cpython/Lib/lib2to3/pgen2/driver.py
Benjamin Peterson d481e3d791 Merged revisions 72494 via svnmerge from
svn+ssh://pythondev@svn.python.org/python/trunk

................
  r72494 | benjamin.peterson | 2009-05-08 20:01:14 -0500 (Fri, 08 May 2009) | 21 lines

  Merged revisions 72491-72493 via svnmerge from
  svn+ssh://pythondev@svn.python.org/sandbox/trunk/2to3/lib2to3

  ........
    r72491 | benjamin.peterson | 2009-05-08 19:33:27 -0500 (Fri, 08 May 2009) | 7 lines

    make 2to3 use unicode internally on 2.x

    This started out as a fix for #2660, but became this large refactoring
    when I realized the dire state this was in. 2to3 now uses
    tokenize.detect_encoding to decode the files correctly into unicode.
  ........
    r72492 | benjamin.peterson | 2009-05-08 19:35:38 -0500 (Fri, 08 May 2009) | 1 line

    remove compat code
  ........
    r72493 | benjamin.peterson | 2009-05-08 19:54:15 -0500 (Fri, 08 May 2009) | 1 line

    add a test for \r\n newlines
  ........
................
2009-05-09 19:42:23 +00:00

148 lines
4.8 KiB
Python

# Copyright 2004-2005 Elemental Security, Inc. All Rights Reserved.
# Licensed to PSF under a Contributor Agreement.
# Modifications:
# Copyright 2006 Google, Inc. All Rights Reserved.
# Licensed to PSF under a Contributor Agreement.
"""Parser driver.
This provides a high-level interface to parse a file into a syntax tree.
"""
__author__ = "Guido van Rossum <guido@python.org>"
__all__ = ["Driver", "load_grammar"]
# Python imports
import codecs
import os
import logging
import sys
# Pgen imports
from . import grammar, parse, token, tokenize, pgen
class Driver(object):
def __init__(self, grammar, convert=None, logger=None):
self.grammar = grammar
if logger is None:
logger = logging.getLogger()
self.logger = logger
self.convert = convert
def parse_tokens(self, tokens, debug=False):
"""Parse a series of tokens and return the syntax tree."""
# XXX Move the prefix computation into a wrapper around tokenize.
p = parse.Parser(self.grammar, self.convert)
p.setup()
lineno = 1
column = 0
type = value = start = end = line_text = None
prefix = ""
for quintuple in tokens:
type, value, start, end, line_text = quintuple
if start != (lineno, column):
assert (lineno, column) <= start, ((lineno, column), start)
s_lineno, s_column = start
if lineno < s_lineno:
prefix += "\n" * (s_lineno - lineno)
lineno = s_lineno
column = 0
if column < s_column:
prefix += line_text[column:s_column]
column = s_column
if type in (tokenize.COMMENT, tokenize.NL):
prefix += value
lineno, column = end
if value.endswith("\n"):
lineno += 1
column = 0
continue
if type == token.OP:
type = grammar.opmap[value]
if debug:
self.logger.debug("%s %r (prefix=%r)",
token.tok_name[type], value, prefix)
if p.addtoken(type, value, (prefix, start)):
if debug:
self.logger.debug("Stop.")
break
prefix = ""
lineno, column = end
if value.endswith("\n"):
lineno += 1
column = 0
else:
# We never broke out -- EOF is too soon (how can this happen???)
raise parse.ParseError("incomplete input",
type, value, (prefix, start))
return p.rootnode
def parse_stream_raw(self, stream, debug=False):
"""Parse a stream and return the syntax tree."""
tokens = tokenize.generate_tokens(stream.readline)
return self.parse_tokens(tokens, debug)
def parse_stream(self, stream, debug=False):
"""Parse a stream and return the syntax tree."""
return self.parse_stream_raw(stream, debug)
def parse_file(self, filename, encoding=None, debug=False):
"""Parse a file and return the syntax tree."""
stream = codecs.open(filename, "r", encoding)
try:
return self.parse_stream(stream, debug)
finally:
stream.close()
def parse_string(self, text, debug=False):
"""Parse a string and return the syntax tree."""
tokens = tokenize.generate_tokens(generate_lines(text).__next__)
return self.parse_tokens(tokens, debug)
def generate_lines(text):
"""Generator that behaves like readline without using StringIO."""
for line in text.splitlines(True):
yield line
while True:
yield ""
def load_grammar(gt="Grammar.txt", gp=None,
save=True, force=False, logger=None):
"""Load the grammar (maybe from a pickle)."""
if logger is None:
logger = logging.getLogger()
if gp is None:
head, tail = os.path.splitext(gt)
if tail == ".txt":
tail = ""
gp = head + tail + ".".join(map(str, sys.version_info)) + ".pickle"
if force or not _newer(gp, gt):
logger.info("Generating grammar tables from %s", gt)
g = pgen.generate_grammar(gt)
if save:
logger.info("Writing grammar tables to %s", gp)
try:
g.dump(gp)
except IOError as e:
logger.info("Writing failed:"+str(e))
else:
g = grammar.Grammar()
g.load(gp)
return g
def _newer(a, b):
"""Inquire whether file a was written since file b."""
if not os.path.exists(a):
return False
if not os.path.exists(b):
return True
return os.path.getmtime(a) >= os.path.getmtime(b)