# Copyright 2004-2005 Elemental Security, Inc. All Rights Reserved.
# Licensed to PSF under a Contributor Agreement.
# Copyright 2006 Google, Inc. All Rights Reserved.
# Licensed to PSF under a Contributor Agreement.
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"]
from . import grammar, parse, token, tokenize, pgen
def __init__(self, grammar, convert=None, logger=None):
logger = logging.getLogger()
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)
type = value = start = end = line_text = None
type, value, start, end, line_text = quintuple
if start != (lineno, column):
assert (lineno, column) <= start, ((lineno, column), start)
s_lineno, s_column = start
prefix += "\n" * (s_lineno - lineno)
prefix += line_text[column:s_column]
if type in (tokenize.COMMENT, tokenize.NL):
type = grammar.opmap[value]
self.logger.debug("%s %r (prefix=%r)",
token.tok_name[type], value, prefix)
if p.addtoken(type, value, (prefix, start)):
self.logger.debug("Stop.")
# We never broke out -- EOF is too soon (how can this happen???)
raise parse.ParseError("incomplete input",
type, value, (prefix, start))
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)
return self.parse_stream(stream, debug)
def parse_string(self, text, debug=False):
"""Parse a string and return the syntax tree."""
tokens = tokenize.generate_tokens(StringIO.StringIO(text).readline)
return self.parse_tokens(tokens, debug)
def _generate_pickle_name(gt):
head, tail = os.path.splitext(gt)
return head + tail + ".".join(map(str, sys.version_info)) + ".pickle"
def load_grammar(gt="Grammar.txt", gp=None,
save=True, force=False, logger=None):
"""Load the grammar (maybe from a pickle)."""
logger = logging.getLogger()
gp = _generate_pickle_name(gt) if gp is None else gp
if force or not _newer(gp, gt):
logger.info("Generating grammar tables from %s", gt)
g = pgen.generate_grammar(gt)
logger.info("Writing grammar tables to %s", gp)
logger.info("Writing failed: %s", e)
"""Inquire whether file a was written since file b."""
if not os.path.exists(a):
if not os.path.exists(b):
return os.path.getmtime(a) >= os.path.getmtime(b)
def load_packaged_grammar(package, grammar_source):
"""Normally, loads a pickled grammar by doing
pkgutil.get_data(package, pickled_grammar)
where *pickled_grammar* is computed from *grammar_source* by adding the
Python version and using a ``.pickle`` extension.
However, if *grammar_source* is an extant file, load_grammar(grammar_source)
is called instead. This facilitates using a packaged grammar file when needed
but preserves load_grammar's automatic regeneration behavior when possible.
if os.path.isfile(grammar_source):
return load_grammar(grammar_source)
pickled_name = _generate_pickle_name(os.path.basename(grammar_source))
data = pkgutil.get_data(package, pickled_name)
"""Main program, when run as a script: produce grammar pickle files.
Calls load_grammar for each argument, a path to a grammar text file.
logging.basicConfig(level=logging.INFO, stream=sys.stdout,
load_grammar(gt, save=True, force=True)
if __name__ == "__main__":
sys.exit(int(not main()))
It is recommended that you Edit text format, this type of Fix handles quite a lot in one request
