+
    ÖÇÜh§.  ã                   ó8  € ^ RI 5 ^ RIHt ]P	                  R4      t]P	                  R4      t]P	                  R4      t]P	                  R4      t]P	                  R4      t	]P	                  R4      t
]P	                  R	4      t ! R
 R]4      t ! R R]]4      tR# )é    )Ú*)ÚXMLCorpusReaderz<p(?: [^>]*){0,1}>(.*?)</p>z<s(?: [^>]*){0,1}>(.*?)</s>z#<([wc](?: [^>]*){0,1}>)(.*?)</[wc]>z!<[wc](?: [^>]*){0,1}>(.*?)</[wc]>ztype="(.*?)"zana="(.*?)"ztext id="(.*?)"c                   ó:   a € ] tR t^t o RR ltRtR tR tRtV t	R# )ÚTEICorpusViewNc                ód   € W n         Wpn        W0n        W@n        \        P                  WVR 7       R# ))ÚstartposN)Ú_taggedÚ_textidsÚ_group_by_sentÚ_group_by_paraÚStreamBackedCorpusViewÚ__init__)ÚselfÚcorpus_fileÚtaggedÚgroup_by_sentÚgroup_by_paraÚtagsetÚhead_lenÚtextidss   &&&&&&&&Ú;/usr/lib/python3/dist-packages/nltk/corpus/reader/pl196x.pyr   ÚTEICorpusView.__init__   s-   € ð ŒØŒà+ÔØ+Ôä×'Ñ'¨ÀHÐ'ÖMó    i   c           
     ój  € VP                  V P                  4      p\        V4      pVP                  R 4      VP                  R4      8”  g   VP                  R 4      ^ 8X  d,   VP	                  4       p\        V4      ^ 8:  d   M
W#,          pKf  VP                  RR4      p\        P                  V4      pV P                  '       dl   V Fe  pWPP                  9  g   K  VP                  V4      ^,
          pW&R P                  R4      \        R4      ,           pVRV W&V,           R ,           pKg  	  . p\        P                  V4       Fç  p	. p
\        P                  V	4       F“  pV P                  '       g   \        P                  V4      pM2\        \!        V P"                  \$        P                  V4      4      4      pV P&                  '       d   V
P)                  V4       K‚  V
P+                  V4       K•  	  V P,                  '       d   VP)                  V
4       KÖ  VP+                  V
4       Ké  	  V# )z<text idz</text>Ú
Ú N)Ú	readlinesÚ	_pagesizeÚconcatÚcountÚreadlineÚlenÚreplaceÚTEXTIDÚfindallr
   ÚfindÚPARAÚSENTr	   ÚWORDÚlistÚmapÚ
_parse_tagÚ
TAGGEDWORDr   ÚappendÚextendr   )r   ÚstreamÚblockÚtmpr   ÚtidÚbegÚendÚoutputÚpara_strÚparaÚsent_strÚsents   &&           r   Ú
read_blockÚTEICorpusView.read_block,   s±  € Ø× Ñ  §¡Ó0ˆÜ�u“ˆØ�{‰{˜:Ó&¨¯©°YÓ)?Ô?ÀEÇKÁKØóE
àôEð —/‘/Ó#ˆCÜ�3‹x˜1Œ}ØØ�LŠEà—‘˜d BÓ'ˆä—.‘. Ó'ˆØ�=�=ˆ=Û�ØŸm™mÖ+ØŸ*™* S›/¨AÕ-�CØ ˜+×*Ñ*¨9Ó5¼¸I»ÕF�CØ! $ 3˜K¨%°cµ	°Ð*<Õ<’Eñ	 ð ˆÜŸ™ UÖ+ˆHØˆDÜ ŸL™L¨Ö2�Ø—|—|�|ÜŸ<™<¨Ó1‘Dä¤ D§O¡O´Z×5GÑ5GÈÓ5QÓ RÓS�DØ×&×&Ð&Ø—K‘K Ö%à—K‘K Ö%ñ 3ð ×"×"Ð"Ø—‘˜dÖ#à—‘˜dÖ#ñ ,ð ˆr   c                óÔ   € Vw  r#VP                  R 4      '       d(   \        P                  V4      P                  ^4      pW23# \        P                  V4      P                  ^4      pW23# )Úw)Ú
startswithÚANAÚsearchÚgroupÚTYPE)r   Útag_word_tupleÚtagÚwords   &&  r   r,   ÚTEICorpusView._parse_tagS   s]   € Ø$‰ˆØ�>‰>˜#×ÒÜ—*‘*˜S“/×'Ñ'¨Ó*ˆCð ˆyÐô —+‘+˜cÓ"×(Ñ(¨Ó+ˆCØˆyÐr   )r   r   r	   r
   )Nr   N)
Ú__name__Ú
__module__Ú__qualname__Ú__firstlineno__r   r   r;   r,   Ú__static_attributes__Ú__classdictcell__©Ú__classdict__s   @r   r   r      s"   ø‡ € ôNð$ €Iò%÷Nð r   r   c                   ó–   a € ] tR t^\t o RtR tR tR tRR ltR t	RR lt
RR	 ltRR
 ltRR ltRR ltRR ltRR ltRR ltRtV tR# )ÚPl196xCorpusReaderiÒ
  c                óº   € R V9   d   VR ,          V n         MRV n         \        P                  ! V .VO5!   \        P                  W4       V P	                  4        R# )Útextid_fileN)r
   r   r   ÚCategorizedCorpusReaderÚ_init_textids)r   ÚargsÚkwargss   &*,r   r   ÚPl196xCorpusReader.__init___   sJ   € Ø˜FÔ"Ø" =Õ1ˆD�Mà ˆDŒMä× Ò  Ð-¨Ô-Ü×(Ñ(¨Ô6à×ÑÖr   c           
     ó  € \        \        4      V n        \        \        4      V n        V P                  e»   \        V P                  4      ;_uu_ 4       pV FŒ  pVP                  4       pVP                  R^4      w  r4W0P                  4       9  d   \        RV P                  : RV: R24      hVP                  V P                  4       F  pV P                  W54       K  	  KŽ  	  R R R 4       R # R #   + '       g   i     R # ; i)NÚ zIn text_id mapping file z: z
 not found)Údefaultdictr*   Ú_f2tÚ_t2fr
   ÚopenÚstripÚsplitÚfileidsÚ
ValueErrorÚ
_delimiterÚ_add_textids)r   ÚfpÚlineÚfile_idÚtext_idsÚtext_ids   &     r   rU   Ú Pl196xCorpusReader._init_textidsj   sÃ   € Ü¤Ó%ˆŒ	Ü¤Ó%ˆŒ	Ø�=‰=Ò$Ü�d—m‘m×$Ô$¨Û�DØŸ:™:›<�DØ(,¯
©
°3¸Ó(:Ñ%�GØ§l¡l£nÔ4Ý(à#Ÿ}œ}«gð7óð ð $,§>¡>°$·/±/Ö#B˜Ø×)Ñ)¨'Ö;ó $Cñ ÷ %Ñ$ñ %ß$×$Ð$ús   ÁBC3Ã3D	c                óŽ   € V P                   V,          P                  V4       V P                  V,          P                  V4       R # ©N)r\   r.   r]   )r   rg   ri   s   &&&r   rd   ÚPl196xCorpusReader._add_textidsz   s0   € Ø�	‰	�'Õ×!Ñ! 'Ô*Ø�	‰	�'Õ×!Ñ! 'Ö*r   Nc           
     óš  a € R p\        \        \        R WV34      4      4      ^8w  d   \        R4      hVe   VR 3# Ve   S P	                  V4      R 3# Vet   \        V\        4      '       d   V.p\        V 3R lV 4       . 4      p\        4       pV F1  p\        S P                  V,          4      \        V4      ,          Wg&   K3  	  WV3# R # )Nc                 ó
   € V R J # rl   © )Úaccessors   &r   Ú<lambda>Ú-Pl196xCorpusReader._resolve.<locals>.<lambda>„   s	   € ¨°TÑ)9r   z6Specify exactly one of: fileids, categories or textidsc              3   óJ   <"  € T F  pSP                   V,          x € K  	  R # 5irl   )r]   )Ú.0Útr   s   & €r   Ú	<genexpr>Ú.Pl196xCorpusReader._resolve.<locals>.<genexpr>˜   s   øé € Ð7©w¨!˜Ÿ™ 1Ÿš«wùó   ƒ #)r"   r*   Úfilterrb   ra   Ú
isinstanceÚstrÚsumÚdictÚsetr\   )r   ra   Ú
categoriesr   r2   ÚfilesÚtdictÚfs   f&&&    r   Ú_resolveÚPl196xCorpusReader._resolve~   sÚ   ø€ ØˆäÜÜÙ9Ø ¨gÐ6óóóð ôô ØKóð ð ÒØ˜D�=Ð àÒ!Ø—<‘< 
Ó+¨TÐ1Ð1àÒÜ˜'¤3×'Ò'Ø"˜)�ÜÔ7©wÓ7¸Ó<ˆEÜ“FˆEÛ�Ü˜tŸy™y¨�|Ó,¬s°7«|Õ;�“ñ à�<Ðñ r   c                ó   € V# rl   rp   )r   rE   s   &&r   Ú
decode_tagÚPl196xCorpusReader.decode_tagž   s   € àˆ
r   c                óÌ   a € S P                  W4      w  rVf   \        S P                  4      # \        V\        4      '       d   V.p\        \        V 3R lV 4       . 4      4      # )a>  
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.
c              3   óJ   <"  € T F  pSP                   V,          x € K  	  R # 5irl   )r\   )ru   Údr   s   & €r   rw   Ú-Pl196xCorpusReader.textids.<locals>.<genexpr>°   s   øé € Ð9±¨A˜4Ÿ9™9 QŸ<š<³ùry   )r„   Úsortedr]   r{   r|   r}   ©r   ra   r€   Ú_s   f&& r   r   ÚPl196xCorpusReader.textids¢   sU   ø€ ð —]‘] 7Ó7‰
ˆØŠ?Ü˜$Ÿ)™)Ó$Ð$ä�gœs×#Ò#Ø�iˆGÜ”cÔ9±Ó9¸2Ó>Ó?Ð?r   c                ó¸  € V P                  WV4      w  rVf   V P                  pM\        V\        4      '       d   V.pV'       dJ   \	        V Uu. uF3  p\        V P                  V4      RRRV P                  W4,          R7      NK5  	  up4      # \	        V Uu. uF,  p\        V P                  V4      RRRV P                  R7      NK.  	  up4      # u upi u upi )NF©r   r   ©r   ©r„   Ú_fileidsr{   r|   r   r   Úabspathr   ©r   ra   r€   r   Úfileids   &&&& r   ÚwordsÚPl196xCorpusReader.words²   sé   € ØŸ=™=¨¸gÓFÑˆØŠ?Ø—m‘m‰GÜ˜¤×%Ò%Ø�iˆGçÜñ #*ó
ñ #*˜ô "ØŸ™ VÓ,ØØØØ!%§¡Ø '¥÷ñ #*ñ
óð ô ñ #*ó	ñ #*˜ô "ØŸ™ VÓ,ØØØØ!%§¡÷ñ #*ñ	óð ùò
ùò	ó   Á9CÂ2Cc                ó¸  € V P                  WV4      w  rVf   V P                  pM\        V\        4      '       d   V.pV'       dJ   \	        V Uu. uF3  p\        V P                  V4      RRRV P                  W4,          R7      NK5  	  up4      # \	        V Uu. uF,  p\        V P                  V4      RRRV P                  R7      NK.  	  up4      # u upi u upi ©NFTr’   r“   r”   r—   s   &&&& r   ÚsentsÚPl196xCorpusReader.sentsÕ   så   € ØŸ=™=¨¸gÓFÑˆØŠ?Ø—m‘m‰GÜ˜¤×%Ò%Ø�iˆGçÜñ #*ó
ñ #*˜ô "ØŸ™ VÓ,ØØØØ!%§¡Ø '¥÷ñ #*ñ
óð ô ñ
 #*ó	ñ #*˜ô "ØŸ™ VÓ,¨e°T¸5È4Ï=É=÷ñ #*ñ	óð ùò
ùòr›   c                ó¸  € V P                  WV4      w  rVf   V P                  pM\        V\        4      '       d   V.pV'       dJ   \	        V Uu. uF3  p\        V P                  V4      RRRV P                  W4,          R7      NK5  	  up4      # \	        V Uu. uF,  p\        V P                  V4      RRRV P                  R7      NK.  	  up4      # u upi u upi r�   r”   r—   s   &&&& r   ÚparasÚPl196xCorpusReader.parasô   så   € ØŸ=™=¨¸gÓFÑˆØŠ?Ø—m‘m‰GÜ˜¤×%Ò%Ø�iˆGçÜñ #*ó
ñ #*˜ô "ØŸ™ VÓ,ØØØØ!%§¡Ø '¥÷ñ #*ñ
óð ô ñ
 #*ó	ñ #*˜ô "ØŸ™ VÓ,¨e°T¸4È$Ï-É-÷ñ #*ñ	óð ùò
ùòr›   c                ó¸  € V P                  WV4      w  rVf   V P                  pM\        V\        4      '       d   V.pV'       dJ   \	        V Uu. uF3  p\        V P                  V4      RRRV P                  W4,          R7      NK5  	  up4      # \	        V Uu. uF,  p\        V P                  V4      RRRV P                  R7      NK.  	  up4      # u upi u upi ©NTFr’   r“   r”   r—   s   &&&& r   Útagged_wordsÚPl196xCorpusReader.tagged_words  så   € ØŸ=™=¨¸gÓFÑˆØŠ?Ø—m‘m‰GÜ˜¤×%Ò%Ø�iˆGçÜñ #*ó
ñ #*˜ô "ØŸ™ VÓ,ØØØØ!%§¡Ø '¥÷ñ #*ñ
óð ô ñ
 #*ó	ñ #*˜ô "ØŸ™ VÓ,¨d°E¸5È4Ï=É=÷ñ #*ñ	óð ùò
ùòr›   c                ó¸  € V P                  WV4      w  rVf   V P                  pM\        V\        4      '       d   V.pV'       dJ   \	        V Uu. uF3  p\        V P                  V4      RRRV P                  W4,          R7      NK5  	  up4      # \	        V Uu. uF,  p\        V P                  V4      RRRV P                  R7      NK.  	  up4      # u upi u upi r¤   r”   r—   s   &&&& r   Útagged_sentsÚPl196xCorpusReader.tagged_sents2  så   € ØŸ=™=¨¸gÓFÑˆØŠ?Ø—m‘m‰GÜ˜¤×%Ò%Ø�iˆGçÜñ #*ó
ñ #*˜ô "ØŸ™ VÓ,ØØØØ!%§¡Ø '¥÷ñ #*ñ
óð ô ñ
 #*ó	ñ #*˜ô "ØŸ™ VÓ,¨d°D¸%È$Ï-É-÷ñ #*ñ	óð ùò
ùòr›   c                ó¸  € V P                  WV4      w  rVf   V P                  pM\        V\        4      '       d   V.pV'       dJ   \	        V Uu. uF3  p\        V P                  V4      RRRV P                  W4,          R7      NK5  	  up4      # \	        V Uu. uF,  p\        V P                  V4      RRRV P                  R7      NK.  	  up4      # u upi u upi )NTr’   r“   r”   r—   s   &&&& r   Útagged_parasÚPl196xCorpusReader.tagged_parasQ  så   € ØŸ=™=¨¸gÓFÑˆØŠ?Ø—m‘m‰GÜ˜¤×%Ò%Ø�iˆGçÜñ #*ó
ñ #*˜ô "ØŸ™ VÓ,ØØØØ!%§¡Ø '¥÷ñ #*ñ
óð ô ñ
 #*ó	ñ #*˜ô "ØŸ™ VÓ,¨d°D¸$ÈÏÉ÷ñ #*ñ	óð ùò
ùòr›   c                ó˜   € V P                  W4      w  r\        V4      ^8X  d   \        P                  ! W^ ,          4      # \	        R4      h)é   zExpected a single file)r„   r"   r   ÚxmlÚ	TypeErrorrŽ   s   &&& r   r¯   ÚPl196xCorpusReader.xmlp  s?   € Ø—]‘] 7Ó7‰
ˆÜˆw‹<˜1ÔÜ"×&Ò& t°Q­ZÓ8Ð8äÐ4Ó5Ð5r   )r\   r]   r
   rl   )NN)NNN)rH   rI   rJ   rK   r   r   rU   rd   r„   r‡   r   r™   rž   r¡   r¥   r¨   r«   r¯   rL   rM   rN   s   @r   rQ   rQ   \   sS   ø‡ € Ø€Hò	ò<ò +ô ò@ô@ô !ôFô>ô>ô>ô>÷>6ò 6r   rQ   N)Únltk.corpus.reader.apiÚnltk.corpus.reader.xmldocsr   ÚreÚcompiler'   r(   r-   r)   rC   r@   r$   r   r   rT   rQ   rp   r   r   Ú<module>r¶      s•   ðô %Ý 6à	‡z�zÐ0Ó1€Ø	‡z�zÐ0Ó1€à�Z‰ZÐ>Ó?€
Ø	‡z�zÐ6Ó7€à	‡z�z�/Ó"€Ø‡j�j�Ó €à	�‰Ð&Ó	'€ôBÐ*ô BôJY6Ð0°/ö Y6r   