else/SpinalTap/spinal.py

729 lines
21 KiB
Python
Raw Normal View History

2016-03-07 20:25:50 +00:00
import collections
import glob
2016-02-21 07:13:50 +00:00
import json
import os
import shutil
2016-03-07 20:25:50 +00:00
import stat
import string
2016-02-21 07:13:50 +00:00
import sys
import time
2016-07-28 03:41:13 +00:00
sys.path.append('C:\\git\\else\\ratelimiter'); import ratelimiter
BYTE = 1
2016-02-21 07:13:50 +00:00
KIBIBYTE = BYTE * 1024
MIBIBYTE = KIBIBYTE * 1024
GIBIBYTE = MIBIBYTE * 1024
TEBIBYTE = GIBIBYTE * 1024
2016-03-07 20:25:50 +00:00
SIZE_UNITS = (TEBIBYTE, GIBIBYTE, MIBIBYTE, KIBIBYTE, BYTE)
2016-03-07 20:25:50 +00:00
UNIT_STRINGS = {
BYTE: 'b',
KIBIBYTE: 'KiB',
MIBIBYTE: 'MiB',
GIBIBYTE: 'GiB',
TEBIBYTE: 'TiB',
}
CHUNK_SIZE = 128 * KIBIBYTE
# Number of bytes to read and write at a time
2016-02-21 07:13:50 +00:00
class DestinationIsDirectory(Exception):
pass
class DestinationIsFile(Exception):
pass
class RecursiveDirectory(Exception):
pass
class SourceNotDirectory(Exception):
pass
class SourceNotFile(Exception):
pass
class SpinalError(Exception):
2016-02-21 07:13:50 +00:00
pass
2016-03-07 20:25:50 +00:00
class FilePath:
2016-05-10 08:00:29 +00:00
'''
Class for consolidating lots of `os.path` operations,
and caching `os.stat` results.
'''
2016-03-07 20:25:50 +00:00
def __init__(self, path):
self.path = os.path.abspath(path)
self._stat = None
self._isdir = None
self._isfile = None
self._islink = None
self._size = None
def __hash__(self):
return self.path.__hash__()
def __repr__(self):
2016-05-10 08:00:29 +00:00
return 'FilePath(%s)' % repr(self.path)
2016-03-07 20:25:50 +00:00
2016-07-05 07:24:08 +00:00
@property
def basename(self):
return os.path.basename(self.path)
2016-03-07 20:25:50 +00:00
@property
def isdir(self):
return self.type_getter('_isdir', stat.S_ISDIR)
@property
def isfile(self):
return self.type_getter('_isfile', stat.S_ISREG)
@property
def islink(self):
return self.type_getter('_islink', stat.S_ISLNK)
@property
def size(self):
if self._size is None:
if self.stat is False:
self._size = None
else:
self._size = self.stat.st_size
return self._size
@property
def stat(self):
if self._stat is None:
try:
self._stat = os.stat(self.path)
except FileNotFoundError:
self._stat = False
return self._stat
def type_getter(self, attr, resolution):
2016-05-10 08:00:29 +00:00
'''
Try to return the cached type. Call resolution(self.stat.st_mode) if
we don't have the stat data yet.
'''
value = getattr(self, attr)
if value is None:
2016-03-07 20:25:50 +00:00
if self.stat is False:
return False
else:
2016-05-10 08:00:29 +00:00
value = resolution(self.stat.st_mode)
setattr(self, attr, value)
return value
2016-03-07 20:25:50 +00:00
def bytes_to_unit_string(bytes):
size_unit = 1
for unit in SIZE_UNITS:
if bytes >= unit:
size_unit = unit
break
size_unit_string = UNIT_STRINGS[size_unit]
size_string = '%.3f %s' % ((bytes / size_unit), size_unit_string)
return size_string
2016-02-21 07:13:50 +00:00
def callback_exclusion(name, path_type):
'''
Example of an exclusion callback function.
'''
2016-05-10 08:00:29 +00:00
print('Excluding', path_type, name)
2016-03-07 20:25:50 +00:00
def callback_v1(fpobj, written_bytes, total_bytes):
2016-02-21 07:13:50 +00:00
'''
Example of a copy callback function.
2016-05-10 08:00:29 +00:00
Prints "filename written/total (percent%)"
'''
2016-03-07 20:25:50 +00:00
filename = fpobj.path.encode('ascii', 'replace').decode()
2016-02-21 07:13:50 +00:00
if written_bytes >= total_bytes:
ends = '\n'
else:
ends = ''
percent = (100 * written_bytes) / total_bytes
2016-03-07 20:25:50 +00:00
percent = '%07.3f' % percent
2016-02-21 07:13:50 +00:00
written = '{:,}'.format(written_bytes)
total = '{:,}'.format(total_bytes)
written = written.rjust(len(total), ' ')
status = '{filename} {written}/{total} ({percent}%)\r'
status = status.format(filename=filename, written=written, total=total, percent=percent)
print(status, end=ends)
sys.stdout.flush()
2016-03-07 20:25:50 +00:00
def copy(source, file_args=None, file_kwargs=None, dir_args=None, dir_kwargs=None):
'''
Perform copy_dir or copy_file as appropriate for the source path.
'''
source = str_to_fp(source)
if source.isfile:
file_args = file_args or tuple()
file_kwargs = file_kwargs or dict()
return copy_file(source, *file_args, **file_kwargs)
elif source.isdir:
dir_args = dir_args or tuple()
dir_kwargs = dir_kwargs or dict()
return copy_dir(source, *dir_args, **dir_kwargs)
raise SpinalError('Neither file nor dir: %s' % source)
def copy_dir(
2016-02-21 07:13:50 +00:00
source,
2016-02-21 07:18:26 +00:00
destination=None,
destination_new_root=None,
2016-02-21 07:13:50 +00:00
bytes_per_second=None,
2016-03-07 20:25:50 +00:00
callback_directory=None,
2016-05-10 08:00:29 +00:00
callback_exclusion=None,
2016-03-07 20:25:50 +00:00
callback_file=None,
callback_permission_denied=None,
2016-05-10 08:00:29 +00:00
callback_verbose=None,
2016-02-21 07:13:50 +00:00
dry_run=False,
2016-03-07 20:25:50 +00:00
exclude_directories=None,
exclude_filenames=None,
2016-05-10 08:00:29 +00:00
files_per_second=None,
2016-02-21 07:13:50 +00:00
overwrite_old=True,
2016-03-07 20:25:50 +00:00
precalcsize=False,
2016-02-21 07:13:50 +00:00
):
'''
2016-03-07 20:25:50 +00:00
Copy all of the contents from source to destination,
including subdirectories.
2016-02-21 07:13:50 +00:00
source:
2016-03-07 20:25:50 +00:00
The directory which will be copied.
2016-02-21 07:13:50 +00:00
destination:
2016-03-07 20:25:50 +00:00
The directory in which copied files are placed. Alternatively, use
2016-02-21 07:18:26 +00:00
destination_new_root.
destination_new_root:
Determine the destination path by calling
2016-03-07 20:25:50 +00:00
`new_root(source, destination_new_root)`.
2016-02-21 07:18:26 +00:00
Thus, this path acts as a root and the rest of the path is matched.
2016-02-21 07:13:50 +00:00
2016-05-10 08:00:29 +00:00
`destination` and `destination_new_root` are mutually exclusive.
2016-02-21 07:13:50 +00:00
bytes_per_second:
Restrict file copying to this many bytes per second. Can be an integer
or an existing Ratelimiter object.
The provided BYTE, KIBIBYTE, etc constants may help.
Default = None
2016-03-07 20:25:50 +00:00
callback_directory:
This function will be called after each file copy with three parameters:
name of file copied, number of bytes written to destination so far,
total bytes needed (from precalcsize).
2016-05-10 08:00:29 +00:00
If `precalcsize` is False, this function will receive written bytes
for both written and total, showing 100% always.
Default = None
callback_exclusion:
Passed directly into `walk_generator`.
2016-03-07 20:25:50 +00:00
Default = None
callback_file:
Will be passed into each individual `copy_file` operation as the
`callback` for that file.
Default = None
callback_permission_denied:
Will be passed into each individual `copy_file` operation as the
`callback_permission_denied` for that file.
2016-02-21 07:13:50 +00:00
Default = None
2016-05-10 08:00:29 +00:00
callback_verbose:
If provided, this function will be called with some operation notes.
Default = None
2016-02-21 07:13:50 +00:00
dry_run:
Do everything except the actual file copying.
2016-02-21 07:13:50 +00:00
Default = False
2016-03-07 20:25:50 +00:00
exclude_filenames:
Passed directly into `walk_generator`.
Default = None
exclude_directories:
Passed directly into `walk_generator`.
Default = None
2016-05-10 08:00:29 +00:00
files_per_second:
Maximum number of files to be processed per second. Helps to keep CPU usage
low.
2016-03-07 20:25:50 +00:00
Default = None
2016-02-21 07:13:50 +00:00
overwrite_old:
If True, overwrite the destination file if the source file
has a more recent "last modified" timestamp.
2016-02-21 07:13:50 +00:00
Default = True
2016-03-07 20:25:50 +00:00
precalcsize:
If True, calculate the size of source before beginning the
operation. This number can be used in the callback_directory function.
Else, callback_directory will receive written bytes as total bytes
(showing 100% always).
This can take a long time.
Default = False
Returns: [destination path, number of bytes written to destination]
(Written bytes is 0 if all files already existed.)
'''
2016-03-07 20:25:50 +00:00
2016-02-21 07:13:50 +00:00
# Prepare parameters
2016-02-21 07:18:26 +00:00
if not is_xor(destination, destination_new_root):
m = 'One and only one of `destination` and '
2016-05-10 08:00:29 +00:00
m += '`destination_new_root` can be passed.'
2016-02-21 07:18:26 +00:00
raise ValueError(m)
2016-03-07 20:25:50 +00:00
source = str_to_fp(source)
source = get_path_casing(source)
2016-02-21 07:18:26 +00:00
if destination_new_root is not None:
destination = new_root(source, destination_new_root)
2016-03-07 20:25:50 +00:00
destination = str_to_fp(destination)
2016-02-21 07:18:26 +00:00
2016-05-10 08:00:29 +00:00
callback_directory = callback_directory or do_nothing
callback_verbose = callback_verbose or do_nothing
2016-03-07 20:25:50 +00:00
if is_subfolder(source, destination):
raise RecursiveDirectory(source, destination)
2016-03-07 20:25:50 +00:00
if not source.isdir:
raise SourceNotDirectory(source)
2016-02-21 07:13:50 +00:00
2016-03-07 20:25:50 +00:00
if destination.isfile:
raise DestinationIsFile(destination)
if precalcsize is True:
total_bytes = get_dir_size(source)
else:
total_bytes = 0
2016-02-21 07:13:50 +00:00
2016-05-10 08:00:29 +00:00
bytes_per_second = limiter_or_none(bytes_per_second)
files_per_second = limiter_or_none(files_per_second)
2016-02-21 07:13:50 +00:00
2016-03-07 20:25:50 +00:00
# Copy
written_bytes = 0
walker = walk_generator(
source,
2016-05-10 08:00:29 +00:00
callback_exclusion=callback_exclusion,
callback_verbose=callback_verbose,
2016-03-07 20:25:50 +00:00
exclude_directories=exclude_directories,
exclude_filenames=exclude_filenames,
)
for (source_abspath) in walker:
# Terminology:
# abspath: C:\folder\subfolder\filename.txt
# location: C:\folder\subfolder
# base_name: filename.txt
# folder: subfolder
2016-02-21 07:13:50 +00:00
2016-03-07 20:25:50 +00:00
destination_abspath = source_abspath.path.replace(source.path, destination.path)
destination_abspath = str_to_fp(destination_abspath)
2016-02-21 07:13:50 +00:00
2016-03-07 20:25:50 +00:00
if destination_abspath.isdir:
raise DestinationIsDirectory(destination_abspath)
2016-02-21 07:13:50 +00:00
2016-03-07 20:25:50 +00:00
destination_location = os.path.split(destination_abspath.path)[0]
2016-05-10 08:00:29 +00:00
os.makedirs(destination_location, exist_ok=True)
2016-03-07 20:25:50 +00:00
copied = copy_file(
source_abspath,
destination_abspath,
2016-05-10 08:00:29 +00:00
bytes_per_second=bytes_per_second,
2016-03-07 20:25:50 +00:00
callback=callback_file,
callback_permission_denied=callback_permission_denied,
2016-05-10 08:00:29 +00:00
callback_verbose=callback_verbose,
2016-03-07 20:25:50 +00:00
dry_run=dry_run,
overwrite_old=overwrite_old,
)
2016-02-21 07:13:50 +00:00
2016-03-07 20:25:50 +00:00
copiedname = copied[0]
written_bytes += copied[1]
2016-02-21 07:13:50 +00:00
2016-05-10 08:00:29 +00:00
if precalcsize is False:
callback_directory(copiedname, written_bytes, written_bytes)
else:
callback_directory(copiedname, written_bytes, total_bytes)
if files_per_second is not None:
files_per_second.limit(1)
2016-02-21 07:13:50 +00:00
return [destination, written_bytes]
2016-03-07 20:25:50 +00:00
def copy_file(
source,
destination=None,
2016-02-21 07:13:50 +00:00
destination_new_root=None,
bytes_per_second=None,
2016-03-07 20:25:50 +00:00
callback=None,
2016-05-10 08:00:29 +00:00
callback_verbose=None,
2016-02-21 07:13:50 +00:00
dry_run=False,
overwrite_old=True,
2016-03-07 20:25:50 +00:00
callback_permission_denied=None,
2016-02-21 07:13:50 +00:00
):
'''
2016-03-07 20:25:50 +00:00
Copy a file from one place to another.
2016-03-07 20:25:50 +00:00
source:
The file to copy.
2016-03-07 20:25:50 +00:00
destination:
The filename of the new copy. Alternatively, use
2016-02-21 07:13:50 +00:00
destination_new_root.
2016-02-21 07:13:50 +00:00
destination_new_root:
Determine the destination path by calling
`new_root(source_dir, destination_new_root)`.
Thus, this path acts as a root and the rest of the path is matched.
bytes_per_second:
Restrict file copying to this many bytes per second. Can be an integer
or an existing Ratelimiter object.
The provided BYTE, KIBIBYTE, etc constants may help.
Default = None
2016-03-07 20:25:50 +00:00
callback:
If provided, this function will be called after writing
each CHUNK_SIZE bytes to destination with three parameters:
the FilePath object being copied, number of bytes written so far,
total number of bytes needed.
2016-02-21 07:13:50 +00:00
Default = None
2016-03-07 20:25:50 +00:00
callback_permission_denied:
If provided, this function will be called when a source file denies
read access, with the file path and the exception object as parameters.
THE OPERATION WILL RETURN NORMALLY.
If not provided, the PermissionError is raised.
2016-02-21 07:13:50 +00:00
Default = None
2016-05-10 08:00:29 +00:00
callback_verbose:
If provided, this function will be called with some operation notes.
Default = None
2016-02-21 07:13:50 +00:00
dry_run:
Do everything except the actual file copying.
2016-02-21 07:13:50 +00:00
Default = False
2016-02-21 07:13:50 +00:00
overwrite_old:
If True, overwrite the destination file if the source file
has a more recent "last modified" timestamp.
2016-02-21 07:13:50 +00:00
Default = True
2016-03-07 20:25:50 +00:00
Returns: [destination filename, number of bytes written to destination]
(Written bytes is 0 if the file already existed.)
'''
2016-02-21 07:13:50 +00:00
# Prepare parameters
2016-03-07 20:25:50 +00:00
if not is_xor(destination, destination_new_root):
m = 'One and only one of `destination` and '
2016-02-21 07:13:50 +00:00
m += '`destination_new_root` can be passed'
raise ValueError(m)
2016-03-07 20:25:50 +00:00
source = str_to_fp(source)
source = get_path_casing(source)
2016-02-21 07:13:50 +00:00
2016-03-07 20:25:50 +00:00
if destination_new_root is not None:
destination = new_root(source, destination_new_root)
destination = str_to_fp(destination)
2016-02-21 07:13:50 +00:00
2016-05-10 08:00:29 +00:00
callback = callback or do_nothing
callback_verbose = callback_verbose or do_nothing
2016-03-07 20:25:50 +00:00
if not source.isfile:
raise SourceNotFile(source)
2016-03-07 20:25:50 +00:00
if destination.isdir:
raise DestinationIsDirectory(destination)
2016-02-21 07:13:50 +00:00
2016-05-10 08:00:29 +00:00
bytes_per_second = limiter_or_none(bytes_per_second)
2016-03-07 20:25:50 +00:00
# Determine overwrite
if destination.stat is not False:
destination_modtime = destination.stat.st_mtime
if overwrite_old is False:
return [destination, 0]
source_modtime = source.stat.st_mtime
if source_modtime == destination_modtime:
return [destination, 0]
2016-02-21 07:13:50 +00:00
# Copy
2016-03-07 20:25:50 +00:00
if dry_run:
if callback is not None:
callback(destination, 0, 0)
return [destination, 0]
source_bytes = source.size
destination_location = os.path.split(destination.path)[0]
2016-05-10 08:00:29 +00:00
os.makedirs(destination_location, exist_ok=True)
2016-02-21 07:13:50 +00:00
written_bytes = 0
2016-03-07 20:25:50 +00:00
try:
2016-05-10 08:00:29 +00:00
callback_verbose('Opening handles.')
2016-03-07 20:25:50 +00:00
source_file = open(source.path, 'rb')
destination_file = open(destination.path, 'wb')
except PermissionError as exception:
if callback_permission_denied is not None:
callback_permission_denied(source, exception)
return [destination, 0]
else:
raise
2016-02-21 07:13:50 +00:00
2016-03-07 20:25:50 +00:00
while True:
data_chunk = source_file.read(CHUNK_SIZE)
data_bytes = len(data_chunk)
if data_bytes == 0:
break
2016-02-21 07:13:50 +00:00
2016-03-07 20:25:50 +00:00
destination_file.write(data_chunk)
written_bytes += data_bytes
2016-02-21 07:13:50 +00:00
2016-05-10 08:00:29 +00:00
if bytes_per_second is not None:
bytes_per_second.limit(data_bytes)
2016-05-10 08:00:29 +00:00
callback(destination, written_bytes, source_bytes)
2016-03-07 20:25:50 +00:00
# Fin
2016-05-10 08:00:29 +00:00
callback_verbose('Closing handles.')
2016-03-07 20:25:50 +00:00
source_file.close()
destination_file.close()
2016-05-10 08:00:29 +00:00
callback_verbose('Copying metadata')
2016-03-07 20:25:50 +00:00
shutil.copystat(source.path, destination.path)
return [destination, written_bytes]
2016-05-10 08:00:29 +00:00
def do_nothing(*args):
'''
Used by other functions as the default callback.
'''
return
2016-03-07 20:25:50 +00:00
def get_path_casing(path):
'''
2016-03-07 20:25:50 +00:00
Take what is perhaps incorrectly cased input and get the path's actual
casing according to the filesystem.
2016-05-10 08:00:29 +00:00
Thank you:
2016-03-07 20:25:50 +00:00
Ethan Furman http://stackoverflow.com/a/7133137/5430534
xvorsx http://stackoverflow.com/a/14742779/5430534
'''
p = str_to_fp(path)
path = p.path
(drive, subpath) = os.path.splitdrive(path)
2016-07-10 04:38:49 +00:00
subpath = subpath.lstrip(os.sep)
def patternize(piece):
'''
Create a pattern like "[u]ser" from "user", forcing glob to look up the
correct path name, and guaranteeing that the only result will be the correct path.
Special cases are:
!, because in glob syntax, [!x] tells glob to look for paths that don't contain
"x". [!] is invalid syntax, so we pick the first non-! character to put
in the brackets.
[, because this starts a capture group
'''
piece = glob.escape(piece)
for character in piece:
2016-07-20 03:31:47 +00:00
if character not in '![]':
2016-07-10 04:38:49 +00:00
replacement = '[%s]' % character
2016-07-20 03:31:47 +00:00
#print(piece, character, replacement)
2016-07-10 04:38:49 +00:00
piece = piece.replace(character, replacement, 1)
break
return piece
pattern = [patternize(piece) for piece in subpath.split(os.sep)]
2016-03-07 20:25:50 +00:00
pattern = os.sep.join(pattern)
pattern = drive.upper() + os.sep + pattern
2016-07-20 03:31:47 +00:00
#print(pattern)
2016-03-07 20:25:50 +00:00
try:
return str_to_fp(glob.glob(pattern)[0])
except IndexError:
return p
def get_dir_size(path):
2016-02-21 07:13:50 +00:00
'''
Calculate the total number of bytes across all files in this directory
and its subdirectories.
'''
2016-03-07 20:25:50 +00:00
path = str_to_fp(path)
2016-03-07 20:25:50 +00:00
if not path.isdir:
raise SourceNotDirectory(path)
2016-02-21 07:13:50 +00:00
total_bytes = 0
2016-05-10 08:00:29 +00:00
for filepath in walk_generator(path):
total_bytes += filepath.size
2016-02-21 07:13:50 +00:00
return total_bytes
2016-02-21 07:13:50 +00:00
def is_subfolder(parent, child):
'''
2016-02-21 07:13:50 +00:00
Determine whether parent contains child.
'''
2016-03-07 20:25:50 +00:00
parent = normalize(str_to_fp(parent).path) + os.sep
child = normalize(str_to_fp(child).path) + os.sep
2016-02-21 07:13:50 +00:00
return child.startswith(parent)
2016-02-21 07:13:50 +00:00
def is_xor(*args):
'''
Return True if and only if one arg is truthy.
'''
2016-02-21 07:13:50 +00:00
return [bool(a) for a in args].count(True) == 1
2016-05-10 08:00:29 +00:00
def limiter_or_none(value):
if isinstance(value, ratelimiter.Ratelimiter):
limiter = value
elif value is not None:
limiter = ratelimiter.Ratelimiter(allowance_per_period=value, period=1)
else:
limiter = None
return limiter
2016-02-21 07:13:50 +00:00
def new_root(filepath, root):
'''
Prepend `root` to `filepath`, drive letter included. For example:
"C:\\folder\\subfolder\\file.txt" and "C:\\backups" becomes
"C:\\backups\\C\\folder\\subfolder\\file.txt"
2016-02-21 07:13:50 +00:00
I use this so that my G: drive can have backups from my C: and D: drives
while preserving directory structure in G:\\D and G:\\C.
'''
2016-03-07 20:25:50 +00:00
filepath = str_to_fp(filepath).path
root = str_to_fp(root).path
2016-02-21 07:13:50 +00:00
filepath = filepath.replace(':', os.sep)
filepath = os.path.normpath(filepath)
filepath = os.path.join(root, filepath)
2016-03-07 20:25:50 +00:00
return str_to_fp(filepath)
2016-02-21 07:13:50 +00:00
def normalize(text):
'''
Apply os.path.normpath and os.path.normcase.
'''
return os.path.normpath(os.path.normcase(text))
2016-03-07 20:25:50 +00:00
def str_to_fp(path):
2016-02-21 07:13:50 +00:00
'''
2016-03-07 20:25:50 +00:00
If `path` is a string, create a FilePath object, otherwise just return it.
'''
if isinstance(path, str):
path = FilePath(path)
return path
def walk_generator(
2016-07-05 07:24:08 +00:00
path='.',
2016-05-10 08:00:29 +00:00
callback_exclusion=None,
callback_verbose=None,
2016-03-07 20:25:50 +00:00
exclude_directories=None,
exclude_filenames=None,
):
2016-02-21 07:13:50 +00:00
'''
2016-05-10 08:00:29 +00:00
Yield FilePath objects from the file tree similar to os.walk.
callback_exclusion:
This function will be called when a file or directory is excluded with
two parameters: the path, and 'file' or 'directory'.
Default = None
callback_verbose:
If provided, this function will be called with some operation notes.
Default = None
2016-03-07 20:25:50 +00:00
exclude_filenames:
A set of filenames that will not be copied. Entries can be absolute
paths to exclude that particular file, or plain names to exclude
all matches. For example:
{'C:\\folder\\file.txt', 'desktop.ini'}
Default = None
exclude_directories:
A set of directories that will not be copied. Entries can be
absolute paths to exclude that particular directory, or plain names
to exclude all matches. For example:
{'C:\\folder', 'thumbnails'}
Default = None
'''
if exclude_directories is None:
exclude_directories = set()
if exclude_filenames is None:
exclude_filenames = set()
2016-05-10 08:00:29 +00:00
callback_exclusion = callback_exclusion or do_nothing
callback_verbose = callback_verbose or do_nothing
2016-03-07 20:25:50 +00:00
exclude_filenames = {normalize(f) for f in exclude_filenames}
exclude_directories = {normalize(f) for f in exclude_directories}
path = str_to_fp(path).path
if normalize(path) in exclude_directories:
2016-05-10 08:00:29 +00:00
callback_exclusion(path, 'directory')
2016-03-07 20:25:50 +00:00
return
if normalize(os.path.split(path)[1]) in exclude_directories:
2016-05-10 08:00:29 +00:00
callback_exclusion(path, 'directory')
2016-03-07 20:25:50 +00:00
return
directory_queue = collections.deque()
directory_queue.append(path)
# This is a recursion-free workplace.
# Thank you for your cooperation.
while len(directory_queue) > 0:
location = directory_queue.popleft()
2016-05-10 08:00:29 +00:00
callback_verbose('listdir: %s' % location)
2016-03-07 20:25:50 +00:00
contents = os.listdir(location)
2016-05-10 08:00:29 +00:00
callback_verbose('received %d items' % len(contents))
2016-03-07 20:25:50 +00:00
directories = []
for base_name in contents:
absolute_name = os.path.join(location, base_name)
if os.path.isdir(absolute_name):
if normalize(absolute_name) in exclude_directories:
2016-05-10 08:00:29 +00:00
callback_exclusion(absolute_name, 'directory')
2016-03-07 20:25:50 +00:00
continue
if normalize(base_name) in exclude_directories:
2016-05-10 08:00:29 +00:00
callback_exclusion(absolute_name, 'directory')
2016-03-07 20:25:50 +00:00
continue
directories.append(absolute_name)
else:
if normalize(base_name) in exclude_filenames:
2016-05-10 08:00:29 +00:00
callback_exclusion(absolute_name, 'file')
2016-03-07 20:25:50 +00:00
continue
if normalize(absolute_name) in exclude_filenames:
2016-05-10 08:00:29 +00:00
callback_exclusion(absolute_filename, 'file')
2016-03-07 20:25:50 +00:00
continue
yield(str_to_fp(absolute_name))
2016-05-10 08:00:29 +00:00
# Extendleft causes them to get reversed, so flip it first.
2016-03-07 20:25:50 +00:00
directories.reverse()
directory_queue.extendleft(directories)