o
    OZh&                    @   sH
  d dl Z d dlZd dlZd dlmZ d dlmZ g dZG dd deZ	G dd deZ
G dd	 d	eZG d
d deZG dd de jZe ej eZeeB ZeeB eB Zed eeiZeeB eB eB eB e B Z!ee"B e#B e$B e%B e&B eB Z'e(ej)Z*e(ej+Z,e*e,B Z-e(ej.Z/e(ej0Z1e(de(dgB Z2e-e(dB Z3e-e(dB Z4dZ5e6 Z7e7d Z8de8> d Z9eeeee"ee#ee e$eeee%e&dZ:ee"B Z;d Z<ee"B Z=ee=B Z>d d ed e"e"e=e=iZ?ddddZ@dZAG dd dZBeB ZCeDeAE D ]
\ZFZGeHeCeGeF qdEddZIdd  ZJd!d" ZKd#d$ ZLd%d& ZMd'd( ZNd)d* ZOd+d, ZPdFd.d/ZQd0d1 ZRd2d3 ZSd4d5 ZTd6d7 ZUd8d9 ZVd:d; ZWd<d= ZXd>d?d@dAZYdBdC ZZdDdE Z[dFdG Z\dHdI Z]dJdK Z^dLdM Z_dNdO Z`dPdQ ZadRdS ZbdTdU ZcdVdW ZddXdY ZedZd[ Zfd\d] Zgd^d_ Zhd`da Zidbdc Zjddde Zkdfdg Zldhdi Zmdjdk Zndldm Zodndo Zpdpdq Zqdrds Zrdtdu Zsdvdw Ztdxdy ZudGdzd{Zvd|d} Zwd~d Zxdd Zydd Zzdd Z{dd Z|dd Z}dd Z~dd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd ZedE ZedE ZdHddZdd Zdd Zdd Zdd ZdZdZdZdZdZdZdddZe<de"dede=diZdd ZG dd dZG ddĄ deZG ddƄ deZG ddȄ deZG ddʄ deZG dd̄ deZG dd΄ deZG ddЄ deZG dd҄ deZG ddԄ deZG ddք deZG dd؄ deZG ddڄ deZG dd܄ deZG ddބ deZG dd deZG dd deZG dd deZG dd deZG dd deZG dd deZG dd deZG dd deZG dd deZG dd dZG dd deZG dd deZG dd deZG dd deZG dd deZG dd deZG dd  d eZG dd deZG dd deZG dd deZG dd deZG d	d
 d
eZG dd deZG dd deZG dd deZG dd deǃZG dd deǃZG dd deǃZG dd deǃZG dd deZG dd deZG dd de̓ZG dd  d eZG d!d" d"eZG d#d$ d$eZG d%d& d&eуZG d'd( d(eZG d)d* d*ZG d+d, d,ZՐd-d. Z֐d/d0 ZG d1d2 d2Ze١ Zi Zeڠܡ D ]7\Z\ZZe۠edi f\ZZeeeed3Zeefee< eߠܡ D ]\ZZeeedeed3ee< qqsd4d5d6d7d8d9d:d;Zedd<d=edd<d-edd>d=edd?d=edd?d-edd@d=edd@d-dAZeσ e ed-e eЃ e e dBZeeZee ed-e e dC e e e ẽ dDZdS (I      N)defaultdict)%AASCIIB	BESTMATCHDDEBUGEENHANCEMATCHFFULLCASEI
IGNORECASELLOCALEM	MULTILINEPPOSIXRREVERSESDOTALLTTEMPLATEUUNICODEV0VERSION0V1VERSION1WWORDXVERBOSEerrorScanner	RegexFlagc                   @   s   e Zd ZdZdddZdS )r%   a  Exception raised for invalid regular expressions.

    Attributes:

        msg: The unformatted error message
        pattern: The regular expression pattern
        pos: The position in the pattern where compilation failed, or None
        lineno: The line number where compilation failed, unless pos is None
        colno: The column number where compilation failed, unless pos is None
    Nc                 C   s   t |trdnd}|| _|| _|| _|d urB|d urB||d|d | _|||d| | _d	||}||v rB|d	| j| j7 }t
| | d S )N
   
r      z{} at position {}z (line {}, column {}))
isinstancestrmsgpatternposcountlinenorfindcolnoformat	Exception__init__)selfmessager.   r/   newline r:   @/var/www/auris/lib/python3.10/site-packages/regex/_regex_core.pyr6   *   s   
zerror.__init__)NN)__name__
__module____qualname____doc__r6   r:   r:   r:   r;   r%      s    r%   c                   @      e Zd ZdS )_UnscopedFlagSetNr<   r=   r>   r:   r:   r:   r;   rA   =       rA   c                   @   r@   )
ParseErrorNrB   r:   r:   r:   r;   rD   A   rC   rD   c                   @   r@   )_FirstSetErrorNrB   r:   r:   r:   r;   rE   E   rC   rE   c                   @   s   e Zd Zd ZZd ZZd ZZd Z	Z
d ZZd ZZd ZZd ZZd	 ZZd
 ZZd ZZd ZZd ZZd ZZd ZZ d Z!Z"d Z#Z$dd Z%e&j'Z'dS )r'      i   i   i   i @           i   i          i       i   @   r*   c                 C   s   | j d ur
d| j  S | j}g }|dk }|r| }| jD ]}||j@ r0||j M }|d|j   q|r:|t| d|}|rQt|dkrMd| }|S d| }|S )Nzregex.%sr   |r*   z~(%s)z~%s)_name__value_	__class__appendhexjoinlen)r7   valuemembersnegativemresr:   r:   r;   __repr__]   s*   




zRegexFlag.__repr__N)(r<   r=   r>   r   r   r   r   r   r   r	   r
   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r    r!   r"   r#   r$   r   r   r[   object__str__r:   r:   r:   r;   r'   I   s(    
r'   z()|?*+{^$.[\# z -z &_-.)||~~&&--rI   r*   )abefir   rY   prsur   r   wxrG   rH   )rm   rk   r   a  
FAILURE
SUCCESS
ANY
ANY_ALL
ANY_ALL_REV
ANY_REV
ANY_U
ANY_U_REV
ATOMIC
BOUNDARY
BRANCH
CALL_REF
CHARACTER
CHARACTER_IGN
CHARACTER_IGN_REV
CHARACTER_REV
CONDITIONAL
DEFAULT_BOUNDARY
DEFAULT_END_OF_WORD
DEFAULT_START_OF_WORD
END
END_OF_LINE
END_OF_LINE_U
END_OF_STRING
END_OF_STRING_LINE
END_OF_STRING_LINE_U
END_OF_WORD
FUZZY
GRAPHEME_BOUNDARY
GREEDY_REPEAT
GROUP
GROUP_CALL
GROUP_EXISTS
KEEP
LAZY_REPEAT
LOOKAROUND
NEXT
PROPERTY
PROPERTY_IGN
PROPERTY_IGN_REV
PROPERTY_REV
PRUNE
RANGE
RANGE_IGN
RANGE_IGN_REV
RANGE_REV
REF_GROUP
REF_GROUP_FLD
REF_GROUP_FLD_REV
REF_GROUP_IGN
REF_GROUP_IGN_REV
REF_GROUP_REV
SEARCH_ANCHOR
SET_DIFF
SET_DIFF_IGN
SET_DIFF_IGN_REV
SET_DIFF_REV
SET_INTER
SET_INTER_IGN
SET_INTER_IGN_REV
SET_INTER_REV
SET_SYM_DIFF
SET_SYM_DIFF_IGN
SET_SYM_DIFF_IGN_REV
SET_SYM_DIFF_REV
SET_UNION
SET_UNION_IGN
SET_UNION_IGN_REV
SET_UNION_REV
SKIP
START_OF_LINE
START_OF_LINE_U
START_OF_STRING
START_OF_WORD
STRING
STRING_FLD
STRING_FLD_REV
STRING_IGN
STRING_IGN_REV
STRING_REV
FUZZY_EXT
c                   @   r@   )	NamespaceNrB   r:   r:   r:   r;   rn     rC   rn      c              	   C   s   t |  }t|| }|dk rdS || | }ddl}t|ds#dS |||D ]}	z| |	= W q) ty9   Y q)w |  i }
t | D ]&\}}}}}}|||||||f< z|||f |
||f< W qD tyj   Y qDw |  ||
 dS )aC  Make room in the given cache.

    Args:
        cache_dict: The cache dictionary to modify.
        args_dict: The dictionary of named list args used by patterns.
        max_length: Maximum # of entries in cache_dict before it is shrunk.
        divisor: Cache will shrink to max_length - 1/divisor*max_length items.
    r   Nsample)	tuplekeysrU   randomhasattrrp   KeyErrorclearupdate)Z
cache_dictZ	args_dictZlocale_sensitive
max_lengthZdivisorZ
cache_keysZoverageZnumber_to_tossrs   Z
doomed_keyZsensitivity_dictr.   Zpattern_typeflagsargsZdefault_versionlocaler:   r:   r;   _shrink_cache  s2   

r|   c                 C   s(   | j }|t@ dkr|| jO }t||S )zFolds the case of a string.r   )ry   _ALL_ENCODINGSguess_encoding_regex	fold_case)infostringry   r:   r:   r;   
_fold_case?  s   
r   c                 C   s   t t| j|dkS z$Checks whether a character is cased.r*   )rU   r   get_all_casesry   )r   charr:   r:   r;   
is_cased_iG  s   r   c                 C   s   t t| |dkS r   )rU   r   r   )ry   r   r:   r:   r;   
is_cased_fK     r   c                 C   s,   t | jt@ }t| ||}|sg S ||S )z&Compiles the firstset for the pattern.)boolry   r   _check_firstsetcompile)r   fsreverser:   r:   r;   _compile_firstsetO  s
   
r   c                 C   s   |rd|v rdS t  }t}|D ]}t|tr|js dS ||jO }||jtd q|tt	B kr3dS t
| t||t @ dd}|j| |dd}|S )z$Checks the firstset for the pattern.N
case_flagsTr   	zerowidthin_set)setNOCASEr+   	Characterpositiver   add
with_flagsr   r   SetUnionlistoptimise)r   r   r   rW   r   rg   r:   r:   r;   r   Y  s    
r   c                 C   s   g }| D ]}| | q|S )z(Flattens the code from a list of tuples.)extend)codeZ	flat_codecr:   r:   r;   _flatten_codex  s   r   c                 C   s"   | j t@ }| j t@ r|t M }|S )zMakes the case flags.)ry   
CASE_FLAGSr   r   )r   ry   r:   r:   r;   make_case_flags  s   


r   Fc                 C   s   |rt |S t |t| dS )zMakes a character literal.r   )r   r   )r   rV   r   r:   r:   r;   make_character  s   r   c                 C   s   t | ||t| dS )zMakes a group reference.r   )RefGroupr   )r   namepositionr:   r:   r;   make_ref_group  r   r   c                 C   s   t | |t| dS )zMakes a string set.r   )	StringSetr   )r   r   r:   r:   r;   make_string_set     r   c                 C   s   |r|S |j t| dS )zMakes a property.r   )r   r   )r   propr   r:   r:   r;   make_property  s   r   c                 C   sL   t | |g}| dr|t | | | dst|dkr"|d S t|S )zParses a pattern, eg. 'a|b|c'.rN   r*   r   )parse_sequencematchrR   rU   Branch)sourcer   branchesr:   r:   r;   _parse_pattern  s   

r   c           	      C   sR  dg}t |}	 | j}|  }|tv r|dv r|| _n|dkr+|t| |d n|dkrCt| |}|du r=t |}n|| n|dkrf|jt@ rS|t	  n|jt
@ r_|t  n|t  n|dkrs|t| | n|d	kr|jt@ r|jt
@ r|t  n|t  n|t  n|d
kr|jt@ r|jt
@ r|t  nr|t  nk|jt
@ r|t  n_|t  nX|dv rt| ||}|rt| |||||| |d n;t| |||}|rt| ||||| |d n#|tt||d n|tt||d n|tt||d qdd |D }t|S )zParses a sequence, eg. 'abc'.NTz)|\F(.[^$z?*+{r   c                 S   s   g | ]}|d ur|qS Nr:   ).0itemr:   r:   r;   
<listcomp>      z"parse_sequence.<locals>.<listcomp>)r   r/   getSPECIAL_CHARSrR   parse_escapeparse_parenry   r   AnyAllr"   AnyUAny	parse_setr   StartOfLineUStartOfLineStartOfString
EndOfLineU	EndOfLineEndOfStringLineUEndOfStringLineparse_quantifierapply_quantifierparse_fuzzyapply_constraintr   ordSequence)	r   r   sequencer   	saved_poschelementcountsconstraintsr:   r:   r;   r     sv   












Mr   c                 C   s   |  }|d u r|rtd| j|td| j|t|tttfr'td| j||\}}	| j}|  }|dkr9t}
n|dkr@t}
n|| _t}
|	 sW|dksQ|	dkrW|
|||	}|
| d S )Nzmultiple repeatznothing to repeat?+r*   )popr%   r   r+   GreedyRepeat
LazyRepeatPossessiveRepeatr/   r   is_emptyrR   )r   r   r   r   r   r   r   r   	min_count	max_countZrepeatedr:   r:   r;   r     s&   r   c                 C   sX   |  }|d u rtd| j|t|tr"t|j||_|| d S |t|| d S )Nznothing for fuzzy constraint)r   r%   r   r+   GroupFuzzy
subpatternrR   )r   r   r   r   r   r   r   r:   r:   r;   r     s   
r   )r   r*   r   Nr*   N)r   *r   c                 C   s.   t |}|r	|S |dkrt| }|r|S dS )zParses a quantifier.{N)_QUANTIFIERSr   parse_limited_quantifier)r   r   r   qr   r:   r:   r;   r   /  s   
r   c                 C   s   | duo| t kS )z,Checks whether a count is above the maximum.N	UNLIMITEDr0   r:   r:   r;   is_above_limit>  s   r   c                 C   s   | j }t| }| drt| }t|pd}|rt|nd}n|s&|| _ dS t| }}| ds6|| _ dS t|s>t|rEtd| j||durT||krTtd| j|||fS )zParses a limited quantifier.,r   N}zrepeat count too bigz"min repeat greater than max repeat)r/   parse_countr   intr   r%   r   )r   r   r   r   r:   r:   r;   r   B  s(   

r   c                 C   s   | j }|dkr	dS i }zt| | | dr t| | | dsW n ty.   || _ Y dS w | dr<t| |||d< | dsItd| j| j |S )z#Parses a fuzzy setting, if present.r   Nr   :testr   z
expected })r/   parse_fuzzy_itemr   rD   parse_fuzzy_testr%   r   )r   r   r   r   r   r   r:   r:   r;   r   `  s&   





r   c                 C   s<   | j }zt| | W dS  ty   || _ t| | Y dS w )zParses a fuzzy setting item.N)r/   parse_cost_constraintrD   parse_cost_equation)r   r   r   r:   r:   r;   r   x  s   r   c           
      C   s6  | j }|  }|tv r?t| ||}t| }|du rd||< dS | j }t| }|s,|d8 }|dk r7td| j|d|f||< dS |tv r|| _ | j }t| }t| }	|	du rXt	 t| ||  }t| }|du rkt	 | j }t| }|	sx|d7 }|s~|d8 }d|  kr|ksn td| j|||f||< dS t	 )zParses a cost constraint.Nr   r*   r   bad fuzzy cost limit)
r/   r   ALPHAparse_constraintparse_fuzzy_compareparse_cost_limitr%   r   DIGITSrD   )
r   r   r   r   
constraintmax_inccost_posmax_costZmin_costZmin_incr:   r:   r;   r     sF   r   c                 C   s:   | j }t| }zt|W S  ty   Y nw td| j|)zParses a cost limit.r   )r/   r   r   
ValueErrorr%   r   )r   r  digitsr:   r:   r;   r     s   
r   c                 C   s    |dvrt  ||v rt  |S )zParses a constraint.Zdeis)rD   )r   r   r   r:   r:   r;   r     s
   r   c                 C   s    |  drdS |  drdS dS )zParses a cost comparator.z<=T<FNr   r   r:   r:   r;   r     s
   

r   c                 C   s   d|v rt d| j| ji }t| | | dr"t| | | dst| }|du r-t tt| }|s9|d8 }|dk rEt d| j| j||d< ||d< dS )	zParses a cost equation.costzmore than one cost equationr   Nr*   r   r   max)	r%   r   r/   parse_cost_termr   r   rD   r   r   )r   r   r  r  r  r:   r:   r;   r     s"   



r   c                 C   sJ   t | }|  }|dvrt ||v rtd| j| jt|pd||< dS )zParses a cost equation term.diszrepeated fuzzy costr*   N)r   r   rD   r%   r   r/   r   )r   r  coeffr   r:   r:   r;   r    s   r  c                 C   s   | j }|  }|tv r<|dkrt| |dS |dkr,|jt@ r!t S |jt@ r)t S t	 S |dkr5t
| |S td| j||rFtt||dS td| j|)Nr   Fr   r   zexpected character setr   )r/   r   r   r   ry   r   r   r"   r   r   r   r%   r   r   r   )r   r   r   r   r   r:   r:   r;   r     s"   


r   c                 C   s
   |  tS )z0Parses a quantifier's count, which can be empty.)	get_whiler  r
  r:   r:   r;   r     s   
r   c              	   C   sj  | j }| d}|dkr| j }| d}|dkrn| j }|  }|dv r,t| |d|dkS || _ t| }||}| d |j}zt| |}	| d W ||_t|jt	@ | _
n||_t|jt	@ | _
w |  t|||	S |dv r{t| |d|dkS |d	krt| |S |d
krt| S |dkrt| |S |dkrt| |S |dkrt| |S |dksd|  krdkrn nt| |||S |dkrt| ||S || _ t| |S |dkr| j }| jtddd}
|
dd  rt|
}|std| j|| d |S || _ | }|j}zt| |}	| d W ||_t|jt	@ | _
n||_t|jt	@ | _
w |  t|||	S )z[Parses a parenthesised subpattern or a flag. Returns FLAGS if it's an
    inline flag.
    Tr   r  =!r  >)Fr   #r   rN   r   09&r   )>includeNr*   zunknown verb)r/   r   parse_lookaround
parse_name
open_groupexpectry   r   r   r$   ignore_spaceclose_groupr   parse_extensionparse_commentparse_conditionalparse_atomicparse_commonparse_call_groupparse_call_named_groupparse_flags_subpatternr  r   isalphaVERBSr%   r   )r   r   r   r   Zsaved_pos_2Zsaved_pos_3r   groupsaved_flagsr   wordverbr:   r:   r;   r   #  s~   








 



r   c              	   C   s  | j }|  }|dkrJt| }||}| d |j}zt| |}| d W ||_t|jt@ | _	n||_t|jt@ | _	w |
  t|||S |dkrkt| dd}| d ||retd| j|t|||S |dkss|dkryt| ||S || _ td	| j|)
zParses a Python extension.r  r  r  r  T)allow_numericcannot refer to an open groupr  unknown extension)r/   r   r  r  r   ry   r   r   r$   r!  r"  r   is_open_groupr%   r   r   r)  )r   r   r   r   r   r-  r.  r   r:   r:   r;   r#  }  s8   




r#  c                 C   sH   	 | j }| d}|r|dkrn
|dkr| d}q|| _ | d dS )zParses a comment.Tr  r   N)r/   r   r   )r   r   r   r:   r:   r;   r$    s   



r$  c              	   C   sZ   |j }zt| |}| d W ||_ t|j t@ | _n||_ t|j t@ | _w t|||S )zParses a lookaround.r  )ry   r   r   r   r$   r!  
LookAround)r   r   behindr   r.  r   r:   r:   r;   r    s   
r  c              	   C   s"  |j }| j}|  }|dkr?|  }|dv rt| |d|dkS |dkr4|  }|dv r4t| |d|dkS || _td| j| j|| _z/t| d}| d t| |}| 	d	r]t| |}nt
 }| d W ||_ t|j t@ | _n||_ t|j t@ | _w | r| rt
 S t|||||S )
z Parses a conditional subpattern.r   r  Fr  r  Tzexpected lookaround conditionalr  rN   )ry   r/   r   parse_lookaround_conditionalr%   r   r  r   r   r   r   r   r$   r!  r   Conditional)r   r   r.  r   r   r-  
yes_branch	no_branchr:   r:   r;   r%    sB   




r%  c              	   C   s   |j }zt| |}| d W ||_ t|j t@ | _n||_ t|j t@ | _w t| |}| dr7t| |}nt }| d t	|||||S )Nr  rN   )
ry   r   r   r   r$   r!  r   r   r   LookAroundConditional)r   r   r6  r   r.  r   r9  r:  r:   r:   r;   r7    s    




r7  c              	   C   sT   |j }zt| |}| d W ||_ t|j t@ | _t|S ||_ t|j t@ | _w )zParses an atomic subpattern.r  )ry   r   r   r   r$   r!  Atomic)r   r   r.  r   r:   r:   r;   r&    s   
r&  c                 C   sz   |j }t| |g}|j }| dr'||_ |t| | t||j }| ds||_ | d t|dkr9|d S t|S )zParses a common groups branch.rN   r  r*   r   )group_countr   r   rR   r  r   rU   r   )r   r   Zinitial_group_countr   Zfinal_group_countr:   r:   r;   r'  	  s   


r'  c                 C   s2   |dkrd}n||  t }| d t|||S )zParses a call to a group.r   r  r  )r  r  r   	CallGroup)r   r   r   r/   r-  r:   r:   r;   r(    s
   
r(  c                 C   s   t | }| d t|||S )zParses a call to a named group.r  )r  r   r>  )r   r   r/   r-  r:   r:   r;   r)  &  s   
r)  c                 C   sR   d}z	 | j }|  }|dkr||  7 }|t| O }q ty(   || _ Y |S w )zParses a set of inline flags.r   TV)r/   r   REGEX_FLAGSru   )r   ry   r   r   r:   r:   r;   parse_flag_set-  s   rA  c                 C   sJ   t | }| drt | }|std| j| jnd}|t@ r!d|_||fS )z!Parses flags being turned on/off.-z$bad inline flags: no flags after '-'r   T)rA  r   r%   r   r/   r   inline_localer   r   flags_on	flags_offr:   r:   r;   parse_flags=  s   
rG  c              	   C   sr   |j }|j |B | @ |_ t|j t@ | _zt| |}| d W ||_ t|j t@ | _|S ||_ t|j t@ | _w )z&Parses a subpattern with scoped flags.r  )ry   r   r$   r!  r   r   )r   r   rE  rF  r.  r   r:   r:   r;   parse_subpatternN  s   
rH  c                 C   s   t | |\}}|t@ rtd| j| j||@ rtd| j| j||j @ t@ }|r5| j|O  _t|j|t M }| drFt| |||S | drTt	| ||| dS td| j| j)zParses a flags subpattern. It could be inline flags or a subpattern
    possibly with local flags. If it's a subpattern, then that's returned;
    if it's a inline flags, then None is returned.
    z-bad inline flags: cannot turn off global flagz(bad inline flags: flag turned on and offr   r  Nr3  )
rG  GLOBAL_FLAGSr%   r   r/   global_flagsrA   r   rH  parse_positional_flags)r   r   rE  rF  Znew_global_flagsr:   r:   r;   r*  \  s(   



r*  c                 C   s&   |j |B | @ |_ t|j t@ | _dS )zParses positional flags.N)ry   r   r$   r!  rD  r:   r:   r;   rK    s   rK  c                 C   sz   | j tddd}|std| j| j| r/|rdnd}|r%t||k r-td| j| j|S | s;td| j| j|S )zParses a name.r  Fr  zmissing group namer   r*   zbad character in group name)r  r   r%   r   r/   isdigitr   isidentifier)r   r1  Zallow_group_0r   Z	min_groupr:   r:   r;   r    s   r  c                 C      t dd | D S )z!Checks whether a string is octal.c                 s       | ]}|t v V  qd S r   )
OCT_DIGITSr   r   r:   r:   r;   	<genexpr>      zis_octal.<locals>.<genexpr>allr   r:   r:   r;   is_octal  r   rW  c                 C   rN  )z#Checks whether a string is decimal.c                 s   rO  r   )r  rQ  r:   r:   r;   rR    rS  zis_decimal.<locals>.<genexpr>rT  rV  r:   r:   r;   
is_decimal  r   rX  c                 C   rN  )z'Checks whether a string is hexadecimal.c                 s   rO  r   )
HEX_DIGITSrQ  r:   r:   r;   rR    rS  z!is_hexadecimal.<locals>.<genexpr>rT  rV  r:   r:   r;   is_hexadecimal  r   rZ  c                 C   s  | j }d| _ |  }|| _ |std| j| j|tv r&t| ||t| ||S |dkrJ|sJ| j}zt| |W S  tyA   || _Y nw t|t	||S |dkrS|sSt
 S |dkr^|s^t| |S |dkrht| ||S |dv rut| ||dk|S |d	kr|sg d
}|jtkr|g d tttddgt|dd |D gS |dkr|st S |tv r|s|jt@ rt|}nt|}|r|S t|}|r|S t|}|rtt	|S td| | j| j|tv rt| |||S t|t	||S )zParses an escape sequence.Fbad escape (end of pattern)gGr   NZpPrh   r   )
            )   i(   i)   rb  r_  c                 S      g | ]}t |qS r:   r   r   r   r:   r:   r;   r     s    z parse_escape.<locals>.<listcomp>r#   bad escape \%s) r!  r   r%   r   r/   HEX_ESCAPESparse_hex_escapeparse_group_refr   r   SearchAnchorparse_string_setparse_named_charparse_propertyr~   r   r   r<  r   Stringr   Graphemer   ry   r"   WORD_POSITION_ESCAPESPOSITION_ESCAPESCHARSET_ESCAPESCHARACTER_ESCAPESr   r  parse_numeric_escape)r   r   r   saved_ignorer   r   charsetrV   r:   r:   r;   r     sd   






r   c           	      C   s   |s|dkrt | ||g|S |}| j}|  }|tv rN||7 }| j}|  }t|rN|tv rN|jt@ }|tks;|t	kr>d}nd}t
|| d|@ }t||S || _||r^td| j| jt||| jS )z!Parses a numeric escape sequence.r       rI   r2  )parse_octal_escaper/   r   r  rW  rP  ry   r}   r   r   r   r   r4  r%   r   r   )	r   r   r   r   r  r   encoding
octal_maskrV   r:   r:   r;   ru    s(   


ru  c                 C   s   | j }|  }t|dk r'|tv r'|| | j }|  }t|dk r'|tv s|| _ ztd|d}t|||W S  ty_   |d tv rSt	dd| | j
| j t	d|d  | j
| j w )z Parses an octal escape sequence.   r^   rI   r   zincomplete escape \%srg  )r/   r   rU   rP  rR   r   rT   r   r  r%   r   )r   r   r  r   r   r   rV   r:   r:   r;   rz    s(   
rz  c                 C   s   | j }g }t|D ]}|  }	|	tvr!td|d|f | j|||	 q	z
td|d}
W n	 t	y:   Y nw |
dk rEt
||
|S td|d|f | j|)zParses a hex escape sequence.incomplete escape \%s%sr^   rJ   i   zbad hex escape \%s%s)r/   ranger   rY  r%   rT   r   rR   r   r  r   )r   r   escexpected_lenr   typer   r  rg   r   rV   r:   r:   r;   ri  "  s&   ri  c                 C   sJ   |  d | j}t| d}|  d ||rtd| j| jt|||S )zParses a group reference.r  Tr  r2  )r   r/   r  r4  r%   r   r   )r   r   r   r   r:   r:   r;   rj  9  s   



rj  c                 C   sJ   |  d t| d}|  d |du s||jvr td| j| jt||S )zParses a string set reference.r  Tr  Nzundefined named list)r   r  kwargsr%   r   r/   r   )r   r   r   r:   r:   r;   rl  D  s   



rl  c                 C   sz   | j }| dr2| jtdd}| dr2zt|}t|t||W S  ty1   t	d| j
| j w || _ t|td|S )zParses a named character.r   T)keep_spacesr   undefined character namer^  )r/   r   r  NAMED_CHAR_PARTunicodedatalookupr   r   ru   r%   r   )r   r   r   r   r   rV   r:   r:   r;   rm  N  s   


rm  c           
      C   s   | j }|  }|dkr+| d}t| \}}| dr*t||||k| }	t||	|S n|r>|dv r>td||| }	t||	|S || _ |rEdnd}t|t||S )zParses a Unicode property.r   r   r   ZCLMNPSZNrh   r   )r/   r   r   parse_property_namelookup_propertyr   r   r   )
r   r   r   r   r   r   negate	prop_namer   r   r:   r:   r;   rn  ^  s   

rn  c                 C   sh   |  t}| j}|  }|r+|dv r+|}|  ttdB  }|r%| j}nd|}}nd}|| _||fS )z/Parses a property name, which may be qualified.z:=z &_-./N)r  PROPERTY_NAME_PARTr/   r   ALNUMr   strip)r   r   r   r   r  r:   r:   r;   r  s  s   
r  c                 C   s   |j t@ pt}| j}d| _| d}z!|tkrt| |}nt| |}| ds/td| j	| j
W || _n|| _w |rB|j|j d}|jt|d}|S )zParses a character set.Fr   ]	missing ]r   r   )ry   _ALL_VERSIONSDEFAULT_VERSIONr!  r   r   parse_set_imp_unionparse_set_unionr%   r   r/   r   r   r   )r   r   versionrv  r  r   r:   r:   r;   r     s    


r   c                 C   N   t | |g}| dr|t | | | dst|dkr"|d S t||S )zParses a set union ([x||y]).r_   r*   r   )parse_set_symm_diffr   rR   rU   r   r   r   itemsr:   r:   r;   r       


r  c                 C   r  )z+Parses a set symmetric difference ([x~~y]).r`   r*   r   )parse_set_interr   rR   rU   
SetSymDiffr  r:   r:   r;   r    r  r  c                 C   r  )z#Parses a set intersection ([x&&y]).ra   r*   r   )parse_set_diffr   rR   rU   SetInterr  r:   r:   r;   r    r  r  c                 C   r  )z!Parses a set difference ([x--y]).rb   r*   r   )r  r   rR   rU   SetDiffr  r:   r:   r;   r    r  r  c                    s   |j t@ pt}t |g}	  j} dr| _n|tkr-t fddtD r-| _n	|	t | qt
|dkr@|d S t||S )z#Parses a set implicit union ([xy]).Tr  c                 3   s    | ]}  |V  qd S r   r	  )r   opr
  r:   r;   rR        z&parse_set_imp_union.<locals>.<genexpr>r*   r   )ry   r  r  parse_set_memberr/   r   r    anySET_OPSrR   rU   r   )r   r   r  r  r   r:   r
  r;   r    s   

r  c                 C   s   t | |}| j}t|tr|jr| ds|S |jt@ pt}| j}|t	kr/| dr/|| _|S | drB|| _t
||ttdgS t | |}t|trO|js[t
||ttd|gS |j|jkritd| j| j|j|jkrq|S t|j|jS )z#Parses a member in a character set.rB  r  zbad character range)parse_set_itemr/   r+   r   r   r   ry   r  r  r    r   r   rV   r%   r   Range)r   r   startZ
saved_pos1r  Z
saved_pos2endr:   r:   r;   r    s,   


r  c                 C   s   |j t@ pt}| drt| |dS | j}| dr-zt| |W S  ty,   || _Y nw |tkrY| drY| d}t	| |}| dsMt
d| j| j|rW|j|j d}|S |  }|sgt
d	| j| jtt|S )
z"Parses an item in a character set.r   Tz[:r   r   r  r  r  zunterminated character set)ry   r  r  r   r   r/   parse_posix_classrD   r    r  r%   r   r   r   r   r   r   )r   r   r  r   r  r   r   r:   r:   r;   r    s,   





r  c                 C   s:   |  d}t| \}}|  dst t||| | ddS )zParses a POSIX character class.r   z:]T)posix)r   r  rD   r  )r   r   r  r  r   r:   r:   r;   r  1  s
   

r  c                 C   sD   t | }| | }t|dk r|dfS td| \}}|| | |fS )z$Converts a float to a rational pair.g-C6?r*   g      ?)r   absfloat_to_rational)ZfltZint_partr%   dennumr:   r:   r;   r  :  s   r  c                 C   s   | dd dkr| d | dd }} nd}|  d}t|dkr2tt|d t|d  \}}nt|dkrCtt|d \}}nt d|||}|d	rX|dd
 S |S )z<Converts a numeric string to a rational string, if possible.Nr*   rB  r   r^   /rG   z{}{}/{}z/1)splitrU   r  floatr  r4   endswith)numericsignpartsr  r  resultr:   r:   r;   numeric_to_rationalE  s   
"
r  c              	   C   s@   zt d| W S  ttfy   ddd | D   Y S w )z&Standardises a property or value name.r^   c                 s   s    | ]	}|d vr|V  qdS )z_- Nr:   rQ  r:   r:   r;   rR  _      z#standardise_name.<locals>.<genexpr>)r  rT   r  ZeroDivisionErrorupper)r   r:   r:   r;   standardise_nameZ  s
   r  zALNUM DIGIT PUNCT XDIGITzYES Y NO N TRUE T FALSE Fc           
      C   s   | rt | nd} t |}| |fdkrdd| } }}|r)| s)| tv r)d| }| rdt| }|s@|s8tdtd|j|j|\}}||}|du r[|sStdtd|j|jt|d> |B |S d	D ]} t| \}}||}|durt|d> |B |  S qft|}|r|\}}t	|t
krt|d> d
B |S t|d> | S |drt|dd }|r|\}}d|v rt|d> d
B |S dD ](\}	} ||	rt| \}}||dd }|durt|d> |B |  S q|stdtd|j|j)zLooks up a property.N)GENERALCATEGORYZASSIGNEDr  Z
UNASSIGNEDr   zunknown propertyzunknown property valuerJ   )ZGCSCRIPTBLOCKr*   ISrG   ZYES))r  r  )INr  )r  r  _POSIX_CLASSES
PROPERTIESr   r%   r   r/   Propertyr   _BINARY_VALUES
startswith)
propertyrV   r   r   r  r   prop_idZ
value_dictval_idprefixr:   r:   r;   r  e  s^   





r  c                 C   s  |   }|tv rTt |}|rdt|gfS |tv r+|dks |r+dt| t| |gfS |dkr7dt| |gfS |dkrJ|rJt| }|durJd|gfS td| | j	| j
t| jtr]d}nd	}|d
kr|}t|dk r| j
}|   }|tvrz|| _
n
||7 }t|dk skdt|d|@ gfS |tv r|}| j
}|   }|tv r||7 }| j
}|   }|rt|| rdt|| d|@ gfS || _
dt|gfS |dkrdtdgfS |std| j	| j
dtdt|gfS )z0Compiles a replacement template escape sequence.Frm   r\  Tr^  Nrg  rx  ry  r  r}  rI   r   r[  )r   r   rt  r   rh  parse_repl_hex_escapecompile_repl_groupparse_repl_named_charr%   r   r/   r+   sepbytesrU   rP  r   r  rW  )r   r.   Z
is_unicoder   rV   r|  r  r   r:   r:   r;   _compile_replacement  sZ   

r  c                 C   sZ   g }t |D ]}|  }|tvrtd|d|f | j| j|| qtd|dS )z5Parses a hex escape sequence in a replacement string.r~  r^   rJ   )	r  r   rY  r%   rT   r   r/   rR   r   )r   r  r  r  rg   r   r:   r:   r;   r    s   r  c                 C   sj   | j }| dr0| ttdB }| dr0z
t|}t|W S  ty/   t	d| j
| j w || _ dS )z1Parses a named character in a replacement string.r    r   r  N)r/   r   r  r   r   r  r  r   ru   r%   r   )r   r   r   rV   r:   r:   r;   r    s   



r  c                 C   s   |  d t| dd}|  d | r.t|}d|  kr#|jks,n td| j| j|S z|j| W S  t	y?   t
dw )z0Compiles a replacement template group reference.r  Tr  r   invalid group referenceunknown group)r   r  rL  r   groupsr%   r   r/   
groupindexru   
IndexError)r   r.   r   indexr:   r:   r;   r    s   

r  z  rJ   z	NON-MATCHMATCHFTz SIMPLE_IGNORE_CASEz FULL_IGNORE_CASEc                 C   s   t | dkr
| d S t| S Nr*   r   )rU   r   r  r:   r:   r;   make_sequence0  s   r  c                   @   s   e Zd Zdd Zd%ddZdd Zdd	 Zd
d Zdd Zdd Z	dd Z
dd Zdd Zdd Zd&ddZdd Zdd Zdd  Zd!d" Zd#d$ ZdS )'	RegexBasec                 C   s   | j | _d S r   )rQ   _keyr7   r:   r:   r;   r6   7     zRegexBase.__init__Nc                 C   s|   |d u r| j }nt|}|d u r| j}nt|t@  }|d u r"| j}nt|}|| j kr7|| jkr7|| jkr7| S | |||S r   )r   r   r   CASE_FLAGS_COMBINATIONSr   r   rebuildr7   r   r   r   r:   r:   r;   r   :  s   
zRegexBase.with_flagsc                 C   s   d S r   r:   r7   r.   r   fuzzyr:   r:   r;   
fix_groupsN     zRegexBase.fix_groupsc                 C      | S r   r:   r7   r   r   r:   r:   r;   r   Q  r  zRegexBase.optimisec                 C   r  r   r:   r7   r   r:   r:   r;   pack_charactersT  r  zRegexBase.pack_charactersc                 C   r  r   r:   r  r:   r:   r;   remove_capturesW  r  zRegexBase.remove_capturesc                 C      dS NTr:   r  r:   r:   r;   	is_atomicZ  r  zRegexBase.is_atomicc                 C   r  r  r:   r  r:   r:   r;   can_be_affix]  r  zRegexBase.can_be_affixc                 C   r  NFr:   r  r:   r:   r;   contains_group`  r  zRegexBase.contains_groupc                 C   s   t  r   )rE   r7   r   r:   r:   r;   get_firstsetc  s   zRegexBase.get_firstsetc                 C   r  r  r:   r  r:   r:   r;   has_simple_startf  r  zRegexBase.has_simple_startFc                 C   s   |  ||S r   )_compiler7   r   r  r:   r:   r;   r   i  r  zRegexBase.compilec                 C   r  r  r:   r  r:   r:   r;   r   l  r  zRegexBase.is_emptyc                 C   
   t | jS r   )hashr  r  r:   r:   r;   __hash__o     
zRegexBase.__hash__c                 C      t | t |u o| j|jkS r   )r  r  r7   otherr:   r:   r;   __eq__r     zRegexBase.__eq__c                 C   s   |  | S r   )r  r  r:   r:   r;   __ne__u  r  zRegexBase.__ne__c                 C      |   d fS r   	max_widthr  r:   r:   r;   get_required_stringx  r  zRegexBase.get_required_string)NNNFF)r<   r=   r>   r6   r   r  r   r  r  r  r  r  r  r  r   r   r  r  r  r  r:   r:   r:   r;   r  6  s$    

r  c                   @   s6   e Zd ZdddZdd Zdd Zdd	 Zd
d ZdS )ZeroWidthBaseTc                 C   s&   t |  t|| _| j| jf| _d S r   )r  r6   r   r   rQ   r  )r7   r   r:   r:   r;   r6   }  s   

zZeroWidthBase.__init__c                 C   s
   t d gS r   r   r  r:   r:   r;   r    r  zZeroWidthBase.get_firstsetc                 C   s6   d}| j r	|tO }|r|tO }|r|tO }| j|fgS Nr   )r   POSITIVE_OPFUZZY_OP
REVERSE_OP_opcoder7   r   r  ry   r:   r:   r;   r    s   zZeroWidthBase._compilec                 C   "   t dt| | jt| j  d S )Nz{}{} {})printr4   INDENT_op_namePOS_TEXTr   r7   indentr   r:   r:   r;   dump     
zZeroWidthBase.dumpc                 C   r  r  r:   r  r:   r:   r;   r    r  zZeroWidthBase.max_widthNT)r<   r=   r>   r6   r  r  r  r  r:   r:   r:   r;   r	  |  s    

r	  c                   @   s>   e Zd ZejejdZdZdd Zdd Z	dd Z
d	d
 ZdS )r   r  ANYc                 C   r  r  r:   r  r:   r:   r;   r    r  zAny.has_simple_startc                 C   s    d}|r|t O }| j| |fgS r  )r  r  r  r:   r:   r;   r    s   zAny._compilec                 C      t dt| | j d S )N{}{})r  r4   r  r  r  r:   r:   r;   r       zAny.dumpc                 C   r  Nr*   r:   r  r:   r:   r;   r    r  zAny.max_widthN)r<   r=   r>   OPr  ZANY_REVr  r  r  r  r  r  r:   r:   r:   r;   r     s    r   c                   @      e Zd ZejejdZdZdS )r   r  ANY_ALLN)r<   r=   r>   r   r"  ZANY_ALL_REVr  r  r:   r:   r:   r;   r         r   c                   @   r!  )r   r  ANY_UN)r<   r=   r>   r   r$  Z	ANY_U_REVr  r  r:   r:   r:   r;   r     r#  r   c                   @      e Zd Zdd Zdd Zdd Zdd Zd	d
 Zdd Zdd Z	dd Z
dd Zdd Zdd Zdd Zdd Zdd Zdd ZdS ) r<  c                 C      t |  || _d S r   )r  r6   r   )r7   r   r:   r:   r;   r6        

zAtomic.__init__c                 C      | j ||| d S r   r   r  r  r:   r:   r;   r       zAtomic.fix_groupsc                 C   s$   | j ||| _ | j  r| j S | S r   )r   r   r   r  r:   r:   r;   r     s   
zAtomic.optimisec                 C      | j || _ | S r   r   r  r  r:   r:   r;   r       zAtomic.pack_charactersc                 C      | j  | _ | S r   r   r  r  r:   r:   r;   r       zAtomic.remove_capturesc                 C   
   | j  S r   r   r  r  r:   r:   r;   r    r  zAtomic.can_be_affixc                 C   r1  r   r   r  r  r:   r:   r;   r    r  zAtomic.contains_groupc                 C      | j |S r   r   r  r  r:   r:   r;   r    r  zAtomic.get_firstsetc                 C   r1  r   r   r  r  r:   r:   r;   r    r  zAtomic.has_simple_startc                 C   s"   t jfg| j|| t jfg S r   )r   ATOMICr   r   ENDr  r:   r:   r;   r    s   zAtomic._compilec                 C   s(   t dt|  | j|d | d S )N{}ATOMICr*   )r  r4   r  r   r  r  r:   r:   r;   r    s   zAtomic.dumpc                 C   r1  r   r   r   r  r:   r:   r;   r     r  zAtomic.is_emptyc                 C   r  r   )r  r   r  r:   r:   r;   r    s   zAtomic.__eq__c                 C   r1  r   r   r  r  r:   r:   r;   r    r  zAtomic.max_widthc                 C   r4  r   r   r  r  r:   r:   r;   r    r  zAtomic.get_required_stringN)r<   r=   r>   r6   r  r   r  r  r  r  r  r  r  r  r   r  r  r  r:   r:   r:   r;   r<    s     r<  c                   @      e Zd ZejZdZdS )BoundaryBOUNDARYN)r<   r=   r>   r   r?  r  r  r:   r:   r:   r;   r>        r>  c                   @   s  e Zd Zdd Zdd Zdd Zdd Zd	d
 Zdd Zdd Z	dd Z
dd Zdd Zdd Zdd Zedd Zedd Zedd Zedd  Zed!d" Zed#d$ Zed%d& Zed'd( Zed)d* Zed+d, Zed-d. Zed/d0 Zd1d2 Zd3d4 Zd5d6 Zd7S )8r   c                 C   r&  r   )r  r6   r   )r7   r   r:   r:   r;   r6     r'  zBranch.__init__c                 C      | j D ]	}|||| qd S r   )r   r  )r7   r.   r   r  rd   r:   r:   r;   r       
zBranch.fix_groupsc                 C   s   | j stg S t||| j }|rt||\}}g }n
t||\}}g }t|||}t|dkrTt|g}|r<|sS| |||}|rS|rM|	| n	|
d| n|}t|| | S r  )r   r   r   _flatten_branches_split_common_suffix_split_common_prefix_reduce_to_setrU   _add_precheckrR   insertr  )r7   r   r   r   suffixr  r   Zfirstsetr:   r:   r;   r     s(   
zBranch.optimisec                 C   sf   t  }|rdnd}|D ]}t|tu r!|jtkr!||j|  q d S |s(d S t||dd |D S )Nr   c                 S   rd  r:   re  rf  r:   r:   r;   r   '      z(Branch._add_precheck.<locals>.<listcomp>)r   r  Literalr   r   r   
charactersr   )r7   r   r   r   rw  r/   branchr:   r:   r;   rG    s   zBranch._add_precheckc                    s    fdd| j D | _ | S )Nc                    s   g | ]}|  qS r:   )r  r   rd   r   r:   r;   r   *  s    z*Branch.pack_characters.<locals>.<listcomp>r   r  r:   rP  r;   r  )  s   zBranch.pack_charactersc                 C      dd | j D | _ | S )Nc                 S      g | ]}|  qS r:   r  rO  r:   r:   r;   r   .  rK  z*Branch.remove_captures.<locals>.<listcomp>rQ  r  r:   r:   r;   r  -     zBranch.remove_capturesc                 C      t dd | jD S )Nc                 s       | ]}|  V  qd S r   r  rO  r:   r:   r;   rR  2  rS  z#Branch.is_atomic.<locals>.<genexpr>rU  r   r  r:   r:   r;   r  1  r*  zBranch.is_atomicc                 C   rV  )Nc                 s   rW  r   )r  rO  r:   r:   r;   rR  5  rS  z&Branch.can_be_affix.<locals>.<genexpr>rY  r  r:   r:   r;   r  4  r*  zBranch.can_be_affixc                 C   rV  )Nc                 s   rW  r   r  rO  r:   r:   r;   rR  8  rS  z(Branch.contains_group.<locals>.<genexpr>)r  r   r  r:   r:   r;   r  7  r*  zBranch.contains_groupc                 C   s.   t  }| jD ]	}|||O }q|pt d gS r   )r   r   r  )r7   r   r   rd   r:   r:   r;   r  :  s   
zBranch.get_firstsetc                 C   sP   | j sg S tjfg}| j D ]}|||| |tjf qtjf|d< |S NrJ  )r   r   BRANCHr   r   rR   NEXTr8  )r7   r   r  r   rd   r:   r:   r;   r  A  s   

zBranch._compilec                 C   sb   t dt|  | jd |d | | jdd  D ]}t dt|  ||d | qd S )Nz{}BRANCHr   r*   {}OR)r  r4   r  r   r  )r7   r  r   rd   r:   r:   r;   r  N  s   zBranch.dumpc                 C   s@   g }|D ]}| | |}t|tr||j q|| q|S r   )r   r+   r   r   r   rR   )r   r   r   new_branchesrd   r:   r:   r;   rC  U  s   
zBranch._flatten_branchesc                    sN  g }|D ]}t |tr||j q||g qtdd |D }|d d|}|k rZ  rZtfdd|D rZd7 |k rZ  rZtfdd|D s@ | jt@ r  dkrt fdd|D s d8   dkrt fdd|D rr dkrg |fS g }|D ]}|t	| d   qd   |fS )Nc                 s       | ]}t |V  qd S r   rU   r   rc   r:   r:   r;   rR  n  rS  z.Branch._split_common_prefix.<locals>.<genexpr>r   c                 3        | ]}|    kV  qd S r   r:   rb  )r/   r  r:   r;   rR  t      r*   c                 3       | ]	}t | V  qd S r   )r   
_can_splitrb  r   r:   r;   rR  }  r  
r+   r   rR   r  minr  rU  ry   r   r  r   r   Zalternativesrd   r   end_posr_  rc   r:   )r0   r/   r  r;   rE  b  s@   
""
zBranch._split_common_prefixc                    sV  g }|D ]}t |tr||j q||g qtdd |D }|d dd| }|kr\  r\tfdd|D r\d8 |kr\  r\tfdd|D sBd  | jt@ r dkrt fdd|D s d8   dkrt fdd|D rt dkrg |fS g }|D ]}|t	|d     q  d  |fS )Nc                 s   r`  r   ra  rb  r:   r:   r;   rR    rS  z.Branch._split_common_suffix.<locals>.<genexpr>r   rJ  c                 3   rc  r   r:   rb  )r/   rI  r:   r;   rR    rd  r*   c                 3   re  r   )r   _can_split_revrb  r   r:   r;   rR    r  rg  ri  r:   )r0   r/   rI  r;   rD    s>   
""
zBranch._split_common_suffixc                 C   s   t | |d s
dS t | |sdS t | |d |d  r!dS t | |d r8t | |d |d  r8dS t | |d rOt | |d |d  rOdS dS )Nr*   TFrG   )r   _is_full_case
_is_folded)r  r0   r:   r:   r;   rf    s   zBranch._can_splitc                 C   s   t | }t| || sdS t| || d sdS t| || d || d  r-dS t| || d rJt| || d || d  rJdS t| || d rgt| || d || d  rgdS dS )NTr*   FrG   )rU   r   rl  rm  )r  r0   r  r:   r:   r;   rk    s    "  zBranch._can_split_revc                 C   s   t t}i }g }|D ]Q}t|r$||j |g ||jt| q
t|t	rM|j
rMt|j
d rM||j
d j |j
 ||j
d jt| q
t| |||| || q
t| ||| |S r  )r   r   r   _is_simple_characterrV   rR   
setdefaultrU   r+   r   r  _flush_char_prefix)r   r   r   prefixedorderr_  rd   r:   r:   r;   _merge_common_prefixes  s$   
zBranch._merge_common_prefixesc                 C   s   t | to| jo| j S r   )r+   r   r   r   r   r:   r:   r;   rn  	  s   zBranch._is_simple_characterc                 C   s   g }t  }t}|D ]3}t|tttfr.|j|kr$t| |||| |j}|	|j
td q	t| |||| || q	t| |||| |S )Nr   )r   r   r+   r   r  SetBaser   r   _flush_set_membersr   r   rR   )r   r   r   r_  r  r   rd   r:   r:   r;   rF  		  s&   
zBranch._reduce_to_setc                    s   |sd S t |  fdddD ]L\}}t|dkr$|t|d  qg }d}|D ]}	t|	dkr>|t|	dd   q*|sH|t  d}q*tt|t|g}
||
| | q|	   	  d S )Nc                    s    | d  S r  r:   )pairrr  r:   r;   <lambda>+	  r  z+Branch._flush_char_prefix.<locals>.<lambda>keyr*   r   FT)
sortedr  rU   rR   r  r   r   r   r   rv   )r   r   rq  rr  r_  rV   r   Zsubbranchesoptionalrd   r   r:   rx  r;   rp  %	  s$    zBranch._flush_char_prefixc                 C   sV   |sd S t |dkrt|d }nt| t|| |}||j|d |  d S )Nr*   r   r   )rU   r   r   r   rR   r   rv   )r   r   r  r   r_  r   r:   r:   r;   rv  ?	  s   zBranch._flush_set_membersc                 C   sF   d|  krt | k sdS  dS | | }t|to"|jo"|jt@ tkS )Nr   F)rU   r+   r   r   r   FULLIGNORECASE)r  rg   r   r:   r:   r;   rl  N	  s   zBranch._is_full_casec                 C   s   t | dk rdS | D ]}t|tr|jr|js dS q
ddd | D }tt|}t	 }|D ]}|tt|kr> dS q1dS )NrG   Fr^   c                 s   s    | ]}t |jV  qd S r   )chrrV   r   rg   r:   r:   r;   rR  a	  r  z$Branch._is_folded.<locals>.<genexpr>T)
rU   r+   r   r   r   rT   r   r   FULL_CASE_FOLDINGget_expand_on_folding)r  rg   foldedexpanding_charsr   r:   r:   r;   rm  W	  s    zBranch._is_foldedc                 C   rV  )Nc                 s   rW  r   r   rO  r:   r:   r;   rR  n	  rS  z"Branch.is_empty.<locals>.<genexpr>rY  r  r:   r:   r;   r   m	  r*  zBranch.is_emptyc                 C   r  r   )r  r   r  r:   r:   r;   r  p	  r  zBranch.__eq__c                 C   rV  )Nc                 s   rW  r   r  rO  r:   r:   r;   rR  t	  rS  z#Branch.max_width.<locals>.<genexpr>)r  r   r  r:   r:   r;   r  s	  r*  zBranch.max_widthN)r<   r=   r>   r6   r  r   rG  r  r  r  r  r  r  r  r  staticmethodrC  rE  rD  rf  rk  rs  rn  rF  rp  rv  rl  rm  r   r  r  r:   r:   r:   r;   r     sP    #

)
(








r   c                   @   sL   e Zd Zdd Zdd Zdd Zdd Zd	d
 Zdd Zdd Z	dd Z
dS )r>  c                 C   s.   t |  || _|| _|| _| j| jf| _d S r   )r  r6   r   r-  r   rQ   r  )r7   r   r-  r   r:   r:   r;   r6   w	  s
   
zCallGroup.__init__c                 C   s   zt | j| _W n" ty*   z
| jj| j | _W n ty'   td|| jw Y nw d| j  kr8| jjks@n td|| j| jdkrU| jj	| j dkrUtd|| j| jj
| ||f | j| jf| _d S )Nr  r   r  r*   zambiguous group reference)r   r-  r  r   group_indexru   r%   r   r=  open_group_countgroup_callsrR   rQ   r  r  r:   r:   r;   r  	  s    zCallGroup.fix_groupsc                 C      t dt| jNzgroup reference not allowedr%   r.   r   r  r:   r:   r;   r  	     zCallGroup.remove_capturesc                 C   s   t j| jfgS r   )r   Z
GROUP_CALLcall_refr  r:   r:   r;   r  	  r  zCallGroup._compilec                 C   r  )Nz{}GROUP_CALL {})r  r4   r  r-  r  r:   r:   r;   r  	  r  zCallGroup.dumpc                 C   r  r   )r  r-  r  r:   r:   r;   r  	  r  zCallGroup.__eq__c                 C      t S r   r   r  r:   r:   r;   r  	  r  zCallGroup.max_widthc                 C   
   d | _ d S r   rP  r  r:   r:   r;   __del__	  r  zCallGroup.__del__N)r<   r=   r>   r6   r  r  r  r  r  r  r  r:   r:   r:   r;   r>  v	  s    r>  c                   @      e Zd Zdd Zdd ZdS )CallRefc                 C   s   || _ || _d S r   )refparsed)r7   r  r  r:   r:   r;   r6   	  s   
zCallRef.__init__c                 C   s&   t j| jfg| j|| t jfg S r   )r   CALL_REFr  r  r  r8  r  r:   r:   r;   r  	  s
   zCallRef._compileNr<   r=   r>   r6   r  r:   r:   r:   r;   r  	  s    r  c                   @   s   e Zd Zedfejedfejedfeje	dfejedfej
edfejedfej
e	dfejiZdedfddZdd ZdddZd	d
 Zdd Zdd Zdd Zdd Zdd Zdd ZdS )r   FTc                 C   s   t |  || _t|| _t| | _t|| _| jr,| jt@ tkr,t	
tt| j| _nt| j| _| j| j| j| j| jf| _d S r   )r  r6   rV   r   r   r  r   r   r~  r   r   r  r  r  rQ   r  r7   rV   r   r   r   r:   r:   r;   r6   	  s   




zCharacter.__init__c                 C      t | j|||S r   )r   rV   r  r:   r:   r;   r  	     zCharacter.rebuildc                 C   r  r   r:   r7   r   r   r   r:   r:   r;   r   	  r  zCharacter.optimisec                 C   
   t | gS r   r
  r  r:   r:   r;   r  	  r  zCharacter.get_firstsetc                 C   r  r  r:   r  r:   r:   r;   r  	  r  zCharacter.has_simple_startc                 C   s   d}| j r	|tO }| jr|tO }|r|tO }t| j| j|f || jg}t	| j
dkr<t|tdd | j
D | jdg}|||S )Nr   r*   c                 S   rd  r:   r   rf  r:   r:   r;   r   	  rK  z&Character._compile.<locals>.<listcomp>r   )r   r  r   ZEROWIDTH_OPr  PrecompiledCoder  r   rV   rU   r  r   ro  r   )r7   r   r  ry   r   r:   r:   r;   r  	  s   
zCharacter._compilec                 C   s<   t t| jd}tdt| t| j |t	| j
  d S )Nbuz{}CHARACTER {} {}{})asciir  rV   lstripr  r4   r  r  r   	CASE_TEXTr   r7   r  r   displayr:   r:   r;   r  	  s   
zCharacter.dumpc                 C   s   || j k| jkS r   )rV   r   r7   r   r:   r:   r;   matches	  r  zCharacter.matchesc                 C   r  r   )rU   r  r  r:   r:   r;   r  	  r  zCharacter.max_widthc                 C   s(   | j sdS tdd | jD | _d| fS )Nr   c                 s   r`  r   r  rf  r:   r:   r;   rR  	  rS  z0Character.get_required_string.<locals>.<genexpr>r   )r   rq   r  folded_charactersr  r:   r:   r;   r  	  s   zCharacter.get_required_stringNF)r<   r=   r>   r   r   	CHARACTERr   ZCHARACTER_IGNr   r~  ZCHARACTER_REVZCHARACTER_IGN_REVr  r6   r  r   r  r  r  r  r  r  r  r:   r:   r:   r;   r   	  s.    


r   c                   @   r%  ) r8  c                 C   s,   t |  || _|| _|| _|| _|| _d S r   )r  r6   r   r-  yes_itemno_itemr   )r7   r   r-  r  r  r   r:   r:   r;   r6   	  s   

zConditional.__init__c                 C   s   zt | j| _W n- ty5   z
| jj| j | _W n ty2   | jdkr)d| _ntd|| jY nw Y nw d| j  krC| jjksKn td|| j| j	
||| | j
||| d S )NZDEFINEr   r  r  )r   r-  r  r   r  ru   r%   r   r=  r  r  r  r  r:   r:   r;   r   
  s"   
zConditional.fix_groupsc                 C   s0   | j ||}| j||}t|| j||| jS r   )r  r   r  r8  r-  r   )r7   r   r   r  r  r:   r:   r;   r   
  s   zConditional.optimisec                 C   s    | j || _ | j|| _| S r   )r  r  r  r  r:   r:   r;   r  
  s   zConditional.pack_charactersc                 C   s   | j  | _ | j | _d S r   )r  r  r  r  r:   r:   r;   r  
  s   zConditional.remove_capturesc                 C      | j  o	| j S r   )r  r  r  r  r:   r:   r;   r  #
  r*  zConditional.is_atomicc                 C   r  r   )r  r  r  r  r:   r:   r;   r  &
  r*  zConditional.can_be_affixc                 C   s   | j  p	| j S r   )r  r  r  r  r:   r:   r;   r  )
  r*  zConditional.contains_groupc                 C   s   | j || j|B S r   )r  r  r  r  r:   r:   r;   r  ,
  s   

zConditional.get_firstsetc                 C   s^   t j| jfg}|| j|| | j||}|r&|t jf || |t j	f |S r   )
r   ZGROUP_EXISTSr-  r   r  r   r  rR   r]  r8  r7   r   r  r   Zadd_coder:   r:   r;   r  0
  s   
zConditional._compilec                 C   s^   t dt| | j | j|d | | j s-t dt|  | j|d | d S d S )Nz{}GROUP_EXISTS {}r*   r^  )r  r4   r  r-  r  r  r  r   r  r:   r:   r;   r  <
  s   
zConditional.dumpc                 C   r  r   )r  r   r  r  r:   r:   r;   r   C
  r*  zConditional.is_emptyc                 C   0   t | t |u o| j| j| jf|j|j|jfkS r   )r  r-  r  r  r  r:   r:   r;   r  F
  
   zConditional.__eq__c                 C      t | j | j S r   r  r  r  r  r  r:   r:   r;   r  J
     zConditional.max_widthc                 C   r  r   rP  r  r:   r:   r;   r  M
  r  zConditional.__del__N)r<   r=   r>   r6   r  r   r  r  r  r  r  r  r  r  r   r  r  r  r:   r:   r:   r;   r8  	  s     r8  c                   @   r=  )DefaultBoundaryDEFAULT_BOUNDARYN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r  P
  r@  r  c                   @   r=  )DefaultEndOfWordDEFAULT_END_OF_WORDN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r  T
  r@  r  c                   @   r=  )DefaultStartOfWordDEFAULT_START_OF_WORDN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r  X
  r@  r  c                   @   r=  )r   END_OF_LINEN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r   \
  r@  r   c                   @   r=  )r   END_OF_LINE_UN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r   `
  r@  r   c                   @   r=  )EndOfStringEND_OF_STRINGN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r  d
  r@  r  c                   @   r=  )r   END_OF_STRING_LINEN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r   h
  r@  r   c                   @   r=  )r   END_OF_STRING_LINE_UN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r   l
  r@  r   c                   @   r=  )	EndOfWordEND_OF_WORDN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r  p
  r@  r  c                   @      e Zd ZdZdd ZdS )FailureFAILUREc                 C   
   t jfgS r   )r   r  r  r:   r:   r;   r  w
  r  zFailure._compileNr<   r=   r>   r  r  r:   r:   r:   r;   r  t
      r  c                   @   sn   e Zd ZdddZdd Zdd Zdd	 Zd
d Zdd Zdd Z	dd Z
dd Zdd Zdd Zdd ZdS )r   Nc                 C   s   t |  |d u ri }|| _|| _d|v r&dD ]}||d v r%||d qt|td@ r:dD ]}||d q0ndD ]}||d q<|dd d|v r^dD ]
}|d |d qQd S ddd|d d d|d< d S )	Nr  r  r   )r   r   re   r   r*   )drg   rj   r  )r  r6   r   r   ro  r   )r7   r   r   re   r:   r:   r;   r6   {
  s0   

zFuzzy.__init__c                 C   s   | j ||d d S r  r)  r  r:   r:   r;   r  
  r*  zFuzzy.fix_groupsc                 C   r+  r   r,  r  r:   r:   r;   r  
  r-  zFuzzy.pack_charactersc                 C   r.  r   r/  r  r:   r:   r;   r  
  r0  zFuzzy.remove_capturesc                 C   r1  r   r   r  r  r:   r:   r;   r  
  r  zFuzzy.is_atomicc                 C   r1  r   r3  r  r:   r:   r;   r  
  r  zFuzzy.contains_groupc                 C   s  g }dD ]}| j | }||d  ||d d u rtn|d  qdD ]}|| j d |  q$| j d d }||d u r@tn| d}|rK|tO }| j d}|rttj|ft| g||d tj	fg | j
|d tjfg S tj|ft| g| j
|d tjfg S )	NZdiser   r*   r  r  r  r   T)r   rR   r   r  r   r   Z	FUZZY_EXTrq   r   r]  r   r8  ZFUZZY)r7   r   r  	argumentsre   vry   r   r:   r:   r;   r  
  s8   
 
zFuzzy._compilec                 C   s>   |   }|r
d| }tdt| | | j|d | d S )Nr  z	{}FUZZY{}r*   )_constraints_to_stringr  r4   r  r   r  )r7   r  r   r   r:   r:   r;   r  
  s
   z
Fuzzy.dumpc                 C   r1  r   r:  r  r:   r:   r;   r   
  r  zFuzzy.is_emptyc                 C   s(   t | t |u o| j|jko| j|jkS r   )r  r   r   r  r:   r:   r;   r  
  s
   
zFuzzy.__eq__c                 C   r  r   r   r  r:   r:   r;   r  
  r  zFuzzy.max_widthc           	      C   s   g }dD ]-}| j | \}}|dkrqd}|dkrd|}||7 }|d ur,|d|7 }|| qg }dD ]}| j d | }|dkrL|d|| q6| j d d }|d urj|dkrjd	d
||}|| d|S )NZidsr   r^   z{}<=z<={}r  r  r  z{}<={}r   r   )r   r4   rR   rT   )	r7   r   r   rh  r  conr  r  limitr:   r:   r;   r  
  s.   


zFuzzy._constraints_to_stringr   )r<   r=   r>   r6   r  r  r  r  r  r  r  r   r  r  r  r:   r:   r:   r;   r   z
  s    
#r   c                   @   $   e Zd Zdd Zdd Zdd ZdS )rp  c                 C   s(   t ttt dd t g}|||S r  )r<  r   r   r   GraphemeBoundaryr   )r7   r   r  Zgrapheme_matcherr:   r:   r;   r     s   zGrapheme._compilec                 C   s   t dt|  d S )Nz
{}GRAPHEME)r  r4   r  r  r:   r:   r;   r    r  zGrapheme.dumpc                 C   r  r   r   r  r:   r:   r;   r    r  zGrapheme.max_widthN)r<   r=   r>   r  r  r  r:   r:   r:   r;   rp  
  s    rp  c                   @      e Zd Zdd ZdS )r  c                 C   s   t jdfgS r  )r   ZGRAPHEME_BOUNDARYr  r:   r:   r;   r     r  zGraphemeBoundary.compileN)r<   r=   r>   r   r:   r:   r:   r;   r        r  c                   @   s   e Zd ZejZdZdd Zdd Zdd Z	dd	 Z
d
d Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zd S )!r   GREEDY_REPEATc                 C   s    t |  || _|| _|| _d S r   )r  r6   r   r   r   )r7   r   r   r   r:   r:   r;   r6     s   

zGreedyRepeat.__init__c                 C   r(  r   r)  r  r:   r:   r;   r    r*  zGreedyRepeat.fix_groupsc                 C   s"   | j ||}t| || j| jS r   )r   r   r  r   r   r7   r   r   r   r:   r:   r;   r     s   zGreedyRepeat.optimisec                 C   r+  r   r,  r  r:   r:   r;   r  $  r-  zGreedyRepeat.pack_charactersc                 C   r.  r   r/  r  r:   r:   r;   r  (  r0  zGreedyRepeat.remove_capturesc                 C   s   | j | jko
| j S r   )r   r   r   r  r  r:   r:   r;   r  ,  r  zGreedyRepeat.is_atomicc                 C   r  r  r:   r  r:   r:   r;   r  /  r  zGreedyRepeat.can_be_affixc                 C   r1  r   r3  r  r:   r:   r;   r  2  r  zGreedyRepeat.contains_groupc                 C   s$   | j |}| jdkr|d  |S r  )r   r  r   r   )r7   r   r   r:   r:   r;   r  5  s   

zGreedyRepeat.get_firstsetc                 C   s\   | j | jg}| jd u r|t n|| j | j||}|s"g S t|g| tj	fg S r   )
r  r   r   rR   r   r   r   rq   r   r8  )r7   r   r  repeatr   r:   r:   r;   r  <  s   
zGreedyRepeat._compilec                 C   sH   | j d u rd}n| j }tdt| | j| j| | j|d | d S )NINF
{}{} {} {}r*   )r   r  r4   r  r  r   r   r  r7   r  r   r  r:   r:   r;   r  I  s   
zGreedyRepeat.dumpc                 C   r1  r   r:  r  r:   r:   r;   r   S  r  zGreedyRepeat.is_emptyc                 C   r  r   )r  r   r   r   r  r:   r:   r;   r  V  s   zGreedyRepeat.__eq__c                 C   s   | j d u rtS | j | j  S r   )r   r   r   r  r  r:   r:   r;   r  [  s   
zGreedyRepeat.max_widthc                 C   sr   | j d u rtn| j }| jdkr| j | }t|td fS | j|\}}|r+||fS | j | }t|td fS r  )r   r   r   r   r  rh  r  )r7   r   r   rl   ofsreqr:   r:   r;   r  a  s   
z GreedyRepeat.get_required_stringN)r<   r=   r>   r   r  r  r  r6   r  r   r  r  r  r  r  r  r  r  r   r  r  r  r:   r:   r:   r;   r     s$    
r   c                   @   r  )r   c                 C   r  r  r:   r  r:   r:   r;   r  o  r  zPossessiveRepeat.is_atomicc                 C   sh   | j ||}|sg S | j| jg}| jd u r|t n|| j tjft	|g| tj
ftj
fg S r   )r   r   r  r   r   rR   r   r   r7  rq   r8  )r7   r   r  r   r  r:   r:   r;   r  r  s   
zPossessiveRepeat._compilec                 C   s^   t dt|  | jd u rd}n| j}t dt|d  | j| j| | j|d | d S )Nr9  r  r  r*   rG   )r  r4   r  r   r  r   r   r  r  r:   r:   r;   r    s   
zPossessiveRepeat.dumpN)r<   r=   r>   r  r  r  r:   r:   r:   r;   r   n  s    r   c                   @   s   e Zd Zdd Zdd Zdd Zdd Zd	d
 Zdd Zdd Z	dd Z
dd Zdd Zdd Zdd Zdd Zdd Zdd Zdd  Zd!S )"r   c                 C   s&   t |  || _|| _|| _d | _d S r   )r  r6   r   r-  r   r  )r7   r   r-  r   r:   r:   r;   r6     s
   

zGroup.__init__c                 C   s(   | ||f| j j| j< | j||| d S r   )r   defined_groupsr-  r   r  r  r:   r:   r;   r    s   zGroup.fix_groupsc                 C   s   | j ||}t| j| j|S r   )r   r   r   r   r-  r  r:   r:   r;   r     s   zGroup.optimisec                 C   r+  r   r,  r  r:   r:   r;   r    r-  zGroup.pack_charactersc                 C   r1  r   r/  r  r:   r:   r;   r    r  zGroup.remove_capturesc                 C   r1  r   r  r  r:   r:   r;   r    r  zGroup.is_atomicc                 C   r  r  r:   r  r:   r:   r;   r    r  zGroup.can_be_affixc                 C   r  r  r:   r  r:   r:   r;   r    r  zGroup.contains_groupc                 C   r4  r   r5  r  r:   r:   r;   r    r  zGroup.get_firstsetc                 C   r1  r   r6  r  r:   r:   r;   r    r  zGroup.has_simple_startc                 C   s   g }| j  }}|dk r| jj| }| jj| }| j ||f}| jj|}|d ur0|tj|fg7 }|tjt	| ||fg7 }|| j
||7 }|tjfg7 }|d urX|tjfg7 }|S r  )r-  r   private_groupsr=  	call_refsr   r   r  ZGROUPr   r   r   r8  )r7   r   r  r   Zpublic_groupZprivate_groupr{  r  r:   r:   r;   r    s   
zGroup._compilec                 C   s@   | j }|dk rt| }tdt| | | j|d | d S )Nr   z
{}GROUP {}r*   )r-  r  r  r4   r  r   r  )r7   r  r   r-  r:   r:   r;   r    s
   z
Group.dumpc                 C   s(   t | t |u o| j| jf|j|jfkS r   )r  r-  r   r  r:   r:   r;   r    s   
zGroup.__eq__c                 C   r1  r   r;  r  r:   r:   r;   r    r  zGroup.max_widthc                 C   r4  r   r<  r  r:   r:   r;   r    r  zGroup.get_required_stringc                 C   r  r   rP  r  r:   r:   r;   r    r  zGroup.__del__N)r<   r=   r>   r6   r  r   r  r  r  r  r  r  r  r  r  r  r  r  r  r:   r:   r:   r;   r     s"    r   c                   @   r=  )KeepKEEPN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r    r@  r  c                   @   r=  )r   LAZY_REPEATN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r     r@  r   c                   @      e Zd ZdddZdd Zdd Zdd	 Zd
d Zdd Zdd Z	dd Z
dd Zdd Zdd Zdd Zdd Zdd Zdd Zd S )!r5  AHEADBEHINDr  c                 C   s(   t |  t|| _t|| _|| _d S r   )r  r6   r   r6  r   r   )r7   r6  r   r   r:   r:   r;   r6     s   



zLookAround.__init__c                 C   s   | j || j| d S r   )r   r  r6  r  r:   r:   r;   r    r  zLookAround.fix_groupsc                 C   s2   | j || j}| jr| r|S t| j| j|S r   )r   r   r6  r   r   r5  r  r:   r:   r;   r     s   zLookAround.optimisec                 C   r+  r   r,  r  r:   r:   r;   r    r-  zLookAround.pack_charactersc                 C   r1  r   r/  r  r:   r:   r;   r    r  zLookAround.remove_capturesc                 C   r1  r   r  r  r:   r:   r;   r     r  zLookAround.is_atomicc                 C   r1  r   r2  r  r:   r:   r;   r    r  zLookAround.can_be_affixc                 C   r1  r   r3  r  r:   r:   r;   r    r  zLookAround.contains_groupc                 C   s&   | j r| j|kr| j|S td gS r   )r   r6  r   r  r   r  r:   r:   r;   r  	  s   
zLookAround.get_firstsetc                 C   sX   d}| j r	|tO }|r|tO }|r|tO }tj|t| j fg| j	| j tj
fg S r  )r   r  r  r  r   Z
LOOKAROUNDr   r6  r   r   r8  r  r:   r:   r;   r    s   zLookAround._compilec                 C   s<   t dt| | j| j t| j  | j|d | j d S )Nz{}LOOK{} {}r*   )	r  r4   r  	_dir_textr6  r  r   r   r  r  r:   r:   r;   r    s   zLookAround.dumpc                 C   s   | j o| j S r   )r   r   r   r  r:   r:   r;   r      r  zLookAround.is_emptyc                 C   r  r   )r  r6  r   r   r  r:   r:   r;   r  #  r  zLookAround.__eq__c                 C   r  r  r:   r  r:   r:   r;   r  '  r  zLookAround.max_widthN)r<   r=   r>   r  r6   r  r   r  r  r  r  r  r  r  r  r   r  r  r:   r:   r:   r;   r5    s     
r5  c                   @   r  )!r;  r  r  r  c                 C   s4   t |  t|| _t|| _|| _|| _|| _d S r   )r  r6   r   r6  r   r   r  r  )r7   r6  r   r   r  r  r:   r:   r;   r6   -  s   



zLookAroundConditional.__init__c                 C   s4   | j ||| | j||| | j||| d S r   )r   r  r  r  r  r:   r:   r;   r  5  s   z LookAroundConditional.fix_groupsc                 C   sD   | j || j}| j|| j}| j|| j}t| j| j|||S r   )r   r   r6  r  r  r;  r   )r7   r   r   r   r  r  r:   r:   r;   r   :  s   zLookAroundConditional.optimisec                 C   s.   | j || _ | j|| _| j|| _| S r   )r   r  r  r  r  r:   r:   r;   r  B  s   z%LookAroundConditional.pack_charactersc                 C   s(   | j  | _ | j | _| j | _d S r   )r   r  r  r  r  r:   r:   r;   r  H  s   z%LookAroundConditional.remove_capturesc                 C      | j  o| j o| j S r   )r   r  r  r  r  r:   r:   r;   r  M     zLookAroundConditional.is_atomicc                 C   r  r   )r   r  r  r  r  r:   r:   r;   r  Q  r  z"LookAroundConditional.can_be_affixc                 C   s   | j  p| j p| j S r   )r   r  r  r  r  r:   r:   r;   r  U  s
   
z$LookAroundConditional.contains_groupc                 C   s   t jt| jt| j fg}|| j| j| |t j	f || j
|| | j||}|r?|t j	f || |t jf |S r   )r   ZCONDITIONALr   r   r6  r   r   r   rR   r]  r  r  r8  r  r:   r:   r;   r  Y  s   
zLookAroundConditional._compilec                 C   s   t dt| | j| j t| j  | j|d | j t dt|  | j	|d | | j
 sGt dt|  | j
|d | d S d S )Nz{}CONDITIONAL {} {}r*   z{}EITHERr^  )r  r4   r  r  r6  r  r   r   r  r  r  r   r  r:   r:   r;   r  g  s   
zLookAroundConditional.dumpc                 C   s   | j  r
| j p| j S r   )r   r   r  r  r  r:   r:   r;   r   q  r  zLookAroundConditional.is_emptyc                 C   r  r   )r  r   r  r  r  r:   r:   r;   r  u  r  zLookAroundConditional.__eq__c                 C   r  r   r  r  r:   r:   r;   r  y  r  zLookAroundConditional.max_widthc                 C   r  r   r  r  r:   r:   r;   r  |  r  z)LookAroundConditional.get_required_stringN)r<   r=   r>   r  r6   r  r   r  r  r  r  r  r  r  r   r  r  r  r:   r:   r:   r;   r;  *  s     

r;  c                   @   r  )r  c                 C   s
   || _ d S r   )r   )r7   r   r:   r:   r;   r6     r  zPrecompiledCode.__init__c                 C   s   t | jgS r   )rq   r   r  r:   r:   r;   r    r  zPrecompiledCode._compileNr  r:   r:   r:   r;   r    s    r  c                   @   s   e Zd Zedfejedfejedfeje	dfejedfej
edfejedfej
e	dfejiZdedfddZdd ZdddZd	d
 Zdd Zdd Zdd Zdd Zdd ZdS )r  FTc                 C   sL   t |  || _t|| _t| | _t|| _| j| j| j| j| jf| _	d S r   )
r  r6   rV   r   r   r  r   r   rQ   r  r  r:   r:   r;   r6     s   




zProperty.__init__c                 C   r  r   )r  rV   r  r:   r:   r;   r    r  zProperty.rebuildc                 C   r  r   r:   r  r:   r:   r;   r     r  zProperty.optimisec                 C   r  r   r
  r  r:   r:   r;   r    r  zProperty.get_firstsetc                 C   r  r  r:   r  r:   r:   r;   r    r  zProperty.has_simple_startc                 C   sF   d}| j r	|tO }| jr|tO }|r|tO }| j| j|f || jfgS r  )r   r  r   r  r  r  r   rV   r  r:   r:   r;   r    s   zProperty._compilec              	   C   sT   t | jd?  }|d |d | jd@  }}tdt| t| j ||t| j  d S )NrJ   r   r*   i  z{}PROPERTY {} {}:{}{})	PROPERTY_NAMESrV   r  r4   r  r  r   r  r   )r7   r  r   r   r   rV   r:   r:   r;   r    s
   
zProperty.dumpc                 C   s   t | j|| jkS r   )r   Zhas_property_valuerV   r   r  r:   r:   r;   r    r*  zProperty.matchesc                 C   r  r  r:   r  r:   r:   r;   r    r  zProperty.max_widthNr  )r<   r=   r>   r   r   ZPROPERTYr   ZPROPERTY_IGNr   r~  ZPROPERTY_REVZPROPERTY_IGN_REVr  r6   r  r   r  r  r  r  r  r  r:   r:   r:   r;   r    s(    


r  c                   @   r  )PrunePRUNEc                 C   r  r   )r   r  r  r:   r:   r;   r    r  zPrune._compileNr  r:   r:   r:   r;   r    r  r  c                   @   s   e Zd Zedfejedfejedfeje	dfejedfej
edfejedfej
e	dfejiZdZdedfddZdd Zddd	Zd
d Zdd Zdd Zdd ZdS )r  FTRANGEc                 C   sV   t |  || _|| _t|| _t| | _t|| _| j	| j| j| j| j| jf| _
d S r   )r  r6   lowerr  r   r   r  r   r   rQ   r  )r7   r  r  r   r   r   r:   r:   r;   r6     s   




zRange.__init__c                 C   s   t | j| j|||S r   )r  r  r  r  r:   r:   r;   r    r*  zRange.rebuildc                 C   s   | j r
| jt@ r
|r| S |jt@ r| jt@ tkr| S t }g }|D ]'}| jt	|  kr2| j
krIn q"tt|}|tdd |D | jd q"|sN| S t|| j
| j d k r`|d|  t|S )Nc                 S   rd  r:   r  rf  r:   r:   r;   r     rK  z"Range.optimise.<locals>.<listcomp>r   r*   r   )r   r   r   ry   r   r~  r   r  r  r   r  r   r  rR   ro  rU   rH  r   )r7   r   r   r   r  r  r   r  r:   r:   r;   r     s(    zRange.optimisec                 C   sJ   d}| j r	|tO }| jr|tO }|r|tO }| j| j|f || j| jfgS r  )	r   r  r   r  r  r  r   r  r  r  r:   r:   r;   r    s   zRange._compilec              	   C   sR   t t| jd}t t| jd}tdt| t| j	 ||t
| j  d S )Nr  z{}RANGE {} {} {}{})r  r  r  r  r  r  r4   r  r  r   r  r   )r7   r  r   Zdisplay_lowerZdisplay_upperr:   r:   r;   r     s   
z
Range.dumpc                 C   s(   | j |  ko| jk| jkS   | jkS r   )r  r  r   r  r:   r:   r;   r    s   (zRange.matchesc                 C   r  r  r:   r  r:   r:   r;   r  
  r  zRange.max_widthNr  )r<   r=   r>   r   r   r  r   Z	RANGE_IGNr   r~  Z	RANGE_REVZRANGE_IGN_REVr  r  r6   r  r   r  r  r  r  r:   r:   r:   r;   r    s     

r  c                   @      e Zd Zedfejedfejedfeje	dfej
edfejedfejedfeje	dfejiZefddZdd Zdd Zd	d
 Zdd Zdd Zdd ZdS )r   FTc                 C   s<   t |  || _|| _|| _t| | _| j| j| jf| _d S r   )	r  r6   r   r-  r   r  r   rQ   r  )r7   r   r-  r   r   r:   r:   r;   r6     s   

zRefGroup.__init__c                 C   s   zt | j| _W n" ty*   z
| jj| j | _W n ty'   td|| jw Y nw d| j  kr8| jjks@n td|| j| j	| j| j
f| _d S )Nr  r*   r  )r   r-  r  r   r  ru   r%   r   r=  rQ   r   r  r  r:   r:   r;   r    s   zRefGroup.fix_groupsc                 C   r  r  r  r  r:   r:   r;   r  +  r  zRefGroup.remove_capturesc                 C   s*   d}|r|t O }| j| j|f || jfgS r  )r  r  r   r-  r  r:   r:   r;   r  .  s   zRefGroup._compilec                 C   r  )Nz{}REF_GROUP {}{})r  r4   r  r-  r  r   r  r:   r:   r;   r  4  r  zRefGroup.dumpc                 C   r  r   r   r  r:   r:   r;   r  8  r  zRefGroup.max_widthc                 C   r  r   rP  r  r:   r:   r;   r  ;  r  zRefGroup.__del__N)r<   r=   r>   r   r   Z	REF_GROUPr   ZREF_GROUP_IGNr   r~  ZREF_GROUP_FLDZREF_GROUP_REVZREF_GROUP_IGN_REVZREF_GROUP_FLD_REVr  r6   r  r  r  r  r  r  r:   r:   r:   r;   r     s$    
	r   c                   @   r=  )rk  SEARCH_ANCHORN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   rk  >  r@  rk  c                   @   s   e Zd Zd(ddZdd Zdd Zdd	 Zd
d Zdd Zdd Z	dd Z
dd Zdd Zdd Zdd Zedd Zedd Zedd Zd d! Zd"d# Zd$d% Zd&d' ZdS ))r   Nc                 C   s    t |  |d u rg }|| _d S r   )r  r6   r  )r7   r  r:   r:   r;   r6   C  s   

zSequence.__init__c                 C   rA  r   )r  r  )r7   r.   r   r  rj   r:   r:   r;   r  J  rB  zSequence.fix_groupsc                 C   sF   g }| j D ]}|||}t|tr||j  q|| qt|S r   )r  r   r+   r   r   rR   r  )r7   r   r   r  rj   r:   r:   r;   r   N  s   

zSequence.optimisec                    s
  g }g }t }| jD ]o}t|tu r7|jr7|js7|j|kr0|js%t |jr0t	
 ||| |j}||j q	t|tu sCt|tu rh|j|kra|jsVt fdd|D rat	
 ||| |j}||j q	t	
 ||| ||  q	t	
 ||| t|S )z+Packs sequences of characters into strings.c                 3       | ]}t  |V  qd S r   r   rf  rP  r:   r;   rR  o  r  z+Sequence.pack_characters.<locals>.<genexpr>)r   r  r  r   r   r   r   r   rV   r   _flush_charactersrR   ro  rL  r  r   rM  r  r  )r7   r   r  rM  r   rj   r:   rP  r;   r  Z  s4   


zSequence.pack_charactersc                 C   rR  )Nc                 S   rS  r:   rT  r   rj   r:   r:   r;   r     rK  z,Sequence.remove_captures.<locals>.<listcomp>r  r  r:   r:   r;   r    rU  zSequence.remove_capturesc                 C   rV  )Nc                 s   rW  r   rX  r  r:   r:   r;   rR    rS  z%Sequence.is_atomic.<locals>.<genexpr>rU  r  r  r:   r:   r;   r    r*  zSequence.is_atomicc                 C   r  r  r:   r  r:   r:   r;   r    r  zSequence.can_be_affixc                 C   rV  )Nc                 s   rW  r   rZ  r  r:   r:   r;   rR    rS  z*Sequence.contains_group.<locals>.<genexpr>)r  r  r  r:   r:   r;   r    r*  zSequence.contains_groupc                 C   sX   t  }| j}|r|  |D ]}|||O }d |vr|  S |d  q|t d gB S r   )r   r  r   r  discard)r7   r   r   r  rj   r:   r:   r;   r    s   zSequence.get_firstsetc                 C   s   t | jo| jd  S r  )r   r  r  r  r:   r:   r;   r    s   zSequence.has_simple_startc                 C   s<   | j }|r|d d d }g }|D ]}|||| q|S r[  )r  r   r   )r7   r   r  seqr   rj   r:   r:   r;   r    s   zSequence._compilec                 C   s   | j D ]}||| qd S r   )r  r  )r7   r  r   rj   r:   r:   r;   r    s   
zSequence.dumpc                    s   |sd S |t @ rt fdd|D st}|t@ tkrFt|}|D ]"}|j}t|dkr:|t	|d |j
d q"|t||j
d q"nt|dkrX|t	|d |d n	|t||d g |d d < d S )Nc                 3   r  r   r  rf  rP  r:   r;   rR    r  z-Sequence._flush_characters.<locals>.<genexpr>r*   r   r   )r   r  r   r~  r   _fix_full_casefoldrM  rU   rR   r   r   ro  )r   rM  r   r  literalsr   charsr:   rP  r;   r    s"   
zSequence._flush_charactersc           
      C   s  dd t  D }t tddd | D  }g }|D ]"}||}|dkr?|||t| f |||d }|dks(qd}g }t	
|D ]$\}}	||k r^|t| || td |t| ||	 td |	}qI|t| k r|t| |d  td |S )	Nc                 S   s   g | ]}t t|qS r:   )r   r   r  rf  r:   r:   r;   r     r   z/Sequence._fix_full_casefold.<locals>.<listcomp>r^   c                 s   r`  r   r  rf  r:   r:   r;   rR    s    z.Sequence._fix_full_casefold.<locals>.<genexpr>r   r*   r   )r   r  r   r  rT   r  findrR   rU   r   _merge_chunksrL  r   r~  )
rM  expandedr   chunksre   foundr/   r  r  r  r:   r:   r;   r    s<   
zSequence._fix_full_casefoldc                 C   s|   t | dk r| S |   | d \}}g }| dd  D ]\}}||kr(t||}q|||f ||}}q|||f |S )NrG   r   r*   )rU   sortr  rR   )r  r  r  Z
new_chunksrj   re   r:   r:   r;   r    s   zSequence._merge_chunksc                 C   rV  )Nc                 s   rW  r   r  r  r:   r:   r;   rR     rS  z$Sequence.is_empty.<locals>.<genexpr>r  r  r:   r:   r;   r     r*  zSequence.is_emptyc                 C   r  r   )r  r  r  r:   r:   r;   r    r  zSequence.__eq__c                 C   rV  )Nc                 s   rW  r   r  r  r:   r:   r;   rR    rS  z%Sequence.max_width.<locals>.<genexpr>)sumr  r  r:   r:   r;   r    r*  zSequence.max_widthc                 C   sT   | j }|r|d d d }d}|D ]}||\}}||7 }|r%||f  S q|d fS )NrJ  r   )r  r  )r7   r   r  offsetrj   r  r  r:   r:   r;   r    s   zSequence.get_required_stringr   )r<   r=   r>   r6   r  r   r  r  r  r  r  r  r  r  r  r  r  r  r  r   r  r  r  r:   r:   r:   r;   r   B  s.    
&

!
r   c                   @   s\   e Zd ZdedfddZdd Zdd Zd	d
 Zdd Zdd Z	dd Z
dd Zdd ZdS )ru  TFc                 C   s\   t |  || _t|| _t|| _t| | _t|| _	d| _
| j| j| j| j| j	f| _d S r  )r  r6   r   rq   r  r   r   r  r   r   
char_widthrQ   r  )r7   r   r  r   r   r   r:   r:   r;   r6     s   





zSetBase.__init__c                 C   s"   t | | j| j|||| jdS r  )r  r   r  r   r  r:   r:   r;   r  &  s
   
zSetBase.rebuildc                 C   r  r   r
  r  r:   r:   r;   r  *  r  zSetBase.get_firstsetc                 C   r  r  r:   r  r:   r:   r;   r  -  r  zSetBase.has_simple_startc                 C   sn   d}| j r	|tO }| jr|tO }|r|tO }| j| j|f |fg}| jD ]	}||	  q$|
tjf |S r  )r   r  r   r  r  r  r   r  r   r   rR   r   r8  )r7   r   r  ry   r   rY   r:   r:   r;   r  0  s   
zSetBase._compilec                 C   sF   t dt| | jt| j t| j  | jD ]
}|	|d | qd S )Nz	{}{} {}{}r*   )
r  r4   r  r  r  r   r  r   r  r  )r7   r  r   rg   r:   r:   r;   r  @  s   
zSetBase.dumpc                 C   s   | j r
| jt@ r
|r| S | jjt@ r| jt@ tkr| S t }g }t	 }|D ]'}| 
t|rMtt|}||vrM|tdd |D | jd || q&|sR| S t| g| S )Nc                 S   rd  r:   r  rf  r:   r:   r;   r   Z  rK  z0SetBase._handle_case_folding.<locals>.<listcomp>r   )r   r   r   r   ry   r   r~  r   r  r   r  r   r   r  rR   ro  r   r   )r7   r   r   r  r  seenr   r  r:   r:   r;   _handle_case_foldingF  s.   
zSetBase._handle_case_foldingc                 C   s   | j r| jt@ s
dS | jjt@ r| jt@ tkrdS t }t	 }|D ]}| 
t|r6tt|}|| q"|s;dS tdd |D S )Nr*   c                 s   r`  r   ra  )r   r  r:   r:   r;   rR  {  rS  z$SetBase.max_width.<locals>.<genexpr>)r   r   r   r   ry   r   r~  r   r  r   r  r   r   r  r   r  )r7   r  r  r   r  r:   r:   r;   r  d  s$   
zSetBase.max_widthc                 C   r  r   rP  r  r:   r:   r;   r  }  r  zSetBase.__del__N)r<   r=   r>   r   r6   r  r  r  r  r  r  r  r  r:   r:   r:   r;   ru    s    
ru  c                   @   v   e Zd Zedfejedfejedfeje	dfejedfej
edfejedfej
e	dfejiZdZd	ddZdd ZdS )
r  FTSET_DIFFc                    s~   | j }t|dkr|d t |dd  g}t|dkr,|d j| j| jd |S t fdd|D | _ |  |S )NrG   r   r*   r   c                 3   s     | ]}|j  d dV  qdS )Tr   N)r   )r   rY   r   r   r:   r;   rR    s    z#SetDiff.optimise.<locals>.<genexpr>)	r  rU   r   r   r   r   r   rq   r  )r7   r   r   r   r  r:   r  r;   r     s   

zSetDiff.optimisec                 C   s,   | j d |o| j d | }|| jkS Nr   r*   r  r  r   r7   r   rY   r:   r:   r;   r    s   "
zSetDiff.matchesNr  )r<   r=   r>   r   r   r  r   ZSET_DIFF_IGNr   r~  ZSET_DIFF_REVZSET_DIFF_IGN_REVr  r  r   r  r:   r:   r:   r;   r    s    
r  c                   @   r  )
r  FT	SET_INTERc                 C      g }| j D ]}|j||dd}t|tr|jr||j  q|| qt|dkr:|d j| j	| j
d|||S t|| _ | ||S NTr   r*   r   r   )r  r   r+   r  r   r   rR   rU   r   r   r   rq   r  r7   r   r   r   r  rY   r:   r:   r;   r        


zSetInter.optimisec                    "   t  fdd| jD }|| jkS )Nc                 3       | ]}|  V  qd S r   r  r  r   r:   r;   rR    r  z#SetInter.matches.<locals>.<genexpr>)rU  r  r   r	  r:   r  r;   r       
zSetInter.matchesNr  )r<   r=   r>   r   r   r
  r   ZSET_INTER_IGNr   r~  ZSET_INTER_REVZSET_INTER_IGN_REVr  r  r   r  r:   r:   r:   r;   r    s    

r  c                   @   r  )
r  FTSET_SYM_DIFFc                 C   r  r  )r  r   r+   r  r   r   rR   rU   r   r   r   rq   r  r  r:   r:   r;   r     r  zSetSymDiff.optimisec                 C   s(   d}| j D ]	}|||k}q|| jkS r  r  )r7   r   rY   rg   r:   r:   r;   r    s   

zSetSymDiff.matchesNr  )r<   r=   r>   r   r   r  r   ZSET_SYM_DIFF_IGNr   r~  ZSET_SYM_DIFF_REVZSET_SYM_DIFF_IGN_REVr  r  r   r  r:   r:   r:   r;   r    s    
r  c                   @   s~   e Zd Zedfejedfejedfeje	dfejedfej
edfejedfej
e	dfejiZdZdddZdd Zdd	 Zd
S )r   FT	SET_UNIONc                 C   s   g }| j D ]}|j||dd}t|tr|jr||j  q|| qt|dkrA|d }|j|j| jk| j	| j
d|||S t|| _ | ||S )NTr   r*   r   )r   r   r   )r  r   r+   r   r   r   rR   rU   r   r   r   rq   r  )r7   r   r   r   r  rY   rg   r:   r:   r;   r     s    


zSetUnion.optimisec           
      C   s  d}| j r	|tO }| jr|tO }|r|tO }ttg }}| jD ]}t|t	r1||j  
|j q |
| q | j| j|f |fg}| D ]-\}}	d}|rR|tO }t|	dkrd|
tj||	d f qF|
tj|t|	ft|	  qF|D ]	}||  qv|
tjf |S r  )r   r  r   r  r  r   r   r  r+   r   rR   rV   r  r   rU   r   r  STRINGrq   r   r   r8  )
r7   r   r  ry   rM  ZothersrY   r   r   valuesr:   r:   r;   r    s0   

 zSetUnion._compilec                    r  )Nc                 3   r  r   r  r  r  r:   r;   rR    r  z#SetUnion.matches.<locals>.<genexpr>)r  r  r   r	  r:   r  r;   r    r  zSetUnion.matchesNr  )r<   r=   r>   r   r   r  r   ZSET_UNION_IGNr   r~  ZSET_UNION_REVZSET_UNION_IGN_REVr  r  r   r  r  r:   r:   r:   r;   r     s    

"r   c                   @   s   e Zd ZdZejZdS )SkipSKIPN)r<   r=   r>   r  r   r  r  r:   r:   r:   r;   r    s    
r  c                   @   r=  )r   START_OF_LINEN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r     r@  r   c                   @   r=  )r   START_OF_LINE_UN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r   #  r@  r   c                   @   r=  )r   START_OF_STRINGN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r   '  r@  r   c                   @   r=  )StartOfWordSTART_OF_WORDN)r<   r=   r>   r   r  r  r  r:   r:   r:   r;   r  +  r@  r  c                   @   r  )ro  FTc                 C   s   t || _t| | _| jt@ tkr,g }| jD ]}ttt|}|	dd |D  qn| j}t || _
d| _| j| j| jf| _d S )Nc                 s   r`  r   r  rf  r:   r:   r;   rR  >  rS  z"String.__init__.<locals>.<genexpr>F)rq   rM  r  r   r~  r   r   r  r  r   r  requiredrQ   r  )r7   rM  r   r  r   r  r:   r:   r;   r6   6  s   



zString.__init__c                 C   s(   |rd}nd}t t| j| | jdgS )NrJ  r   r   )r   r   rM  r   )r7   r   r/   r:   r:   r;   r  G  s   
zString.get_firstsetc                 C   r  r  r:   r  r:   r:   r;   r  O  r  zString.has_simple_startc                 C   sB   d}|r|t O }| jr|tO }| j| j|f |t| jf| j gS r  )r  r  REQUIRED_OPr  r   rU   r  r  r:   r:   r;   r  R  s   zString._compilec                 C   s@   t ddd | jD d}tdt| |t| j  d S )Nr^   c                 s   r`  r   r  rf  r:   r:   r;   rR  \  rS  zString.dump.<locals>.<genexpr>r  z{}STRING {}{})	r  rT   rM  r  r  r4   r  r  r   r  r:   r:   r;   r  [  s    
zString.dumpc                 C   r  r   )rU   r  r  r:   r:   r;   r  `  r  zString.max_widthc                 C   s   d| fS r  r:   r  r:   r:   r;   r  c  s   zString.get_required_stringN)r<   r=   r>   r   r   r  r   Z
STRING_IGNr   r~  Z
STRING_FLDZ
STRING_REVZSTRING_IGN_REVZSTRING_FLD_REVr  r6   r  r  r  r  r  r  r:   r:   r:   r;   ro  /  s    	ro  c                   @   r  )rL  c                 C   sD   d dd | jD }t|d}tdt| |t| j  d S )Nr^   c                 s   r`  r   r  rf  r:   r:   r;   rR  h  rS  zLiteral.dump.<locals>.<genexpr>r  z{}LITERAL MATCH {}{})	rT   rM  r  r  r  r4   r  r  r   )r7   r  r   literalr  r:   r:   r;   r  g  s
   
zLiteral.dumpN)r<   r=   r>   r  r:   r:   r:   r;   rL  f  r  rL  c                   @   s(   e Zd ZefddZdd Zdd ZdS )r   c           
         s   || _ || _t  | _| j| j| jf| _|| jf| _| j|jvr)t|j|j| j< | j j| j }| j j	| j }| j | j j
t@ }| B }g }|D ]}	t|	trVdd |	D }	| fdd|	D  qH|jtdd dd |D | _d S )Nc                 S   rd  r:   r  rf  r:   r:   r;   r     rK  z&StringSet.__init__.<locals>.<listcomp>c                    s   g | ]}t | d qS )r   re  rf  r   r:   r;   r     r   T)r{  r   c                 S   rd  r:   )r   )r   choicer:   r:   r;   r     rK  )r   r   r  r   rQ   r  Zset_keynamed_lists_usedrU   r  ry   r}   r+   r,   rR   r  r   )
r7   r   r   r   r  r  r{  Z
fold_flagschoicesr   r:   r   r;   r6   n  s*   


zStringSet.__init__c                 C   r  )Nz{}STRING_SET {}{})r  r4   r  r   r  r   r  r:   r:   r;   r    r  zStringSet.dumpc                 C   r  r   rP  r  r:   r:   r;   r    r  zStringSet.__del__N)r<   r=   r>   r   r6   r  r  r:   r:   r:   r;   r   m  s    !r   c                   @   sX   e Zd ZdZdd ZdddZddd	ZdddZdddZdd Z	dd Z
dd ZdS )Sourcez1Scanner for the regular expression source string.c                 C   sL   t |tr|| _t| _n|d| _dd | _d| _d| _|d d | _d S )Nzlatin-1c                 S   r  r   )r  rt  r:   r:   r;   ry    s   
 z!Source.__init__.<locals>.<lambda>r   F)	r+   r,   r   r  	char_typedecoder/   r!  r  )r7   r   r:   r:   r;   r6     s   

zSource.__init__Fc                 C   s   | j }| j}z,| jr'|s'	 ||  r|d7 }n|| dkr%|d|}nnq|| }|d | _|W S  tyD   || _|d d  Y S  tyW   t|| _|d d  Y S w )NTr*   r  r(   r   )r   r/   r!  isspacer  r  r  rU   )r7   Zoverride_ignorer   r/   r   r:   r:   r;   r     s*   




z
Source.getr*   c                 C   s   | j }| j}zQ| jrDg }t||k r>	 ||  r|d7 }n|| dkr+|d|}nnq|||  |d7 }t||k sd|}n||||  }|t|7 }|| _|W S  tyj   t|| _d| Y S  t	y|   t|| _d| Y S w NTr*   r  r(   r^   )
r   r/   r!  rU   r(  r  rR   rT   r  r  )r7   r0   r   r/   	substringr:   r:   r;   get_many  s8   



zSource.get_manyTc                 C   s2  | j }| j}| jrb|sbz6g }	 ||  r|d7 }n"|| dkr'|d|}n|| |v |kr;|||  |d7 }nnq|| _W n tyO   t|| _Y n ty\   t|| _Y nw d	|S z!|| |v |krw|d7 }|| |v |ksk|| j| }|| _|W S  ty   || j| }|| _| Y S w r)  )
r   r/   r!  r(  r  rR   r  rU   r  rT   )r7   test_setr  r  r   r/   r*  r:   r:   r;   r    sD   




zSource.get_whilec                 C   s   | j }| j}zF| jr3	 ||  r|d7 }n|| dkr#|d|}n|| |v |kr0|d7 }nnqn|| |v |krG|d7 }|| |v |ks;|| _W d S  ty[   t|| _Y d S  tyi   t|| _Y d S w NTr*   r  r(   )r   r/   r!  r(  r  r  rU   r  )r7   r,  r  r   r/   r:   r:   r;   
skip_while  s.   


zSource.skip_whilec                 C   s   | j }| j}| jrQz4|D ]+}	 ||  r|d7 }n|| dkr'|d|}nnq|| |kr3 W dS |d7 }q|| _W dS  tyG   Y dS  tyP   Y dS w |||sYdS |t| | _dS )NTr*   r  r(   F)	r   r/   r!  r(  r  r  r  r  rU   )r7   r*  r   r/   r   r:   r:   r;   r   3  s4   


zSource.matchc                 C   s$   |  |std|| j| jd S )Nz
missing {})r   r%   r4   r   r/   )r7   r*  r:   r:   r;   r   Z  s   
zSource.expectc                 C   s~   | j }| j}z%| jr%	 ||  r|d7 }n|| dkr#|d|}nnq|t|kW S  ty5   Y dS  ty>   Y dS w r-  )r   r/   r!  r(  r  rU   r  r  )r7   r   r/   r:   r:   r;   at_end^  s"   
zSource.at_endNr  r*   )TFr  )r<   r=   r>   r?   r6   r   r+  r  r.  r   r   r/  r:   r:   r:   r;   r%    s    


&
0'r%  c                   @   s:   e Zd ZdZddi fddZdddZdd	 Zd
d ZdS )Infoz"Info about the regular expression.r   Nc                 C   sl   |t |t@ pt O }|| _|| _d| _|| _d| _i | _i | _	|| _
i | _g | _i | _i | _g | _i | _d S )NFr   )DEFAULT_FLAGSr  r  ry   rJ  rC  r  r=  r  
group_namer&  r#  open_groupsr  r  r  r  )r7   ry   r&  r  r:   r:   r;   r6   w  s   
zInfo.__init__c                 C   s   | j |}|d u r-	 |  jd7  _|d u s| j| jvrnq| j}|r-|| j |< || j|< || jv rAt| jd  }|| j|< |}| j| | j|dd | j|< |S )NTr*   r   )	r  r   r=  r3  r4  rU   r  rR   r  )r7   r   r-  Zgroup_aliasr:   r:   r;   r    s$   



zInfo.open_groupc                 C   s   | j   d S r   )r4  r   r  r:   r:   r;   r"    r  zInfo.close_groupc                 C   sB   | j t@ pt}|tkrdS | rt|}n| j|}|| jv S r  )	ry   r  r  r    rL  r   r  r   r4  )r7   r   r  r-  r:   r:   r;   r4    s   

zInfo.is_open_groupr   )r<   r=   r>   r?   r6   r  r"  r4  r:   r:   r:   r;   r1  t  s    
r1  c                 C   s   i }g }| j D ]b\}}}|j||f}||}|du rf|jdkrBt| jt@ }	t|t}
|	|
f||fkrA|t	t
||||f n| j|j }|d }|dd ||fkr^||||f t
|}|||< ||_q|| _|| _dS )zkChecks whether the reverse and fuzzy features of the group calls match
    the groups which they call.
    Nr   r*   )r  r-  r   r   ry   r   r+   r   rR   r  rU   r  r  r  additional_groups)r   r  r  r5  callr   r  r{  r  revZfuzZdef_infor-  r:   r:   r;   _check_group_features  s.   



r8  c                 C   sb   |  t|t@ \}}|r&d|_|tkrd}|j}|t@ s"|t M }|j}nd}d}d}|||fS )z>Gets the required string and related info of a parsed pattern.TrJ  r   r:   )r  r   r   r  r   r   r   r  )r  ry   
req_offsetr  	req_flags	req_charsr:   r:   r;   _get_required_string  s   

r<  c                   @   s   e Zd ZdddZdd ZdS )r&   r   c                    s  || _ g }|D ].\}}t|}t||j t jt@ |_t| }|	 s.t
d|j|j||  qt|  fddt|D }t|}t jt@ }| |}| }t| j\}	}
}t |  jrtt
d|j|jt jt@ }||tjfg }t|}| szt ||}t|}|| }W n	 ty   Y nw  jt@ pt }|dt!t"fvrt#dt$d |t%@ |B |i i i g |	|
|t&|| _'d S )Nzunbalanced parenthesisc                    s    g | ]\}}t  |d  |qS r0  )r   )r   r\  rh   rP  r:   r;   r     s     z$Scanner.__init__.<locals>.<listcomp>z(recursive regex not supported by Scannerr   z5VERSION0 and VERSION1 flags are mutually incompatible)(lexiconr%  r1  r&  r   ry   r$   r!  r   r/  r%   r   r/   rR   r  	enumerater   r   r   r  r<  r8  r  r   r   SUCCESSr   r  r   r  rE   r  r  r   r    r  r   rI  rU   scanner)r7   r=  ry   patternsphraseactionr   r  r   r9  r;  r:  r   Zfs_coder  r:   rP  r;   r6     sX   




zScanner.__init__c           	      C   s   g }|j }| j|j}d}	 | }|sn-| }||krn$| j|jd  d }t|dr7|| _|| | }|d ur?|| |}q|||d  fS )Nr   Tr*   __call__)rR   r@  r   r  r=  	lastindexrt   r-  )	r7   r   r  rR   r   rg   rY   jrC  r:   r:   r;   scan;  s(   
zScanner.scanN)r   )r<   r=   r>   r6   rG  r:   r:   r:   r;   r&     s    
Ir&   rz  r(   	)rc   rd   rf   nri   tr  ZDigitTZBlankZSpaceWord)r  r   hrj   r   rl   r!   )r   rd   r   KrY   r   Z)rd   r   rY   r   )FAILr   r  r  )ro   r  r  r  )enumr   r  collectionsr   Zregex._regexr   __all__r5   r%   rA   rD   rE   IntFlagr'   globalsrw   __members__r    r  r   r  r   r   r   r}   r   r2  r   r   r
   r   r   rI  r   r   r   r"   r$   ZSCOPED_FLAGS	frozensetascii_lettersr   r  r  r  	octdigitsrP  	hexdigitsrY  r   r  r  r  Zget_code_sizeZBYTES_PER_CODEZBITS_PER_CODEr   r@  r   r   r~  r  r  rh  OPCODESrn   r   r>  r  rg   r  setattrr|   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r  r   r   r   r#  r$  r  r%  r7  r&  r'  r(  r)  rA  rG  rH  r*  rK  r  rW  rX  rZ  r   ru  rz  ri  rj  rl  rm  rn  r  r   r  r  r  r  r  r  r  r  r  r  r  r   r  r  r  r  r  r  r  r  r  r  r  r  r   r  r  r  r  r	  r   r   r   r<  r>  r   r>  r  r   r8  r  r  r  r   r   r  r   r   r  r  r   rp  r  r   r   r   r  r   r5  r;  r  r  r  r  r   rk  r   ru  r  r  r  r   r  r   r   r   r  ro  rL  r   r%  r1  r8  r<  r&   Zget_propertiesr  r  r  r  r  r  r   r   Zprop_valuesr  rU   Zval_namer  rt  rs  rr  dictrq  r,  r:   r:   r:   r;   <module>   s  3





U2


T 
=
	Z$+#I!




)&	FH
F8   .	JY \RDU4M1 Vi!B7) _A'a
	