Your IP : 216.73.216.213
B
Z��ZP? � @ s� d dl mZ d dlZyd dlmZ W n ek
rD d dlmZ Y nX ddlmZ dZ dZ
d Z e d
�Z
d d � Zd
d� Zdd� Zdd� Zdd� ZG dd� de�Zdd� ZG dd� de�ZG dd � d e�Ze� ZG d d � d e�ZG d d � d e�ZG d!d"� d"e�ZG d#d$� d$e�Z G d%d&� d&e�Z e� Z d'e _ d(e _ d)d*� Z!d+d,� Z"d-d.� Z#d/d0� Z$e$d1�Z%d2e%_ d3e%_ dDd4d5�Z&d6d7� Z'd8d9� Z(d:d;� Z)d<d=� Z*ee �Z+d>e+_ d?e+_ ee
�Z,d@e,_ dAe,_ ee �Z-dBe-_ dCe-_ dS )E� )�absolute_importN)�maxsize)�maxint� )�ErrorsZbol�eol�eof�
c C s� t | �}|�� d}t|�}g }xj||k r�t|| �}|d }|d7 }x,||k rr|t|| �krr|d7 }|d7 }qHW |�|� |�|� q"W |S )z�
Return a list of character codes consisting of pairs
[code1a, code1b, code2a, code2b,...] which cover all
the characters in |s|.
r r )�list�sort�len�ord�append)�sZ char_list�i�n�result�code1�code2� r �F/opt/alt/python37/lib64/python3.7/site-packages/Cython/Plex/Regexps.py�chars_to_ranges" s
r c C sP t | td��}t|td�d �}||k rHtd�td� }|| || fS dS dS )z�
If the range of characters from code1 to code2-1 includes any
lower case letters, return the corresponding upper case range.
�a�zr �AN)�maxr
�min)r r �code3�code4�dr r r �uppercase_range9 s r c C sP t | td��}t|td�d �}||k rHtd�td� }|| || fS dS dS )z�
If the range of characters from code1 to code2-1 includes any
upper case letters, return the corresponding lower case range.
r �Zr r N)r r
r )r r r r r r r r �lowercase_rangeG s r" c s& � fdd�t dt� �d�D �}t|� S )z�
Given a list of codes as returned by chars_to_ranges, return
an RE which will match a character in any of the ranges.
c s" g | ]}t � | � |d ��qS )r )� CodeRange)�.0r )� code_listr r �
<listcomp>Z s z CodeRanges.<locals>.<listcomp>r � )�ranger �Alt)r% �re_listr )r% r �
CodeRangesU s r+ c C sB | t kr|k r4n n tt| t �ttt d |��S t| |�S dS )z�
CodeRange(code1, code2) is an RE which matches any character
with a code |c| in the range |code1| <= |c| < |code2|.
r N)�nl_coder) � RawCodeRange�
RawNewline)r r r r r r# ^ s
r# c @ sd e Zd ZdZdZdZdZdd� Zdd� Zdd � Z d
d � Z
d d
� Z dd� Z dd� Z
dd� Zdd� ZdS )�REa RE is the base class for regular expression constructors.
The following operators are defined on REs:
re1 + re2 is an RE which matches |re1| followed by |re2|
re1 | re2 is an RE which matches either |re1| or |re2|
r Nc C s t d| jj ��dS )aM
This method should add states to |machine| to implement this
RE, starting at |initial_state| and ending at |final_state|.
If |match_bol| is true, the RE must be able to match at the
beginning of a line. If nocase is true, upper and lower case
letters should be treated as equivalent.
z %s.build_machine not implementedN)�NotImplementedError� __class__�__name__)�self�machine�
initial_state� final_state� match_bol�nocaser r r �
build_machine{ s zRE.build_machinec C s" |� � }|�|� |�||� |S )z~
Given a state |s| of machine |m|, return a new state
reachable from |s| on character |c| or epsilon.
)� new_state�link_to�add_transition)r3 �mr5 �cr r r r � build_opt� s
z RE.build_optc C s
t | |�S )N)�Seq)r3 �otherr r r �__add__� s z
RE.__add__c C s
t | |�S )N)r) )r3 rA r r r �__or__� s z RE.__or__c C s | j r | j S | �� S d S )N)�str�calc_str)r3 r r r �__str__� s z
RE.__str__c C s t |t�s| �||d� d S )NzPlex.RE instance)�
isinstancer/ �
wrong_type)r3 �num�valuer r r �check_re� s
z RE.check_rec C s" t |�t d�kr | �||d� d S )N� �string)�typerH )r3 rI rJ r r r � check_string� s zRE.check_stringc C s8 | � ||� t|�dkr4t�d|| jjt|�f ��d S )Nr zOInvalid value for argument %d of Plex.%s.Expected a string of length 1, got: %s)rO r r ZPlexValueErrorr1 r2 �repr)r3 rI rJ r r r �
check_char� s z
RE.check_charc C sL t |�tjkr$d|jj|jjf }n
t |�j}t�d|| jj||f ��d S )Nz%s.%s instancez<Invalid type for argument %d of Plex.%s (expected %s, got %s)rN �typesZ InstanceTyper1 �
__module__r2 r Z
PlexTypeError)r3 rI rJ ZexpectedZgotr r r rH � s
z
RE.wrong_type)r2 rS � __qualname__�__doc__�nullable�match_nlrD r9 r? rB rC rF rK rO rQ rH r r r r r/ o s
r/ c C s> t | �dkr$tt| �t| �d �}nt| �}dt| � |_|S )z;
Char(c) is an RE which matches the character |c|.
r zChar(%s))r r# r
�
SpecialSymbolrP rD )r> r r r r �Char� s
rY c @ s< e Zd ZdZdZdZdZdZdZdd� Z dd� Z
dd � Z dS )
r- z�
RawCodeRange(code1, code2) is a low-level RE which matches any character
with a code |c| in the range |code1| <= |c| < |code2|, where the range
does not include newline. For internal use only.
r Nc C s& ||f| _ t||�| _t||�| _d S )N)r( r r" )r3 r r r r r �__init__� s
zRawCodeRange.__init__c C sP |r| � ||t�}|�| j|� |rL| jr8|�| j|� | jrL|�| j|� d S )N)r? �BOLr< r( r r" )r3 r= r5 r6 r7 r8 r r r r9 � s zRawCodeRange.build_machinec C s d| j | jf S )NzCodeRange(%d,%d))r r )r3 r r r rE s zRawCodeRange.calc_str) r2 rS rT rU rV rW r( r r" rZ r9 rE r r r r r- � s
r- c @ s e Zd ZdZdZdZdd� ZdS )� _RawNewlinezd
RawNewline is a low-level RE which matches a newline character.
For internal use only.
r r c C s8 |r| � ||t�}| � ||t�}|�ttd f|� d S )Nr )r? r[ �EOLr< r, )r3 r= r5 r6 r7 r8 r r r r r9 s z_RawNewline.build_machineN)r2 rS rT rU rV rW r9 r r r r r\ s r\ c @ s, e Zd ZdZdZdZdZdd� Zdd� ZdS )rX zx
SpecialSymbol(sym) is an RE which matches the special input
symbol |sym|, which is one of BOL, EOL or EOF.
r Nc C s
|| _ d S )N)�sym)r3 r^ r r r rZ # s zSpecialSymbol.__init__c C s. |r | j tkr | �||t�}|�| j |� d S )N)r^ r] r? r[ r< )r3 r= r5 r6 r7 r8 r r r r9 & s zSpecialSymbol.build_machine) r2 rS rT rU rV rW r^ rZ r9 r r r r rX s rX c @ s( e Zd ZdZdd� Zdd� Zdd� ZdS ) r@ z]Seq(re1, re2, re3...) is an RE which matches |re1| followed by
|re2| followed by |re3|...c G s� d}x*t |�D ] \}}| �||� |o*|j}qW || _|| _t|�}d}x,|rt|d8 }|| }|jrjd}P |jsJP qJW || _d S )Nr r )� enumeraterK rV r* r rW )r3 r* rV r �rerW r r r rZ 2 s" z Seq.__init__c C s� | j }t|�dkr |�|� nb|}t|�}xTt|�D ]H\} }
| |d k rR|�� } n|} |
�||| ||� | }|
jpz|oz|
j}q4W d S )Nr r )r* r r; r_ r: r9 rW rV ) r3 r= r5 r6 r7 r8 r* �s1r r r` �s2r r r r9 E s
zSeq.build_machinec C s dd� tt| j�� S )NzSeq(%s)�,)�join�maprD r* )r3 r r r rE U s z Seq.calc_strN)r2 rS rT rU rZ r9 rE r r r r r@ . s r@ c @ s( e Zd ZdZdd� Zdd� Zdd� ZdS ) r) zRAlt(re1, re2, re3...) is an RE which matches either |re1| or
|re2| or |re3|...c G s� || _ d}d}g }g }d}xJ|D ]B}| �||� |jrF|�|� d}n
|�|� |jrZd}|d7 }q W || _|| _|| _|| _d S )Nr r )r* rK rV r rW � nullable_res�non_nullable_res)r3 r* rV rW rf rg r r` r r r rZ ] s&
z Alt.__init__c C s` x | j D ]}|�|||||� qW | jr\|r:| �||t�}x | jD ]}|�|||d|� qBW d S )Nr )rf r9 rg r? r[ )r3 r= r5 r6 r7 r8 r` r r r r9 s s zAlt.build_machinec C s dd� tt| j�� S )NzAlt(%s)rc )rd re rD r* )r3 r r r rE | s z Alt.calc_strN)r2 rS rT rU rZ r9 rE r r r r r) Y s r) c @ s( e Zd ZdZdd� Zdd� Zdd� ZdS ) �Rep1z@Rep1(re) is an RE which matches one or more repetitions of |re|.c C s& | � d|� || _|j| _|j| _d S )Nr )rK r` rV rW )r3 r` r r r rZ � s z
Rep1.__init__c C sN |� � }|� � }|�|� | j�||||p0| jj|� |�|� |�|� d S )N)r: r; r` r9 rW )r3 r= r5 r6 r7 r8 ra rb r r r r9 � s
zRep1.build_machinec C s
d| j S )NzRep1(%s))r` )r3 r r r rE � s z
Rep1.calc_strN)r2 rS rT rU rZ r9 rE r r r r rh � s rh c @ s0 e Zd ZdZdZdZdd� Zdd� Zdd� ZdS ) �
SwitchCasez�
SwitchCase(re, nocase) is an RE which matches the same strings as RE,
but treating upper and lower case letters according to |nocase|. If
|nocase| is true, case is ignored, otherwise it is not.
Nc C s || _ || _|j| _|j| _d S )N)r` r8 rV rW )r3 r` r8 r r r rZ � s zSwitchCase.__init__c C s | j �||||| j� d S )N)r` r9 r8 )r3 r= r5 r6 r7 r8 r r r r9 � s zSwitchCase.build_machinec C s | j r d}nd}d|| jf S )N�NoCase�Casez%s(%s))r8 r` )r3 �namer r r rE � s zSwitchCase.calc_str) r2 rS rT rU r` r8 rZ r9 rE r r r r ri � s ri z8
Empty is an RE which matches the empty string.
�Emptyc C s$ t ttt| ��� }dt| � |_|S )z@
Str1(s) is an RE which matches the literal string |s|.
zStr(%s))r@ �tuplere rY rP rD )r r r r r �Str1� s ro c G sH t | �dkrt| d �S tttt| ��� }dd�tt| �� |_|S dS )z�
Str(s) is an RE which matches the literal string |s|.
Str(s1, s2, s3, ...) is an RE which matches any of |s1| or |s2| or |s3|...
r r zStr(%s)rc N)r ro r) rn re rd rP rD )�strsr r r r �Str� s
rq c C s t t| ��}dt| � |_|S )zH
Any(s) is an RE which matches any character in the string |s|.
zAny(%s))r+ r rP rD )r r r r r �Any� s rr c C s: t | �}|�dt � |�t� t|�}dt| � |_|S )zp
AnyBut(s) is an RE which matches any character (including
newline) which is not in the string |s|.
r z
AnyBut(%s))r �insertr r r+ rP rD )r �rangesr r r r �AnyBut� s
ru rL zT
AnyChar is an RE which matches any single character (including a newline).
�AnyCharc C s� |r*t t| �t|�d �}d| |f |_n\g }x@tdt| �d�D ],}|�t t| | �t| |d �d �� q@W t|� }dt| � |_|S )a
Range(c1, c2) is an RE which matches any single character in the range
|c1| to |c2| inclusive.
Range(s) where |s| is a string of even length is an RE which matches
any single character in the ranges |s[0]| to |s[1]|, |s[2]| to |s[3]|,...
r z Range(%s,%s)r r' z Range(%s))r# r
rD r( r r r) rP )ra rb r rt r r r r �Range� s ,rw c C s t | t�}d| |_|S )zI
Opt(re) is an RE which matches either |re| or the empty string.
zOpt(%s))r) rm rD )r` r r r r �Opt s
rx c C s t t| ��}d| |_|S )zJ
Rep(re) is an RE which matches zero or more repetitions of |re|.
zRep(%s))rx rh rD )r` r r r r �Rep s
ry c C s t | dd�S )z�
NoCase(re) is an RE which matches the same strings as RE, but treating
upper and lower case letters as equivalent.
r )r8 )ri )r` r r r rj s rj c C s t | dd�S )z�
Case(re) is an RE which matches the same strings as RE, but treating
upper and lower case letters as distinct, i.e. it cancels the effect
of any enclosing NoCase().
r )r8 )ri )r` r r r rk s rk z=
Bol is an RE which matches the beginning of a line.
�Bolz7
Eol is an RE which matches the end of a line.
�Eolz9
Eof is an RE which matches the end of the file.
�Eof)N).�
__future__r rR �sysr r � ImportErrorrL r r[ r] ZEOFr
r, r r r" r+ r# �objectr/ rY r- r\ r. rX r@ r) rh ri rm rU rD ro rq rr ru rv rw rx ry rj rk rz r{ r| r r r r �<module> sb n +'!