|
1
|
|
|
import os |
|
2
|
|
|
from multiprocessing import Manager |
|
3
|
|
|
from multiprocessing.pool import Pool |
|
4
|
|
|
|
|
5
|
|
|
|
|
6
|
|
View Code Duplication |
class Sprinter: |
|
|
|
|
|
|
7
|
|
|
|
|
8
|
|
|
def __init__(self, directory, filters, full_paths, pool_size, _printer): |
|
9
|
|
|
"""DirPaths sub class for directory parsing using parallel processing.""" |
|
10
|
|
|
self.directory = directory |
|
11
|
|
|
self.filters = filters |
|
12
|
|
|
self.pool_size = pool_size |
|
13
|
|
|
self._printer = _printer |
|
14
|
|
|
|
|
15
|
|
|
if self.filters: |
|
16
|
|
|
self._printer('Filtering enabled') |
|
17
|
|
|
self.explore_path = self.explore_path_filter |
|
18
|
|
|
else: |
|
19
|
|
|
self._printer('Filtering disabled') |
|
20
|
|
|
self.explore_path = self.explore_path_encompass |
|
21
|
|
|
|
|
22
|
|
|
self.filepaths = Manager().list() |
|
23
|
|
|
self.unsearched = Manager().Queue() |
|
24
|
|
|
|
|
25
|
|
|
if full_paths: |
|
26
|
|
|
self._printer('Absolute paths') |
|
27
|
|
|
self.add_path = self._add_filepath_absolute |
|
28
|
|
|
else: |
|
29
|
|
|
self._printer('Relative paths') |
|
30
|
|
|
self.add_path = self._add_filepath_relative |
|
31
|
|
|
|
|
32
|
|
|
def __iter__(self): |
|
33
|
|
|
return iter(self.filepaths) |
|
34
|
|
|
|
|
35
|
|
|
def __len__(self): |
|
36
|
|
|
return len(self.filepaths) |
|
37
|
|
|
|
|
38
|
|
|
def _add_filepath_relative(self, paths): |
|
39
|
|
|
for directory, fullname in paths: |
|
40
|
|
|
self.filepaths.append(fullname) |
|
41
|
|
|
|
|
42
|
|
|
def _add_filepath_absolute(self, paths): |
|
43
|
|
|
for directory, fullname in paths: |
|
44
|
|
|
self.filepaths.append(os.path.join(directory, fullname)) |
|
45
|
|
|
|
|
46
|
|
|
def _get_root_files(self, directory): |
|
47
|
|
|
"""Retrieve files within the root directory""" |
|
48
|
|
|
if len(self.filepaths) == 0: |
|
49
|
|
|
if self.filters: |
|
50
|
|
|
root_files = [(directory, f) for f in os.listdir(directory) |
|
51
|
|
|
if os.path.isfile(os.path.join(directory, f)) and self.filters.validate(f) |
|
52
|
|
|
and self.filters.get_level(f) == self.filters.max_level] |
|
53
|
|
|
else: |
|
54
|
|
|
root_files = [(directory, f) for f in os.listdir(directory) |
|
55
|
|
|
if os.path.isfile(os.path.join(directory, f))] |
|
56
|
|
|
self.add_path(root_files) |
|
57
|
|
|
|
|
58
|
|
|
def explore_path_filter(self, task_num, dirpath): |
|
59
|
|
|
""" |
|
60
|
|
|
Explore path to discover unsearched directories and save filepaths |
|
61
|
|
|
:param task_num: Processor ID |
|
62
|
|
|
:param dirpath: Tuple (base directory, path), path information pulled from unsearched Queue |
|
63
|
|
|
:return: Directories to add to unsearched Queue |
|
64
|
|
|
""" |
|
65
|
|
|
base, path = dirpath |
|
66
|
|
|
directories = [] |
|
67
|
|
|
nondirectories = [] |
|
68
|
|
|
if self.filters.validate(path): |
|
69
|
|
|
self._printer("Task: " + str(task_num) + " >>> Explored path: " + path, stream=True) |
|
70
|
|
|
# Loop through paths |
|
71
|
|
|
for filename in os.listdir(base + os.sep + path): |
|
72
|
|
|
fullname = os.path.join(path, filename) |
|
73
|
|
|
if self.filters.validate(fullname): |
|
74
|
|
|
# Check that non-empty folders flag is on and we're at the max directory level |
|
75
|
|
|
if self.filters.non_empty_folders and self.filters.get_level(fullname) == self.filters.max_level: |
|
76
|
|
|
# Check that the path is not an empty folder |
|
77
|
|
|
if os.path.isdir(base + os.sep + fullname): |
|
78
|
|
|
# Get paths in folder without walking directory |
|
79
|
|
|
paths = os.listdir(base + os.sep + fullname) |
|
80
|
|
|
|
|
81
|
|
|
# Check that any of the paths are files and not just directories |
|
82
|
|
|
if paths and any(os.path.isfile(os.path.join(base, fullname, p)) for p in paths): |
|
83
|
|
|
nondirectories.append((base, fullname)) |
|
84
|
|
|
else: |
|
85
|
|
|
# Append to directories if dir |
|
86
|
|
|
if os.path.isdir(base + os.sep + fullname): |
|
87
|
|
|
directories.append((base, fullname)) |
|
88
|
|
|
# Pass filters and append to nondirectories if file |
|
89
|
|
|
elif self.filters.validate(fullname): |
|
90
|
|
|
nondirectories.append((base, fullname)) |
|
91
|
|
|
self.add_path(nondirectories) |
|
92
|
|
|
return directories |
|
93
|
|
|
|
|
94
|
|
|
def explore_path_encompass(self, task_num, dirpath): |
|
95
|
|
|
""" |
|
96
|
|
|
Explore path to discover unsearched directories and save filepaths |
|
97
|
|
|
:param task_num: Processor ID |
|
98
|
|
|
:param dirpath: Tuple (base directory, path), path information pulled from unsearched Queue |
|
99
|
|
|
:return: Directories to add to unsearched Queue |
|
100
|
|
|
""" |
|
101
|
|
|
base, path = dirpath |
|
102
|
|
|
directories = [] |
|
103
|
|
|
nondirectories = [] |
|
104
|
|
|
self._printer("Task: " + str(task_num) + " >>> Explored path: " + path, stream=True) |
|
105
|
|
|
# Loop through paths |
|
106
|
|
|
for filename in os.listdir(base + os.sep + path): |
|
107
|
|
|
fullname = os.path.join(path, filename) |
|
108
|
|
|
# Append to directories if dir |
|
109
|
|
|
if os.path.isdir(base + os.sep + fullname): |
|
110
|
|
|
directories.append((base, fullname)) |
|
111
|
|
|
# Pass filters and append to nondirectories if file |
|
112
|
|
|
else: |
|
113
|
|
|
nondirectories.append((base, fullname)) |
|
114
|
|
|
self.add_path(nondirectories) |
|
115
|
|
|
return directories |
|
116
|
|
|
|
|
117
|
|
|
def parallel_worker(self, task_num): |
|
118
|
|
|
""" |
|
119
|
|
|
Continuously pulls directories from the Queue until it is empty. |
|
120
|
|
|
Gets child directories from parent and adds them to Queue. |
|
121
|
|
|
Executes task_done() to remove directory from unsearched Queue |
|
122
|
|
|
:param task_num: Processor ID |
|
123
|
|
|
""" |
|
124
|
|
|
while True: |
|
125
|
|
|
base, path = self.unsearched.get() |
|
126
|
|
|
dirs = self.explore_path(task_num, (base, path)) |
|
127
|
|
|
for newdir in dirs: |
|
128
|
|
|
self.unsearched.put(newdir) |
|
129
|
|
|
self.unsearched.task_done() |
|
130
|
|
|
|
|
131
|
|
|
def sprinter(self): |
|
132
|
|
|
""" |
|
133
|
|
|
Called when parallelize is True. |
|
134
|
|
|
This function will generate the file names in a directory tree by adding directories to a Queue and |
|
135
|
|
|
continuously exploring directories in the Queue until Queue is emptied. |
|
136
|
|
|
Significantly faster than crawler method for larger directory trees. |
|
137
|
|
|
""" |
|
138
|
|
|
self._printer('Multiprocess Walk') |
|
139
|
|
|
# Loop through directories in case there is more than one (1) |
|
140
|
|
|
for directory in self.directory: |
|
141
|
|
|
self._get_root_files(directory) # Add file within root directory if filepaths is empty |
|
142
|
|
|
# acquire the list of paths |
|
143
|
|
|
first_level_dirs = next(os.walk(directory))[1] |
|
144
|
|
|
for path in first_level_dirs: |
|
145
|
|
|
self.unsearched.put((directory, path)) |
|
146
|
|
|
self._printer('Pool Processing STARTED') |
|
147
|
|
|
pool = Pool(self.pool_size) |
|
148
|
|
|
pool.map_async(self.parallel_worker, range(self.pool_size)) |
|
149
|
|
|
pool.close() |
|
150
|
|
|
self.unsearched.join() |
|
151
|
|
|
self._printer('Pool Processing ENDED') |
|
152
|
|
|
return self.filepaths |
|
153
|
|
|
|
sfneal /
dirutility
Loading history...