-iVRdZddlZddlmZddlZdZdZdZ GddZ dS)z Arabic noun stemmer Nc$fd|DS)a@ Verify possible affixes in the resulted segments according to the given affixes list. @param word: the input word. @type word: unicode. @param list_seg: list of word segments indexes (numbers). @type list_seg: list of pairs. @return: list of acceped segments. @rtype: list of pairs. c~g|]9}dd|d|ddgv7|:S-Nrjoin.0s affix_listwords R/var/www/html/speakWrite/venv/lib/python3.11/site-packages/qalsadi/noun_affixer.py z verify_affix..!sL X X X!388T&AaD&\4!<,H#I#IZ#W#WA#W#W#Wrlist_segr s` `r verify_affixrs$ Y X X X Xx X X XXrc`|}d|vr |drdS|ddvrd|vrdSd|vrdSd|vrdS|d vr |d sdS|d vr |d sdSd |vr |dsdS|d r |dsdS|tjkr |dsdS|tjkr|ddkrdSdS)aF Test if the given word from dictionary is compabilbe with affixes tags. @param noun_tuple: the input word attributes given from dictionary. @type noun_tuple: dict. @param affix_tags: a list of tags given by affixes. @type affix_tags:list. @param proclitic_nm: first level prefix vocalized. @type proclitic_nm: unicode. @param enclitic_nm: first level suffix vocalized. @type enclitic_nm: unicode. @param suffix_nm: first level suffix vocalized. @type suffix_nm: unicode. @return: if the tags are compatible. @rtype: Boolean. تنوين mamnou3_sarfFnumber)جمعuجمع تكسيرuجمع مؤنث سالمuجمع مذكر سالمuمثنى)uهمuهنuكماuكمuهما hm_suffix)uهuها ha_suffixالk_prefix kal_prefixw_suffixrT)endswitharWAW TEH_MARBUTA) noun_tuple affix_tags proclitic_nm enclitic_nm suffix_nm proclitics r validate_tagsr+$s#"Iz!!j&@!u(>>> % 3 35 % 3 35  # #5BBB:VaKbBun$$Z -D$u Yz*'=u&!!*\*BuBF:j#9uBN""z(';x'G'Gu 4rchtj|d}tj|d}tj|d}|tj|ddz }d|vr d|vrd|vrdSd|vrd|vrdSd|vrd|vrdSd|vrd|vrdSd|vrd |vrdSd |vrd |vrdSd S) ag Verify if proaffixes (sytaxic affixes) are compatable with affixes ( conjugation) @param proclitic_nm: first level prefix. @type proclitic_nm: unicode. @param enclitic: first level suffix. @type enclitic: unicode. @param suffix: second level suffix. @type suffix: unicode. @return: compatible. @rtype: True/False. tagscasesr تعريفuمضافFru إضافةu لايضافuجرu مجرورT)SNCCOMP_PREFIX_LIST_TAGSCOMP_SUFFIX_LIST_TAGSCONJ_SUFFIX_LIST_TAGSget)r'encliticsuffixproclitic_tags enclitic_tags suffix_tagss rcheck_clitic_affixr:[sN.| ! !l+&E&EutrceZdZdZedZed dZedZedZdZ edZ ed Z ed Z d S) muwaledcdS)Nr)selfs r__initzmuwaled.__inits  rct|g}|tjtjztjtjztjtjtjztjzfvr$|tjz}|||r4|tjtjztjtjzfvr$|tjz}||| tjr,|ddtj z}||| tj r,|ddtj z}|||}|S)u Generate the Noun stem variants according to the affixes. For example مدرستي = >مدرست+ي = > مدرسة +ي. Return a list of possible cases. @param stem: the input stem. @type stem: unicode. @param suffix_nm: suffix (no mark). @type suffix_nm: unicode. @return: list of stem variants. @rtype: list of unicode. N) setr"ALEFTEHYEHr$addNOONr#r! ALEF_MAKSURAHAMZA YEH_HAMZA)stemr)possible_noun_list possible_nounvalidated_lists rget_stem_variantszmuwaled.get_stem_variantssS"!     Gbf  FR^ # F FRW rv %    !2>1M  " "= 1 1 1 2I"&27*:BFRW>II (f!5!5 (BCCyRVRW---  (K  34V|dkr|ddtjz}n|dkr|ddtjz}n|tj tjzr|d|vr&d|vr"|s |stj|dd}n|tjtj ztj tj zfvrtj|dd}n.|tj r|s|r | tjr|ddtjz}n| tj r|ddtjz}n|tjtj zs9|tjtjztj zr6| tjr|ddtjz}|S)u Get the word variant to be joined to the suffix. For example: word = مدرسة, suffix = ي. The word is converted to مدرست. @param word: word found in dictionary. @type word: unicode. @param proclitic: proclitic ( first level). @type proclitic: unicode. @param suffix: suffix ( first level). @type suffix: unicode. @param enclitic: enclitic( second level). @type enclitic: unicode. @return: variant of word. @rtype: unicode. NrArQruلل)r"strip_tashkeelrWr!r$rCrErDrHr`rGr#rIr]r_ WAW_HAMZArJSUKUNFATHATAN)rr*r6r5 word_stemr)r(long_suffix_nms rget_word_variantzmuwaled.get_word_variant s  %f-- '11 "[0' 22    bn - -@ :!!"'BN":;; 4FFR^+FRW$rv-! !*#2# 7II"'BF"222 " 3IcrcN C CII#r)) )#2# 7I"& ' 26) / #2#??  2%%%crcNRV3    0 0$ :B%crcNRV3 ""%crcNRW4   26 1 2 2 : i'')++#,!,/ #2#?? rv/"'1ABBB/ #2#??    ) ) :y :K :  ** :%crcNR\9 ""28,, :%crcNR\9 ""26BH#455 :%%bfrx&7"(&BCC :##BK00 :&crcNR\9 rctj|dd}tj|dd}|||\}}|}t j|ddr |dd}|tjtj}dtj|dvrst j |drYd |dtj |ddg}| tj r |dd}|tjtjtjz}|tjtjtjz}t!jd tjztj|}t!jd tjzd|}| tjtjzrd nd } |||||}||||| \}} d ||| |g} d ||||g} d ||||g} t j| } tjD]1\}}| ||} | ||} 2| | | fS)a Join the noun and its affixes, and get the vocalized form @param noun: noun found in dictionary. @type noun: unicode. @param proclitic: first level prefix. @type proclitic: unicode. @param suffix: second level suffix. @type suffix: unicode. @param enclitic: first level suffix. @type enclitic: unicode. @return: vocalized word. @rtype: unicode. vocalizedrrANr/r-rQrz(%s)+z^(%s)+TFr)r0r1r2rer"rUrarkFATHAis_sunr SHADDAr!rjrCrHrSrTr`rErnr[rhAJUST_VOCAL_PATTERNS)r>nounr*r6r5 proclitic_vocrbencl_voc_non_inflectrcrXrZword_non_irab_markword_vocalized segmentedpatrnrepls rvocalizezmuwaled.vocalize}s 1)<[I!L 0:;GJ -1-F-F &. . * *  <RSS " " 9D||BK22 34Y?G G GBI GM M G77DGRYQRR9::D%%bh// 3 -crc 2 ||BGRX%788||BORX-GHHvg("(D99vh)2t44--26(9::E$$$$T9fhGG ,0+C+C *h, , ( ( WW D"68L M   -z.sS   388T&AaD&\4!<,H#I#IZ#W#WA#W#W#Wrrrs` `rrzmuwaled.verify_affixs4         rN)F) __name__ __module__ __qualname___muwaled__init staticmethodrOr[rernr}rrrrrrr<r<s   ,,\,\&/&/&/\&/P55\5*ZZ\Zx[=[=[=z\\$   \    rr<) __doc__rSpyarabic.arabyarabyr"stem_noun_constr0rr+r:r<rrrrs  Y Y Y444nNNNbY Y Y Y Y Y Y Y Y Y r