#!/usr/bin/python
"""
parse_cpython.py
"""
from __future__ import print_function
import errno
import os
import re
import sys
from core.util import log
from frontend.lexer import C, R
C_DEF = [
R(r'#.*', 'Comment'),
R(r'[ \t\n]+', 'Whitespace'),
# This could be more space-insensitive.
R(r'static.*PyMethodDef (.*)\[\] = ', 'BeginDef'),
C(r'{', 'LBrace'),
C(r'}', 'RBrace'),
C(r',', 'Comma'),
C(r';', 'Semi'),
R(r'"([^"]*)"', 'Str'),
C(r'FILE', 'FILE'),
C(r'PyDoc_STR(', 'LDocStr'),
C(r')', 'RDocStr'),
R(r'[^,}\n]+', 'Opaque'),
]
# EXPRESSION (takes place of ARITH, VSub_ArgUnquoted, VSub_ArgDQ)
# SQ RAW_SQ DQ RAW_DQ
# VS -- a single state here? Or switches into expression state, because }
# is an operator
# Problem: DICT_KEY might be a different state, to accept either a bare word
# foo, or an expression (X=a+2), which is allowed in shell. Python doesn't
# allowed unquoted words, but we want to.
# TODO: There are 4 shared groups here. I think you should test if that
# structure should be preserved through re2c. Do a benchmark.
#
# If a group has no matches, then return Id.Unknown_Tok? And then you can
# chain the groups in order. It might make sense to experiment with the order
# too.
_SIGNIFICANT_SPACE = R(r'[ \t\r]+', Id.WS_Space)
_BACKSLASH = [
R(r'\\[^\n\0]', Id.Lit_EscapedChar),
C('\\\n', Id.Ignored_LineCont),
]
VAR_NAME_RE = r'[a-zA-Z_][a-zA-Z0-9_]*'
# All Kind.VSub
_VARS = [
# Unbraced variables
R(r'\$' + VAR_NAME_RE, Id.VSub_DollarName),
R(r'\$[0-9]', Id.VSub_Number),
C(r'$!', Id.VSub_Bang),
C(r'$@', Id.VSub_At),
# SQ RAW_SQ DQ RAW_DQ
# VS -- a single state here? Or switches into expression state, because }
# is an operator
# Problem: DICT_KEY might be a different state, to accept either a bare word
# foo, or an expression (X=a+2), which is allowed in shell. Python doesn't
# allowed unquoted words, but we want to.
# TODO: There are 4 shared groups here. I think you should test if that
# structure should be preserved through re2c. Do a benchmark.
#
# If a group has no matches, then return Id.Unknown_Tok? And then you can
# chain the groups in order. It might make sense to experiment with the order
# too.
_BACKSLASH = [
R(r'\\[^\n\0]', Id.Lit_EscapedChar),
C('\\\n', Id.Ignored_LineCont),
]
VAR_NAME_RE = r'[a-zA-Z_][a-zA-Z0-9_]*'
# All Kind.VSub
_VARS = [
# Unbraced variables
R(r'\$' + VAR_NAME_RE, Id.VSub_DollarName),
R(r'\$[0-9]', Id.VSub_Number),
C(r'$!', Id.VSub_Bang),
C(r'$@', Id.VSub_At),
C(r'$#', Id.VSub_Pound),
C(r'$$', Id.VSub_Dollar),
C(r'$*', Id.VSub_Star),