« back to all changes in this revision
Viewing changes to bzrlib/win32utils.py
- browse files at revision 5046
- compare with another revision
- download diff
- download tarball
- view history from revision 5046
- Committer: Canonical.com Patch Queue Manager
- Date: 2010-02-18 03:02:28 UTC
- mfrom: (5037.3.2 integration)
- Revision ID: pqm@pqm.ubuntu.com-20100218030228-dtbmt8pdtf21ndo7
(mbp, for doxx) better win32 cmdline splitter
added
removed
bzrlib/win32utils.py
25
25
import struct
26
26
import sys
27
27
28
from bzrlib import cmdline
28
29
29
30
# Windows version
30
31
if sys.platform == 'win32':
522
523
trace.mutter('Unable to set hidden attribute on %r: %s', path, e)
523
524
524
525
525
526
class UnicodeShlex(object):
527
"""This is a very simplified version of shlex.shlex.
528
529
The main change is that it supports non-ascii input streams. The internal
530
structure is quite simplified relative to shlex.shlex, since we aren't
531
trying to handle multiple input streams, etc. In fact, we don't use a
532
file-like api either.
533
"""
534
535
def __init__(self, uni_string):
536
self._input = uni_string
537
self._input_iter = iter(self._input)
538
self._whitespace_match = re.compile(u'\s').match
539
self._word_match = re.compile(u'\S').match
540
self._quote_chars = u'"'
541
# self._quote_match = re.compile(u'[\'"]').match
542
self._escape_match = lambda x: None # Never matches
543
self._escape = '\\'
544
# State can be
545
# ' ' - after whitespace, starting a new token
546
# 'a' - after text, currently working on a token
547
# '"' - after ", currently in a "-delimited quoted section
548
# "\" - after '\', checking the next char
549
self._state = ' '
550
self._token = [] # Current token being parsed
551
552
def _get_token(self):
553
# Were there quote chars as part of this token?
554
quoted = False
555
quoted_state = None
556
for nextchar in self._input_iter:
557
if self._state == ' ':
558
if self._whitespace_match(nextchar):
559
# if self._token: return token
560
continue
561
elif nextchar in self._quote_chars:
562
self._state = nextchar # quoted state
563
elif self._word_match(nextchar):
564
self._token.append(nextchar)
565
self._state = 'a'
566
else:
567
raise AssertionError('wtttf?')
568
elif self._state in self._quote_chars:
569
quoted = True
570
if nextchar == self._state: # End of quote
571
self._state = 'a' # posix allows 'foo'bar to translate to
572
# foobar
573
elif self._state == '"' and nextchar == self._escape:
574
quoted_state = self._state
575
self._state = nextchar
576
else:
577
self._token.append(nextchar)
578
elif self._state == self._escape:
579
if nextchar == '\\':
580
self._token.append('\\')
581
elif nextchar == '"':
582
self._token.append(nextchar)
583
else:
584
self._token.append('\\' + nextchar)
585
self._state = quoted_state
586
elif self._state == 'a':
587
if self._whitespace_match(nextchar):
588
if self._token:
589
break # emit this token
590
else:
591
continue # no token to emit
592
elif nextchar in self._quote_chars:
593
# Start a new quoted section
594
self._state = nextchar
595
# escape?
596
elif (self._word_match(nextchar)
597
or nextchar in self._quote_chars
598
# or whitespace_split?
599
):
600
self._token.append(nextchar)
601
else:
602
raise AssertionError('state == "a", char: %r'
603
% (nextchar,))
604
else:
605
raise AssertionError('unknown state: %r' % (self._state,))
606
result = ''.join(self._token)
607
self._token = []
608
if not quoted and result == '':
609
result = None
610
return quoted, result
611
612
def __iter__(self):
613
return self
614
615
def next(self):
616
quoted, token = self._get_token()
617
if token is None:
618
raise StopIteration
619
return quoted, token
620
621
622
def _command_line_to_argv(command_line):
623
"""Convert a Unicode command line into a set of argv arguments.
624
625
This does wildcard expansion, etc. It is intended to make wildcards act
626
closer to how they work in posix shells, versus how they work by default on
627
Windows.
628
"""
629
s = UnicodeShlex(command_line)
630
# Now that we've split the content, expand globs
526
def _command_line_to_argv(command_line, single_quotes_allowed=False):
527
"""Convert a Unicode command line into a list of argv arguments.
528
529
It performs wildcard expansion to make wildcards act closer to how they
530
work in posix shells, versus how they work by default on Windows. Quoted
531
arguments are left untouched.
532
533
:param command_line: The unicode string to split into an arg list.
534
:param single_quotes_allowed: Whether single quotes are accepted as quoting
535
characters like double quotes. False by
536
default.
537
:return: A list of unicode strings.
538
"""
539
s = cmdline.Splitter(command_line, single_quotes_allowed=single_quotes_allowed)
540
# Now that we've split the content, expand globs if necessary
631
541
# TODO: Use 'globbing' instead of 'glob.glob', this gives us stuff like
632
542
# '**/' style globs
633
543
args = []
641
551
642
552
if has_ctypes and winver != 'Windows 98':
643
553
def get_unicode_argv():
644
LPCWSTR = ctypes.c_wchar_p
645
INT = ctypes.c_int
646
POINTER = ctypes.POINTER
647
prototype = ctypes.WINFUNCTYPE(LPCWSTR)
648
GetCommandLine = prototype(("GetCommandLineW",
649
ctypes.windll.kernel32))
650
prototype = ctypes.WINFUNCTYPE(POINTER(LPCWSTR), LPCWSTR, POINTER(INT))
651
command_line = GetCommandLine()
554
prototype = ctypes.WINFUNCTYPE(ctypes.c_wchar_p)
555
GetCommandLineW = prototype(("GetCommandLineW",
556
ctypes.windll.kernel32))
557
command_line = GetCommandLineW()
558
if command_line is None:
559
raise ctypes.WinError()
652
560
# Skip the first argument, since we only care about parameters
653
561
argv = _command_line_to_argv(command_line)[1:]
654
562
if getattr(sys, 'frozen', None) is None:
Loggerhead is a web-based interface for Breezy
Version: 2.0.1