Filtering os.walk() dirs and files

I'm looking for a way to include/exclude files patterns and exclude directories from a os.walk() call.

Here's what I'm doing by now:

import fnmatch
import os

includes = ['*.doc', '*.odt']
excludes = ['/home/paulo-freitas/Documents']

def _filter(paths):
    for path in paths:
        if os.path.isdir(path) and not path in excludes:
            yield path

        for pattern in (includes + excludes):
            if not os.path.isdir(path) and fnmatch.fnmatch(path, pattern):
                yield path

for root, dirs, files in os.walk('/home/paulo-freitas'):
    dirs[:] = _filter(map(lambda d: os.path.join(root, d), dirs))
    files[:] = _filter(map(lambda f: os.path.join(root, f), files))

    for filename in files:
        filename = os.path.join(root, filename)

        print(filename)

Is there a better way to do this? How?


This solution uses fnmatch.translate to convert glob patterns to regular expressions (it assumes the includes only is used for files):

import fnmatch
import os
import os.path
import re

includes = ['*.doc', '*.odt'] # for files only
excludes = ['/home/paulo-freitas/Documents'] # for dirs and files

# transform glob patterns to regular expressions
includes = r'|'.join([fnmatch.translate(x) for x in includes])
excludes = r'|'.join([fnmatch.translate(x) for x in excludes]) or r'$.'

for root, dirs, files in os.walk('/home/paulo-freitas'):

    # exclude dirs
    dirs[:] = [os.path.join(root, d) for d in dirs]
    dirs[:] = [d for d in dirs if not re.match(excludes, d)]

    # exclude/include files
    files = [os.path.join(root, f) for f in files]
    files = [f for f in files if not re.match(excludes, f)]
    files = [f for f in files if re.match(includes, f)]

    for fname in files:
        print fname

From docs.python.org:

os.walk(top[, topdown=True[, onerror=None[, followlinks=False]]])

When topdown is True, the caller can modify the dirnames list in-place … this can be used to prune the search …

for root, dirs, files in os.walk('/home/paulo-freitas', topdown=True):
    # excludes can be done with fnmatch.filter and complementary set,
    # but it's more annoying to read.
    dirs[:] = [d for d in dirs if d not in excludes] 
    for pat in includes:
        for f in fnmatch.filter(files, pat):
            print os.path.join(root, f)

I should point out that the above code assumes excludes is a pattern, not a full path. You would need to adjust the list comprehension to filter if os.path.join(root, d) not in excludes to match the OP case.


why fnmatch?

import os
excludes=....
for ROOT,DIR,FILES in os.walk("/path"):
    for file in FILES:
       if file.endswith(('doc','odt')):
          print file
    for directory in DIR:
       if not directory in excludes :
          print directory

not exhaustively tested