet class Return a string containing a pretty-printed representation of the given VerbNet frame semantics. :param vnframe: An ElementTree containing the xml contents of a VerbNet frame. rFrxrVry¬r%rw(z, )rc3,K|] }d| yw)rNr)r'rrs rr(zEVerbnetCorpusReader._pprint_semantics_within_frame..usB6%F82eW-6s)rGr)rrNrr predicateargumentrxs ` rrz2VerbnetCorpusReader._pprint_semantics_within_framefs -I;D[;QR;Qx'*;QIR MM$Y/4R8CT9U8VVWX\XaXabkXlWmmno . yyB6BBB Ss A/)Fr)NNNN)r%)%__name__ __module__ __qualname____doc__r recompilermrorer!r*r7rrrLrPrXr]r\rr;r_rKrHrIrJrrrrrrrrrrrrrrrs .67JB"**]+K2 WI! X&88HB 8$> 9 1CD Q I$-6  *D* ( (!0! $ > )8Crr)rrr collectionsrnltk.corpus.reader.xmldocsrrrrrrs( #6_ C/_ Cr