S"jp2dZddlmZddlmZddlmZddlmZm Z m Z ddl m Z m Z mZmZddlmZdd lmZmZmZmZmZmZmZmZmZmZmZmZmZGd d eZ d e!d e e!e effdZ"ededededededededededededi Z#de!deed e!fdZ$y)z( Simple HTML -> Telegram entity parser. )deque)escape) HTMLParser)IterableTupleList) add_surrogate del_surrogatewithin_surrogate strip_text)TLObject) MessageEntityBoldMessageEntityItalicMessageEntityCodeMessageEntityPreMessageEntityEmailMessageEntityUrlMessageEntityTextUrlMessageEntityMentionNameMessageEntityUnderlineMessageEntityStrikeMessageEntityBlockquoteMessageEntityCustomEmojiTypeMessageEntityc0eZdZfdZdZdZdZxZS)HTMLToTelegramParserct|d|_g|_i|_t |_t |_y)N)super__init__textentities_building_entitiesr _open_tags_open_tags_meta)self __class__s Y/var/www/api.ozpay.ru/ozpay/venv/lib/python3.12/site-packages/telethon/extensions/html.pyr!zHTMLToTelegramParser.__init__s7   "$'$wc2|jj||jjdt|}d}i}|dk(s|dk(rt}n:|dk(s|dk(rt }n(|dk(rt }n|dk(s|dk(rt}n |dk(rt}d |vrd |d <n|d k(r( |jd } |dtdd|_ n|d k(r t}d|d<n|dk(r |d}|jdr|tdd}t }n0|j#|k(rt$}nt&}t)||d<d}|jj+|jj|n|dk(r t-|d}t0}||d<|r;||jvr,|dt|j2dd||j|<yyy#t$rYwxYw#t$r t}Y_wxYw#t$rYywxYw#tt.f$rYywxYw)Nstrongbemiudels blockquote expandableT collapsedcodepreclassz language-rlanguageahrefzmailto:urlztg-emojizemoji-id document_idr)offsetlength)r% appendleftr&dictrrrrrr$lenr9KeyErrorrr startswithrget_starttag_textrrr popleftint ValueErrorrr")r'tagattrs EntityTypeargsr7r<emoji_ids r)handle_starttagz$HTMLToTelegramParser.handle_starttagsE ""3' ''-U   (?cSj*J D[C3J,J CZ/J E\SCZ,J L 0Ju$$([! F] /--e4#(>#k2B2C#DCL E\)J!D  CZ Fm~~i(#i./*/ ))+s2!1J!5J"/"4DKC  ( ( *  + +C 0 J  uZ012J"*D  #T%<%<<+5,499~, ,D # #C (=:C  /.  /  "j)  sT$G 4G"G59H GG GG G21G25 HHHHc0t|jdkDr|jdnd}|dk(r|jd}|r|}|jj D]#\}}|xj t|z c_%|xj |z c_y)Nrrr:)rCr%r&r$itemsr?r")r'r" previous_tagr<rJentitys r) handle_dataz HTMLToTelegramParser.handle_datacs-0-AA-Etq)2 3 &&q)C2288: 'KC MMSY &M ' T r*c |jj|jj|jj |d}|r|j j|yy#t$rYFwxYwN)r%rGr& IndexErrorr$popr#append)r'rJrSs r) handle_endtagz"HTMLToTelegramParser.handle_endtagosq  OO # # %  ( ( *((,,S$7  MM  (    s4A11 A=<A=)__name__ __module__ __qualname__r!rOrTrZ __classcell__)r(s@r)rrs'DL )r*rhtmlreturnc2|s|gfSt}|jt|t|j|j }|j j |j jdt||j fS)a Parses the given HTML message and returns its stripped representation plus a list of the MessageEntity's that were found. :param html: the message with HTML to be parsed. :return: a tuple consisting of (clean message, [message entities]). c|jSrV)r>)rSs r)zparse..s FMMr*key) rfeedr r r"r#reversesortr )r_parserr"s r)parserjzsw Rx ! #F KK d#$ fkk6?? 3D OO OO9:   //r*)zz )zz)zz)zz)zzc*|jrddfSddfS)Nz
z
z
)r5e_s r)rcrcs#%&[[!+6D+r*c<dj|jdfS)Nz-
    
        z{}
)formatr9rls r)rcrcs! 6!**%($r*c(dj|dfS)Nzrprnts r)rcrcs&<&C&CA&F%Or*c(dj|dfSNz rrrsrts r)rcrcsO$:$:1$=v#Fr*cNdjt|jdfSrw)rprr<rls r)rcrcs(>(>vaee}(Mv'Vr*c<dj|jdfS)Nzrr)rpuser_idrls r)rcrcs,H,O,OPQPYPY,Z\b+cr*c<dj|jdfS)Nzz )rpr=rls r)rcrcs,F,M,Mamm,\^k+lr*r"r#c|s|S|s t|St|tr|f}t|}g}t |D]\}}|j }|j |j z}tjt|d}|sMt|r |||||}|j|||df|j|| |df|jdt|}|rT|j\} } } t|| r| dz } t|| r|d| | zt|| |z||dz}| }|rTt|d|||dz}t!|S)a= Performs the reverse operation to .parse(), effectively returning HTML given a normal text and its MessageEntity's. :param text: the text to be reconverted into HTML. :param entities: the MessageEntity's applied to the text. :return: a HTML representation of the combination of both inputs. Nrc|d|dfS)Nrr}r@)rus r)rczunparse..s!A$!r*rd)r isinstancerr enumerater>r?ENTITY_TO_FORMATTERgettypecallablerYrhrCrXr r ) r"r# insert_atr/rSr2rm delimiternext_escape_boundatrnwhats r)unparsers  d|(H%;  DIx(4 6 MM MMFMM )'++DL$?  "%fd1Qi8   aIaL1 2   a!Yq\2 34NN-N.D  mmo AttR( !GBtR(CRy4&b1B)C"DDtL]L^G__  $))* +d3D3E.F FD  r*N)%__doc__ collectionsrr_r html.parserrtypingrrrhelpersr r r r tlrtl.typesrrrrrrrrrrrrrrstrrjrrr@r*r)rs"((PPc):c)L00c4(9#::;0(0*,O,OFVcl+2*#**;!<**r*