�`^c@sEdZd
dl
Z
d
dlmZmZ
dejf
d��YZdS(s�Convert graminit.[ch] spit out by pgen to Python code.

Pgen is the Python parser generator.  It is useful to quickly create a
parser from a grammar file in Python's grammar notation.  But I don't
want my parsers to be written in C (yet), so I'm translating the
parsing tables to Python data structures and writing a Python parse
engine.

Note that the token numbers are constants determined by the standard
Python tokenizer.  The standard token module defines these numbers and
their names (the names are not used much).  The token numbers are
hardcoded into the Python tokenizer and into pgen.  A Python
implementation of the Python tokenizer is also available, in the
standard tokenize module.

On the other hand, symbol numbers (representing the grammar's
non-terminals) are assigned by pgen based on the actual grammar
input.

Note: this module is pretty much obsolete; the pgen module generates
equivalent grammar tables directly from the Grammar.txt input file
without having to invoke the Python pgen C program.

i�N(tgrammarttokent	Converterc
Bs2eZ
dZd
�Zd�Zd�Zd�ZRS(s2
Grammar subclass that reads classic pgen output files.

    The run() method reads the tables as produced by the pgen parser
    generator, typically contained in two C files, graminit.h and
    graminit.c.  The other methods are for internal use only.

    See the base class for more documentation.

    cCs(|j|
�
|j
|�
|j�
d
S(s<Load the grammar tables from the text files written by pgen.N(tparse_graminit_htparse_graminit_ct
finish_off(tselft
graminit_ht
graminit_c((s(/sys/lib/python2.7/lib2to3/pgen2/conv.pytrun/s



c	Cs�yt|
�}Wn#t
k
r5
}
d
|
|fGHtSXi|_i|_d}x�|D]�}|d7}tjd|�}|r�|j�r�d|
||j�fGHqU|j�\}}t	|�}||j|<||j|<qUWt
S(s�Parse the .h file written by pgen.  (Internal)

        This file is a sequence of #define statements defining the
        nonterminals of the grammar as numbers.  We build two tables
        mapping the numbers to names and back.

        sCan't open %s: %sii
s^#define\s+(\w+)\s+(\d+)$s%s(%s): can't parse %s(topentIOErrortFalset
symbol2numbert
number2symboltretmatchtstriptgroupstinttTrue(	Rtfilenamet
fterrtlinenotlinetmotsymboltnumber((s(/sys/lib/python2.7/lib2to3/pgen2/conv.pyR5s&




	
	







	




c!CsHyt|
�}Wn#t
k
r5
}
d
|
|fGHtSXd}|d|j�}}|d|j�}}|d|j�}}i}g}x�jd�r�x�jd�r�tjd|�}tt|j	��\}	}
}g}xkt
|�D]]}
|d|j�}}tjd|�}tt|j	��\}}|j||f�
q�d|j�}}|||	|
f<|d|j�}}q�Wtjd|�}tt|j	��\}}g}xxt
|�D]j}
|d|j�}}tjd|�}tt|j	��\}}	}
||	|
f}|j|�
q�W|j|�
|d|j�}}|d|j�}}q�W||_i}tjd	|�}t|j
d��}x:
t
|�D],
}|d|j�}}tjd
|�}|j
d�}tt|j
ddd
d��\}}}}||}|d|j�}}tjd|�}i}t|j
d��}x`t|�D]R\}}t|�}x7t
d�D])}|d|>@r�d||d|<q�q�Wq�W||f||<q�W|d|j�}}||_g}|d|j�}}tjd|�}t|j
d��}x�t
|�D]�}|d|j�}}tjd|�}|j	�\}}t|�}|dkr�d}nt|�}|j||f�
qhW|d|j�}}||_|d|j�}}|d|j�}}tjd|�}t|j
d��}|d|j�}}|d|j�}}tjd|�}t|j
d��}|d|j�}}tjd|�}t|j
d��} | |_|d|j�}}y|d|j�}}Wntk
rC


n
XdS(s�Parse the .c file written by pgen.  (Internal)

        The file looks as follows.  The first two lines are always this:

        #include "pgenheaders.h"
        #include "grammar.h"

        After that come four blocks:

        1) one or more state definitions
        2) a table defining dfas
        3) a table defining labels
        4) a struct defining the grammar

        A state definition has the following form:
        - one or more arc arrays, each of the form:
          static arc arcs_<n>_<m>[<k>] = {
                  {<i>, <j>},
                  ...
          };
        - followed by a state array, of the form:
          static state states_<s>[<t>] = {
                  {<k>, arcs_<n>_<m>},
                  ...
          };

        sCan't open %s: %sii
sstatic arc s)static arc arcs_(\d+)_(\d+)\[(\d+)\] = {$s\s+{(\d+), (\d+)},$s'static state states_(\d+)\[(\d+)\] = {$s\s+{(\d+), arcs_(\d+)_(\d+)},$sstatic dfa dfas\[(\d+)\] = {$s0\s+{(\d+), "(\w+)", (\d+), (\d+), states_(\d+),$iiiis\s+("(?:\\\d\d\d)*")},$is!static label labels\[(\d+)\] = {$s\s+{(\d+), (0|"\w+")},$t
0s
\s+(\d+),$s\s+{(\d+), labels},$s	\s+(\d+)$N(R
RRtnextt
startswithRRtmapRRtrangetappendtstatestgrouptevalt	enumeratetordtdfastNonetlabelststartt
StopIteration(!RRRRRRtallarcsR#Rt
nt
mt
ktarcst
_t
it
jt
st
ttstateR(tndfasRRt
xt
yt
ztfirstt	rawbitsett
ctbyteR*tnlabelsR+((s(/sys/lib/python2.7/lib2to3/pgen2/conv.pyRTs�








	
	















	



	
	
-









	







	

	


	




c
Cs�i|_i|_
xjt|j�D]Y\}
\}}|tjkr_|d
k	r_|
|j|<q"|d
kr"|
|j
|<q"q"Wd
S(s1Create additional useful structures.  (Internal).N(tkeywordsttokensR&R*R
tNAMER)(Rtilabelttypetvalue((s(/sys/lib/python2.7/lib2to3/pgen2/conv.pyR�s	
	
"



(t__name__t
__module__t__doc__R	RRR(((s(/sys/lib/python2.7/lib2to3/pgen2/conv.pyR$s
				�RIRtpgen2RR
tGrammarR(((s(/sys/lib/python2.7/lib2to3/pgen2/conv.pyt<module>s