AeP?fddlmZddlZ ddlmZddlmZdZ dZ d Z e d Z d Zd Zd ZdZdZGddeZdZGddeZGddeZeZGddeZGddeZGddeZGddeZGdd eZeZd!e_d"e_ d#Z!d$Z"d%Z#d&Z$e$d'Z%d(e%_d)e%_ d5d*Z&d+Z'd,Z(d-Z)d.Z*ee Z+d/e+_d0e+_ ee Z,d1e,_d2e,_ ee Z-d3e-_d4e-_ y#e$r ddlmZYwxYw)6)absolute_importN)maxsize)maxint)Errorsboleoleof cRt|}|jd}t|}g}||krwt||}|dz}|dz }||kr2|t||k\r!|dz }|dz }||kr|t||k\r!|j ||j |||krw|S)z Return a list of character codes consisting of pairs [code1a, code1b, code2a, code2b,...] which cover all the characters in |s|. rr)listsortlenordappend)s char_listinresultcode1code2s 3lib/python3.12/site-packages/Cython/Plex/Regexps.pychars_to_rangesr"s QI NN A IA F a%IaL!  Q!eYq\!22 QJE FA!eYq\!22  e e a% Mct|td}t|tddz}||kr!tdtdz }||z||zfSy)z If the range of characters from code1 to code2-1 includes any lower case letters, return the corresponding upper case range. azrANmaxrminrrcode3code4ds ruppercase_ranger'9W s3x E s3x!| $E u} Hs3x  519%%rct|td}t|tddz}||kr!tdtdz }||z||zfSy)z If the range of characters from code1 to code2-1 includes any upper case letters, return the corresponding lower case range. rZrrNr r#s rlowercase_ranger+Gr(rc tdt|dDcgc]}t||||dz}}t|Scc}w)z Given a list of codes as returned by chars_to_ranges, return an RE which will match a character in any of the ranges. rr)ranger CodeRangeAlt) code_listrre_lists r CodeRangesr3UsK CH3y>[\B]^Qy1yQ'78^G^ =_s?c|tcxkr|kr3nn0tt|ttttdz|St||S)z CodeRange(code1, code2) is an RE which matches any character with a code |c| in the range |code1| <= |c| < |code2|. r)nl_coder0 RawCodeRange RawNewlinerrs rr/r/^sG  %<w/! U35 5E5))rcReZdZdZdZdZdZdZdZdZ dZ dZ d Z d Z d Zd Zy) REaRE is the base class for regular expression constructors. The following operators are defined on REs: re1 + re2 is an RE which matches |re1| followed by |re2| re1 | re2 is an RE which matches either |re1| or |re2| rNcFtd|jjz)aM This method should add states to |machine| to implement this RE, starting at |initial_state| and ending at |final_state|. If |match_bol| is true, the RE must be able to match at the beginning of a line. If nocase is true, upper and lower case letters should be treated as equivalent. z %s.build_machine not implemented)NotImplementedError __class____name__)selfmachine initial_state final_state match_bolnocases r build_machinezRE.build_machine{s&""D"&.."9"9#:; ;rcl|j}|j||j|||S)z~ Given a state |s| of machine |m|, return a new state reachable from |s| on character |c| or epsilon. ) new_statelink_toadd_transition)r?mrAcrs r build_optz RE.build_opts2 KKMa $$Q*rct||SN)Seqr?others r__add__z RE.__add__4rct||SrN)r0rPs r__or__z RE.__or__rSrcR|jr |jS|jSrN)strcalc_strr?s r__str__z RE.__str__s 8888O==? "rcLt|ts|j||dyy)NzPlex.RE instance) isinstancer: wrong_typer?numvalues rcheck_rez RE.check_res"%$ OOC(: ;%rcZt|tdk7r|j||dyy)Nstring)typer]r^s r check_stringzRE.check_strings' ;$r( " OOC 1 #rc|j||t|dk7r8tjd||jj t |fzy)NrzOInvalid value for argument %d of Plex.%s.Expected a string of length 1, got: %s)rfrrPlexValueErrorr=r>reprr^s r check_charz RE.check_chars_ #u% u:?'')Q,/1H1H$u+TW)WX X rc&t|tjk(r1|jjd|jj d}nt|j }t jd||jj ||fz)N.z instancezr PlexTypeError)r?r_r`expectedgots rr]z RE.wrong_types ;%,, ,**EOO,D,DFCu+&&C""$:'*DNN,C,CXs=T$TU Ur)r>ro __qualname____doc__nullablematch_nlrWrErLrRrUrZrarfrjr]rrr:r:osGHH C ;  # <2XUrr:ct|dk(r"tt|t|dz}n t|}dt |z|_|S)z; Char(c) is an RE which matches the character |c|. rzChar(%s))rr/r SpecialSymbolrirW)rKrs rCharrzsF 1v{3q63q6A:.q!d1g%FJ Mrc6eZdZdZdZdZdZdZdZdZ dZ dZ y)r6z RawCodeRange(code1, code2) is a low-level RE which matches any character with a code |c| in the range |code1| <= |c| < |code2|, where the range does not include newline. For internal use only. rNcZ||f|_t|||_t|||_yrN)r.r'r+)r?rrs r__init__zRawCodeRange.__init__s+U^ .ue<.ue<rc|r|j||t}|j|j||rR|jr|j|j||j r|j|j |yyyrN)rLBOLrIr.r'r+r?rJrArBrCrDs rrEzRawCodeRange.build_machinesv  NN1mSAM$$TZZ= ##,,T-A-A;O##,,T-A-A;O$ rc8d|j|jfzS)NzCodeRange(%d,%d)r8rYs rrXzRawCodeRange.calc_strs!TZZ$<<rorsrtrurvr.r'r+r}rErXrwrrr6r6s2 HH EOO= P=rr6ceZdZdZdZdZdZy) _RawNewlinezd RawNewline is a low-level RE which matches a newline character. For internal use only. rrc|r|j||t}|j||t}|jttdzf|yNr)rLrEOLrIr5)r?rJrArBrCrDrs rrEz_RawNewline.build_machinesB  NN1mSAM NN1mS 1 '7Q;/=rN)r>rorsrtrurvrErwrrrrsHH>rrc(eZdZdZdZdZdZdZdZy)ryzx SpecialSymbol(sym) is an RE which matches the special input symbol |sym|, which is one of BOL, EOL or EOF. rNc||_yrN)sym)r?rs rr}zSpecialSymbol.__init__#s rc|r*|jtk(r|j||t}|j |j|yrN)rrrLrrIrs rrEzSpecialSymbol.build_machine&s7 S NN1mSAM$$TXX{;r) r>rorsrtrurvrr}rErwrrryrys"HH CKKFI,E"++  GrcZddjtt|jzS)NzSeq(%s),joinmaprWr2rYs rrXz Seq.calc_strU!388CT\\$:;;;rNr>rorsrtr}rErXrwrrrOrO.s"!&G rorsrtrrDr}rErXrwrrrrs" B F$ +*rrz8 Empty is an RE which matches the empty string. Emptycjtttt|}dt |z|_|S)z@ Str1(s) is an RE which matches the literal string |s|. Str(%s))rOtuplerrzrirWrrs rStr1rs.%D! % &FT!W$FJ Mrct|dk(rt|dSttt t|}ddj t t |z|_|S)z Str(s) is an RE which matches the literal string |s|. Str(s1, s2, s3, ...) is an RE which matches any of |s1| or |s2| or |s3|... rrrr)rrr0rrrrirW)strsrs rStrrsR  4yA~DG}eCdO,-#dD/!::  rcTtt|}dt|z|_|S)zH Any(s) is an RE which matches any character in the string |s|. zAny(%s))r3rrirWrs rAnyrs( * +FT!W$FJ Mrct|}|jdt |jtt |}dt |z|_|S)zp AnyBut(s) is an RE which matches any character (including newline) which is not in the string |s|. rz AnyBut(%s))rinsertrrr3rirW)rrangesrs rAnyButrsI Q F MM!fW MM&  FQ'FJ MrrczT AnyChar is an RE which matches any single character (including a newline). AnyCharc N|r1tt|t|dz}d|d|d|_|Sg}tdt |dD];}|j tt||t||dzdz=t |}dt|z|_|S)a Range(c1, c2) is an RE which matches any single character in the range |c1| to |c2| inclusive. Range(s) where |s| is a string of even length is an RE which matches any single character in the ranges |s[0]| to |s[1]|, |s[2]| to |s[3]|,... rzRange(rrrr-z Range(%s))r/rrWr.rrr0ri)rrrrrs rRangers 3r7CGaK0')2.  M q#b'1% EA MM)C1JBq1uI0BC D Ef 48+ Mrc:t|t}d|z|_|S)zI Opt(re) is an RE which matches either |re| or the empty string. zOpt(%s))r0rrWrrs rOptrsU^FRFJ MrcBtt|}d|z|_|S)zJ Rep(re) is an RE which matches zero or more repetitions of |re|. zRep(%s))rrrWrs rReprs!b]FRFJ Mrct|dS)z NoCase(re) is an RE which matches the same strings as RE, but treating upper and lower case letters as equivalent. rrDrrs rrrs b ##rct|dS)z Case(re) is an RE which matches the same strings as RE, but treating upper and lower case letters as distinct, i.e. it cancels the effect of any enclosing NoCase(). rrrrs rrr s b ##rz= Bol is an RE which matches the beginning of a line. Bolz7 Eol is an RE which matches the end of a line. Eolz9 Eof is an RE which matches the end of the file. EofrN). __future__rrmsysrr ImportErrorrcrrrEOFrr5rr'r+r3r/objectr:rzr6rr7ryrOr0rrrrtrWrrrrrrrrrrrrrrwrrrs' %     d).   *"EUEU\ =2=> >" >]