2000-07-29 21:04:17 -03:00
|
|
|
"""distutils.filelist
|
|
|
|
|
|
|
|
Provides the FileList class, used for poking about the filesystem
|
|
|
|
and building lists of files.
|
|
|
|
"""
|
|
|
|
|
|
|
|
__revision__ = "$Id$"
|
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
import os, re
|
2000-07-29 21:04:17 -03:00
|
|
|
import fnmatch
|
|
|
|
from distutils.util import convert_path
|
2000-07-29 21:21:36 -03:00
|
|
|
from distutils.errors import DistutilsTemplateError, DistutilsInternalError
|
2002-06-04 18:04:03 -03:00
|
|
|
from distutils import log
|
2000-07-29 21:04:17 -03:00
|
|
|
|
|
|
|
class FileList:
|
2000-07-29 21:08:13 -03:00
|
|
|
"""A list of files built by on exploring the filesystem and filtered by
|
|
|
|
applying various patterns to what we find there.
|
|
|
|
|
|
|
|
Instance attributes:
|
|
|
|
dir
|
|
|
|
directory from which files will be taken -- only used if
|
|
|
|
'allfiles' not supplied to constructor
|
|
|
|
files
|
|
|
|
list of filenames currently being built/filtered/manipulated
|
|
|
|
allfiles
|
|
|
|
complete list of files under consideration (ie. without any
|
|
|
|
filtering applied)
|
|
|
|
"""
|
2000-07-29 21:04:17 -03:00
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def __init__(self, warn=None, debug_print=None):
|
2002-06-04 17:14:43 -03:00
|
|
|
# ignore argument to FileList, but keep them for backwards
|
|
|
|
# compatibility
|
2000-07-29 22:45:42 -03:00
|
|
|
self.allfiles = None
|
|
|
|
self.files = []
|
2000-07-29 21:04:17 -03:00
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def set_allfiles(self, allfiles):
|
2000-07-29 22:45:42 -03:00
|
|
|
self.allfiles = allfiles
|
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def findall(self, dir=os.curdir):
|
2000-07-29 22:45:42 -03:00
|
|
|
self.allfiles = findall(dir)
|
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def debug_print(self, msg):
|
2000-07-29 21:04:17 -03:00
|
|
|
"""Print 'msg' to stdout if the global DEBUG (taken from the
|
|
|
|
DISTUTILS_DEBUG environment variable) flag is true.
|
|
|
|
"""
|
2002-09-11 13:31:53 -03:00
|
|
|
from distutils.debug import DEBUG
|
2000-07-29 21:04:17 -03:00
|
|
|
if DEBUG:
|
|
|
|
print msg
|
|
|
|
|
2000-07-29 22:45:42 -03:00
|
|
|
# -- List-like methods ---------------------------------------------
|
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def append(self, item):
|
2000-07-29 22:45:42 -03:00
|
|
|
self.files.append(item)
|
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def extend(self, items):
|
2000-07-29 22:45:42 -03:00
|
|
|
self.files.extend(items)
|
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def sort(self):
|
2000-07-29 22:45:42 -03:00
|
|
|
# Not a strict lexical sort!
|
|
|
|
sortable_files = map(os.path.split, self.files)
|
|
|
|
sortable_files.sort()
|
|
|
|
self.files = []
|
|
|
|
for sort_tuple in sortable_files:
|
2009-10-02 21:07:35 -03:00
|
|
|
self.files.append(os.path.join(*sort_tuple))
|
2000-07-29 22:45:42 -03:00
|
|
|
|
|
|
|
|
|
|
|
# -- Other miscellaneous utility methods ---------------------------
|
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def remove_duplicates(self):
|
2000-07-29 22:45:42 -03:00
|
|
|
# Assumes list has been sorted!
|
2002-06-04 17:14:43 -03:00
|
|
|
for i in range(len(self.files) - 1, 0, -1):
|
|
|
|
if self.files[i] == self.files[i - 1]:
|
2000-07-29 22:45:42 -03:00
|
|
|
del self.files[i]
|
|
|
|
|
|
|
|
|
|
|
|
# -- "File template" methods ---------------------------------------
|
2001-12-06 16:51:35 -04:00
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def _parse_template_line(self, line):
|
|
|
|
words = line.split()
|
2000-07-29 21:08:13 -03:00
|
|
|
action = words[0]
|
|
|
|
|
2000-07-29 21:21:36 -03:00
|
|
|
patterns = dir = dir_pattern = None
|
|
|
|
|
|
|
|
if action in ('include', 'exclude',
|
|
|
|
'global-include', 'global-exclude'):
|
2000-09-25 23:12:31 -03:00
|
|
|
if len(words) < 2:
|
2000-07-29 21:21:36 -03:00
|
|
|
raise DistutilsTemplateError, \
|
|
|
|
"'%s' expects <pattern1> <pattern2> ..." % action
|
2000-07-29 21:04:17 -03:00
|
|
|
|
2000-07-29 21:21:36 -03:00
|
|
|
patterns = map(convert_path, words[1:])
|
2000-07-29 21:04:17 -03:00
|
|
|
|
2000-07-29 21:21:36 -03:00
|
|
|
elif action in ('recursive-include', 'recursive-exclude'):
|
2000-09-25 23:12:31 -03:00
|
|
|
if len(words) < 3:
|
2000-07-29 21:21:36 -03:00
|
|
|
raise DistutilsTemplateError, \
|
|
|
|
"'%s' expects <dir> <pattern1> <pattern2> ..." % action
|
2000-07-29 21:04:17 -03:00
|
|
|
|
2000-07-29 21:08:13 -03:00
|
|
|
dir = convert_path(words[1])
|
2000-07-29 21:21:36 -03:00
|
|
|
patterns = map(convert_path, words[2:])
|
2000-07-29 21:04:17 -03:00
|
|
|
|
2000-07-29 21:21:36 -03:00
|
|
|
elif action in ('graft', 'prune'):
|
2000-09-25 23:12:31 -03:00
|
|
|
if len(words) != 2:
|
2000-07-29 21:21:36 -03:00
|
|
|
raise DistutilsTemplateError, \
|
|
|
|
"'%s' expects a single <dir_pattern>" % action
|
2000-07-29 21:04:17 -03:00
|
|
|
|
2000-07-29 21:21:36 -03:00
|
|
|
dir_pattern = convert_path(words[1])
|
2000-07-29 21:08:13 -03:00
|
|
|
|
|
|
|
else:
|
2000-07-29 21:21:36 -03:00
|
|
|
raise DistutilsTemplateError, "unknown action '%s'" % action
|
|
|
|
|
2000-07-29 22:04:22 -03:00
|
|
|
return (action, patterns, dir, dir_pattern)
|
2000-07-29 21:21:36 -03:00
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def process_template_line(self, line):
|
2000-07-29 21:21:36 -03:00
|
|
|
# Parse the line: split it up, make sure the right number of words
|
2000-07-29 21:36:25 -03:00
|
|
|
# is there, and return the relevant words. 'action' is always
|
2000-07-29 21:21:36 -03:00
|
|
|
# defined: it's the first word of the line. Which of the other
|
|
|
|
# three are defined depends on the action; it'll be either
|
|
|
|
# patterns, (dir and patterns), or (dir_pattern).
|
2009-08-18 05:16:33 -03:00
|
|
|
action, patterns, dir, dir_pattern = self._parse_template_line(line)
|
2000-07-29 21:08:13 -03:00
|
|
|
|
|
|
|
# OK, now we know that the action is valid and we have the
|
|
|
|
# right number of words on the line for that action -- so we
|
2000-07-29 21:21:36 -03:00
|
|
|
# can proceed with minimal error-checking.
|
2000-07-29 21:08:13 -03:00
|
|
|
if action == 'include':
|
2009-08-17 18:48:22 -03:00
|
|
|
self.debug_print("include " + ' '.join(patterns))
|
2000-07-29 21:21:36 -03:00
|
|
|
for pattern in patterns:
|
2000-09-25 23:12:31 -03:00
|
|
|
if not self.include_pattern(pattern, anchor=1):
|
2002-06-04 17:14:43 -03:00
|
|
|
log.warn("warning: no files found matching '%s'",
|
|
|
|
pattern)
|
2000-07-29 21:08:13 -03:00
|
|
|
|
|
|
|
elif action == 'exclude':
|
2009-08-17 18:48:22 -03:00
|
|
|
self.debug_print("exclude " + ' '.join(patterns))
|
2000-07-29 21:21:36 -03:00
|
|
|
for pattern in patterns:
|
2000-09-25 23:12:31 -03:00
|
|
|
if not self.exclude_pattern(pattern, anchor=1):
|
2002-06-04 17:14:43 -03:00
|
|
|
log.warn(("warning: no previously-included files "
|
|
|
|
"found matching '%s'"), pattern)
|
2000-07-29 21:08:13 -03:00
|
|
|
|
|
|
|
elif action == 'global-include':
|
2009-08-17 18:48:22 -03:00
|
|
|
self.debug_print("global-include " + ' '.join(patterns))
|
2000-07-29 21:21:36 -03:00
|
|
|
for pattern in patterns:
|
2000-09-25 23:12:31 -03:00
|
|
|
if not self.include_pattern(pattern, anchor=0):
|
2002-06-04 17:14:43 -03:00
|
|
|
log.warn(("warning: no files found matching '%s' " +
|
|
|
|
"anywhere in distribution"), pattern)
|
2000-07-29 21:08:13 -03:00
|
|
|
|
|
|
|
elif action == 'global-exclude':
|
2009-08-17 18:48:22 -03:00
|
|
|
self.debug_print("global-exclude " + ' '.join(patterns))
|
2000-07-29 21:21:36 -03:00
|
|
|
for pattern in patterns:
|
2000-09-25 23:12:31 -03:00
|
|
|
if not self.exclude_pattern(pattern, anchor=0):
|
2002-06-04 17:14:43 -03:00
|
|
|
log.warn(("warning: no previously-included files matching "
|
|
|
|
"'%s' found anywhere in distribution"),
|
|
|
|
pattern)
|
2000-07-29 21:08:13 -03:00
|
|
|
|
|
|
|
elif action == 'recursive-include':
|
|
|
|
self.debug_print("recursive-include %s %s" %
|
2009-08-17 18:48:22 -03:00
|
|
|
(dir, ' '.join(patterns)))
|
2000-07-29 21:21:36 -03:00
|
|
|
for pattern in patterns:
|
2000-09-25 23:12:31 -03:00
|
|
|
if not self.include_pattern(pattern, prefix=dir):
|
2004-05-31 12:12:27 -03:00
|
|
|
log.warn(("warning: no files found matching '%s' " +
|
2004-07-18 03:16:08 -03:00
|
|
|
"under directory '%s'"),
|
2002-06-04 17:14:43 -03:00
|
|
|
pattern, dir)
|
2000-07-29 21:08:13 -03:00
|
|
|
|
|
|
|
elif action == 'recursive-exclude':
|
|
|
|
self.debug_print("recursive-exclude %s %s" %
|
2009-08-17 18:48:22 -03:00
|
|
|
(dir, ' '.join(patterns)))
|
2000-07-29 21:21:36 -03:00
|
|
|
for pattern in patterns:
|
2000-07-29 21:08:13 -03:00
|
|
|
if not self.exclude_pattern(pattern, prefix=dir):
|
2002-06-04 17:14:43 -03:00
|
|
|
log.warn(("warning: no previously-included files matching "
|
|
|
|
"'%s' found under directory '%s'"),
|
|
|
|
pattern, dir)
|
2001-12-06 16:51:35 -04:00
|
|
|
|
2000-07-29 21:08:13 -03:00
|
|
|
elif action == 'graft':
|
|
|
|
self.debug_print("graft " + dir_pattern)
|
2000-07-29 21:36:25 -03:00
|
|
|
if not self.include_pattern(None, prefix=dir_pattern):
|
2002-06-04 17:14:43 -03:00
|
|
|
log.warn("warning: no directories found matching '%s'",
|
|
|
|
dir_pattern)
|
2000-07-29 21:08:13 -03:00
|
|
|
|
|
|
|
elif action == 'prune':
|
|
|
|
self.debug_print("prune " + dir_pattern)
|
|
|
|
if not self.exclude_pattern(None, prefix=dir_pattern):
|
2002-06-04 17:14:43 -03:00
|
|
|
log.warn(("no previously-included directories found " +
|
|
|
|
"matching '%s'"), dir_pattern)
|
2000-07-29 21:08:13 -03:00
|
|
|
else:
|
2000-07-29 21:21:36 -03:00
|
|
|
raise DistutilsInternalError, \
|
2000-07-29 21:08:13 -03:00
|
|
|
"this cannot happen: invalid action '%s'" % action
|
2000-07-29 21:04:17 -03:00
|
|
|
|
2000-07-29 22:45:42 -03:00
|
|
|
# -- Filtering/selection methods -----------------------------------
|
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def include_pattern(self, pattern, anchor=1, prefix=None, is_regex=0):
|
2000-07-29 21:36:25 -03:00
|
|
|
"""Select strings (presumably filenames) from 'self.files' that
|
2009-08-17 18:48:22 -03:00
|
|
|
match 'pattern', a Unix-style wildcard (glob) pattern.
|
|
|
|
|
|
|
|
Patterns are not quite the same as implemented by the 'fnmatch'
|
|
|
|
module: '*' and '?' match non-special characters, where "special"
|
|
|
|
is platform-dependent: slash on Unix; colon, slash, and backslash on
|
2000-07-29 21:36:25 -03:00
|
|
|
DOS/Windows; and colon on Mac OS.
|
2000-07-29 21:04:17 -03:00
|
|
|
|
|
|
|
If 'anchor' is true (the default), then the pattern match is more
|
|
|
|
stringent: "*.py" will match "foo.py" but not "foo/bar.py". If
|
|
|
|
'anchor' is false, both of these will match.
|
|
|
|
|
|
|
|
If 'prefix' is supplied, then only filenames starting with 'prefix'
|
|
|
|
(itself a pattern) and ending with 'pattern', with anything in between
|
|
|
|
them, will match. 'anchor' is ignored in this case.
|
|
|
|
|
|
|
|
If 'is_regex' is true, 'anchor' and 'prefix' are ignored, and
|
|
|
|
'pattern' is assumed to be either a string containing a regex or a
|
|
|
|
regex object -- no translation is done, the regex is just compiled
|
|
|
|
and used as-is.
|
|
|
|
|
|
|
|
Selected strings will be added to self.files.
|
|
|
|
|
|
|
|
Return 1 if files are found.
|
|
|
|
"""
|
|
|
|
files_found = 0
|
2000-09-25 23:12:31 -03:00
|
|
|
pattern_re = translate_pattern(pattern, anchor, prefix, is_regex)
|
2000-07-29 21:36:25 -03:00
|
|
|
self.debug_print("include_pattern: applying regex r'%s'" %
|
2000-07-29 21:04:17 -03:00
|
|
|
pattern_re.pattern)
|
|
|
|
|
|
|
|
# delayed loading of allfiles list
|
2000-07-29 22:45:42 -03:00
|
|
|
if self.allfiles is None:
|
|
|
|
self.findall()
|
2000-07-29 21:04:17 -03:00
|
|
|
|
|
|
|
for name in self.allfiles:
|
2000-09-25 23:12:31 -03:00
|
|
|
if pattern_re.search(name):
|
2000-07-29 21:04:17 -03:00
|
|
|
self.debug_print(" adding " + name)
|
2000-09-25 23:12:31 -03:00
|
|
|
self.files.append(name)
|
2000-07-29 21:04:17 -03:00
|
|
|
files_found = 1
|
2001-12-06 16:51:35 -04:00
|
|
|
|
2000-07-29 21:04:17 -03:00
|
|
|
return files_found
|
|
|
|
|
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def exclude_pattern(self, pattern, anchor=1, prefix=None, is_regex=0):
|
2000-07-29 21:04:17 -03:00
|
|
|
"""Remove strings (presumably filenames) from 'files' that match
|
2009-08-17 18:48:22 -03:00
|
|
|
'pattern'.
|
|
|
|
|
|
|
|
Other parameters are the same as for 'include_pattern()', above.
|
|
|
|
The list 'self.files' is modified in place. Return 1 if files are
|
|
|
|
found.
|
2000-07-29 21:04:17 -03:00
|
|
|
"""
|
|
|
|
files_found = 0
|
2000-09-25 23:12:31 -03:00
|
|
|
pattern_re = translate_pattern(pattern, anchor, prefix, is_regex)
|
2000-07-29 21:04:17 -03:00
|
|
|
self.debug_print("exclude_pattern: applying regex r'%s'" %
|
|
|
|
pattern_re.pattern)
|
2000-09-25 23:12:31 -03:00
|
|
|
for i in range(len(self.files)-1, -1, -1):
|
|
|
|
if pattern_re.search(self.files[i]):
|
2000-07-29 21:04:17 -03:00
|
|
|
self.debug_print(" removing " + self.files[i])
|
|
|
|
del self.files[i]
|
|
|
|
files_found = 1
|
2001-12-06 16:51:35 -04:00
|
|
|
|
2000-07-29 21:04:17 -03:00
|
|
|
return files_found
|
|
|
|
|
|
|
|
|
|
|
|
# ----------------------------------------------------------------------
|
|
|
|
# Utility functions
|
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def findall(dir = os.curdir):
|
2000-07-29 21:04:17 -03:00
|
|
|
"""Find all files under 'dir' and return the list of full filenames
|
|
|
|
(relative to 'dir').
|
|
|
|
"""
|
|
|
|
from stat import ST_MODE, S_ISREG, S_ISDIR, S_ISLNK
|
|
|
|
|
|
|
|
list = []
|
|
|
|
stack = [dir]
|
|
|
|
pop = stack.pop
|
|
|
|
push = stack.append
|
|
|
|
|
|
|
|
while stack:
|
|
|
|
dir = pop()
|
2000-09-25 23:12:31 -03:00
|
|
|
names = os.listdir(dir)
|
2000-07-29 21:04:17 -03:00
|
|
|
|
|
|
|
for name in names:
|
|
|
|
if dir != os.curdir: # avoid the dreaded "./" syndrome
|
2000-09-25 23:12:31 -03:00
|
|
|
fullname = os.path.join(dir, name)
|
2000-07-29 21:04:17 -03:00
|
|
|
else:
|
|
|
|
fullname = name
|
|
|
|
|
|
|
|
# Avoid excess stat calls -- just one will do, thank you!
|
|
|
|
stat = os.stat(fullname)
|
|
|
|
mode = stat[ST_MODE]
|
|
|
|
if S_ISREG(mode):
|
2000-09-25 23:12:31 -03:00
|
|
|
list.append(fullname)
|
2000-07-29 21:04:17 -03:00
|
|
|
elif S_ISDIR(mode) and not S_ISLNK(mode):
|
2000-09-25 23:12:31 -03:00
|
|
|
push(fullname)
|
2000-07-29 21:04:17 -03:00
|
|
|
|
|
|
|
return list
|
|
|
|
|
|
|
|
|
2009-04-05 18:44:08 -03:00
|
|
|
def glob_to_re(pattern):
|
2009-08-17 18:48:22 -03:00
|
|
|
"""Translate a shell-like glob pattern to a regular expression.
|
|
|
|
|
|
|
|
Return a string containing the regex. Differs from
|
|
|
|
'fnmatch.translate()' in that '*' does not match "special characters"
|
|
|
|
(which are platform-specific).
|
2000-07-29 21:04:17 -03:00
|
|
|
"""
|
2000-09-25 23:12:31 -03:00
|
|
|
pattern_re = fnmatch.translate(pattern)
|
2000-07-29 21:04:17 -03:00
|
|
|
|
|
|
|
# '?' and '*' in the glob pattern become '.' and '.*' in the RE, which
|
|
|
|
# IMHO is wrong -- '?' and '*' aren't supposed to match slash in Unix,
|
|
|
|
# and by extension they shouldn't match such "special characters" under
|
|
|
|
# any OS. So change all non-escaped dots in the RE to match any
|
|
|
|
# character except the special characters.
|
|
|
|
# XXX currently the "special characters" are just slash -- i.e. this is
|
|
|
|
# Unix-only.
|
2009-04-05 18:44:08 -03:00
|
|
|
pattern_re = re.sub(r'((?<!\\)(\\\\)*)\.', r'\1[^/]', pattern_re)
|
|
|
|
|
2000-07-29 21:04:17 -03:00
|
|
|
return pattern_re
|
|
|
|
|
|
|
|
|
2009-08-17 18:48:22 -03:00
|
|
|
def translate_pattern(pattern, anchor=1, prefix=None, is_regex=0):
|
2000-07-29 21:04:17 -03:00
|
|
|
"""Translate a shell-like wildcard pattern to a compiled regular
|
2009-08-17 18:48:22 -03:00
|
|
|
expression.
|
|
|
|
|
|
|
|
Return the compiled regex. If 'is_regex' true,
|
2000-07-29 21:04:17 -03:00
|
|
|
then 'pattern' is directly compiled to a regex (if it's a string)
|
|
|
|
or just returned as-is (assumes it's a regex object).
|
|
|
|
"""
|
|
|
|
if is_regex:
|
2009-08-17 18:48:22 -03:00
|
|
|
if isinstance(pattern, str):
|
2000-07-29 21:04:17 -03:00
|
|
|
return re.compile(pattern)
|
|
|
|
else:
|
|
|
|
return pattern
|
|
|
|
|
|
|
|
if pattern:
|
2000-09-25 23:12:31 -03:00
|
|
|
pattern_re = glob_to_re(pattern)
|
2000-07-29 21:04:17 -03:00
|
|
|
else:
|
|
|
|
pattern_re = ''
|
2001-12-06 16:51:35 -04:00
|
|
|
|
2000-07-29 21:04:17 -03:00
|
|
|
if prefix is not None:
|
2009-08-17 18:28:34 -03:00
|
|
|
# ditch end of pattern character
|
|
|
|
empty_pattern = glob_to_re('')
|
2009-08-17 18:48:22 -03:00
|
|
|
prefix_re = glob_to_re(prefix)[:-len(empty_pattern)]
|
2000-09-25 23:12:31 -03:00
|
|
|
pattern_re = "^" + os.path.join(prefix_re, ".*" + pattern_re)
|
2000-07-29 21:04:17 -03:00
|
|
|
else: # no prefix -- respect anchor flag
|
|
|
|
if anchor:
|
|
|
|
pattern_re = "^" + pattern_re
|
2001-12-06 16:51:35 -04:00
|
|
|
|
2000-09-25 23:12:31 -03:00
|
|
|
return re.compile(pattern_re)
|