This repository has been archived by the owner on Dec 16, 2022. It is now read-only.
/
file_utils.py
460 lines (374 loc) · 15.7 KB
/
file_utils.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
"""
Utilities for working with the local dataset cache.
"""
import glob
import os
import logging
import tempfile
import json
from urllib.parse import urlparse
from pathlib import Path
from typing import Optional, Tuple, Union, IO, Callable, Set, List
from hashlib import sha256
from functools import wraps
from zipfile import ZipFile, is_zipfile
import tarfile
import shutil
import boto3
import botocore
from botocore.exceptions import ClientError, EndpointConnectionError
from filelock import FileLock
import requests
from requests.adapters import HTTPAdapter
from requests.exceptions import ConnectionError
from requests.packages.urllib3.util.retry import Retry
from allennlp.common.tqdm import Tqdm
logger = logging.getLogger(__name__)
CACHE_ROOT = Path(os.getenv("ALLENNLP_CACHE_ROOT", Path.home() / ".allennlp"))
CACHE_DIRECTORY = str(CACHE_ROOT / "cache")
DEPRECATED_CACHE_DIRECTORY = str(CACHE_ROOT / "datasets")
# This variable was deprecated in 0.7.2 since we use a single folder for caching
# all types of files (datasets, models, etc.)
DATASET_CACHE = CACHE_DIRECTORY
# Warn if the user is still using the deprecated cache directory.
if os.path.exists(DEPRECATED_CACHE_DIRECTORY):
logger = logging.getLogger(__name__)
logger.warning(
f"Deprecated cache directory found ({DEPRECATED_CACHE_DIRECTORY}). "
f"Please remove this directory from your system to free up space."
)
def url_to_filename(url: str, etag: str = None) -> str:
"""
Convert `url` into a hashed filename in a repeatable way.
If `etag` is specified, append its hash to the url's, delimited
by a period.
"""
url_bytes = url.encode("utf-8")
url_hash = sha256(url_bytes)
filename = url_hash.hexdigest()
if etag:
etag_bytes = etag.encode("utf-8")
etag_hash = sha256(etag_bytes)
filename += "." + etag_hash.hexdigest()
return filename
def filename_to_url(filename: str, cache_dir: Union[str, Path] = None) -> Tuple[str, str]:
"""
Return the url and etag (which may be `None`) stored for `filename`.
Raise `FileNotFoundError` if `filename` or its stored metadata do not exist.
"""
if cache_dir is None:
cache_dir = CACHE_DIRECTORY
cache_path = os.path.join(cache_dir, filename)
if not os.path.exists(cache_path):
raise FileNotFoundError("file {} not found".format(cache_path))
meta_path = cache_path + ".json"
if not os.path.exists(meta_path):
raise FileNotFoundError("file {} not found".format(meta_path))
with open(meta_path) as meta_file:
metadata = json.load(meta_file)
url = metadata["url"]
etag = metadata["etag"]
return url, etag
def cached_path(
url_or_filename: Union[str, Path],
cache_dir: Union[str, Path] = None,
extract_archive: bool = False,
force_extract: bool = False,
) -> str:
"""
Given something that might be a URL (or might be a local path),
determine which. If it's a URL, download the file and cache it, and
return the path to the cached file. If it's already a local path,
make sure the file exists and then return the path.
# Parameters
url_or_filename : `Union[str, Path]`
A URL or local file to parse and possibly download.
cache_dir : `Union[str, Path]`, optional (default = `None`)
The directory to cache downloads.
extract_archive : `bool`, optional (default = `False`)
If `True`, then zip or tar.gz archives will be automatically extracted.
In which case the directory is returned.
force_extract : `bool`, optional (default = `False`)
If `True` and the file is an archive file, it will be extracted regardless
of whether or not the extracted directory already exists.
"""
if cache_dir is None:
cache_dir = CACHE_DIRECTORY
if isinstance(url_or_filename, Path):
url_or_filename = str(url_or_filename)
url_or_filename = os.path.expanduser(url_or_filename)
parsed = urlparse(url_or_filename)
file_path: str
extraction_path: Optional[str] = None
if parsed.scheme in ("http", "https", "s3"):
# URL, so get it from the cache (downloading if necessary)
file_path = get_from_cache(url_or_filename, cache_dir)
if extract_archive and (is_zipfile(file_path) or tarfile.is_tarfile(file_path)):
# This is the path the file should be extracted to.
# For example ~/.allennlp/cache/234234.21341 -> ~/.allennlp/cache/234234.21341-extracted
extraction_path = file_path + "-extracted"
elif os.path.exists(url_or_filename):
# File, and it exists.
file_path = url_or_filename
if extract_archive and (is_zipfile(file_path) or tarfile.is_tarfile(file_path)):
# This is the path the file should be extracted to.
# For example model.tar.gz -> model-tar-gz-extracted
extraction_dir, extraction_name = os.path.split(file_path)
extraction_name = extraction_name.replace(".", "-") + "-extracted"
extraction_path = os.path.join(extraction_dir, extraction_name)
elif parsed.scheme == "":
# File, but it doesn't exist.
raise FileNotFoundError("file {} not found".format(url_or_filename))
else:
# Something unknown
raise ValueError("unable to parse {} as a URL or as a local path".format(url_or_filename))
if extraction_path is not None:
# No need to extract again.
if os.path.isdir(extraction_path) and os.listdir(extraction_path) and not force_extract:
return extraction_path
# Extract it.
with FileLock(file_path + ".lock"):
shutil.rmtree(extraction_path, ignore_errors=True)
os.makedirs(extraction_path)
if is_zipfile(file_path):
with ZipFile(file_path, "r") as zip_file:
zip_file.extractall(extraction_path)
zip_file.close()
else:
tar_file = tarfile.open(file_path)
tar_file.extractall(extraction_path)
tar_file.close()
return extraction_path
return file_path
def is_url_or_existing_file(url_or_filename: Union[str, Path, None]) -> bool:
"""
Given something that might be a URL (or might be a local path),
determine check if it's url or an existing file path.
"""
if url_or_filename is None:
return False
url_or_filename = os.path.expanduser(str(url_or_filename))
parsed = urlparse(url_or_filename)
return parsed.scheme in ("http", "https", "s3") or os.path.exists(url_or_filename)
def _split_s3_path(url: str) -> Tuple[str, str]:
"""Split a full s3 path into the bucket name and path."""
parsed = urlparse(url)
if not parsed.netloc or not parsed.path:
raise ValueError("bad s3 path {}".format(url))
bucket_name = parsed.netloc
s3_path = parsed.path
# Remove '/' at beginning of path.
if s3_path.startswith("/"):
s3_path = s3_path[1:]
return bucket_name, s3_path
def _s3_request(func: Callable):
"""
Wrapper function for s3 requests in order to create more helpful error
messages.
"""
@wraps(func)
def wrapper(url: str, *args, **kwargs):
try:
return func(url, *args, **kwargs)
except ClientError as exc:
if int(exc.response["Error"]["Code"]) == 404:
raise FileNotFoundError("file {} not found".format(url))
else:
raise
return wrapper
def _get_s3_resource():
session = boto3.session.Session()
if session.get_credentials() is None:
# Use unsigned requests.
s3_resource = session.resource(
"s3", config=botocore.client.Config(signature_version=botocore.UNSIGNED)
)
else:
s3_resource = session.resource("s3")
return s3_resource
@_s3_request
def _s3_etag(url: str) -> Optional[str]:
"""Check ETag on S3 object."""
s3_resource = _get_s3_resource()
bucket_name, s3_path = _split_s3_path(url)
s3_object = s3_resource.Object(bucket_name, s3_path)
return s3_object.e_tag
@_s3_request
def _s3_get(url: str, temp_file: IO) -> None:
"""Pull a file directly from S3."""
s3_resource = _get_s3_resource()
bucket_name, s3_path = _split_s3_path(url)
s3_resource.Bucket(bucket_name).download_fileobj(s3_path, temp_file)
def _session_with_backoff() -> requests.Session:
"""
We ran into an issue where http requests to s3 were timing out,
possibly because we were making too many requests too quickly.
This helper function returns a requests session that has retry-with-backoff
built in. See
<https://stackoverflow.com/questions/23267409/how-to-implement-retry-mechanism-into-python-requests-library>.
"""
session = requests.Session()
retries = Retry(total=5, backoff_factor=1, status_forcelist=[502, 503, 504])
session.mount("http://", HTTPAdapter(max_retries=retries))
session.mount("https://", HTTPAdapter(max_retries=retries))
return session
def _http_etag(url: str) -> Optional[str]:
with _session_with_backoff() as session:
response = session.head(url, allow_redirects=True)
if response.status_code != 200:
raise IOError(
"HEAD request failed for url {} with status code {}".format(url, response.status_code)
)
return response.headers.get("ETag")
def _http_get(url: str, temp_file: IO) -> None:
with _session_with_backoff() as session:
req = session.get(url, stream=True)
content_length = req.headers.get("Content-Length")
total = int(content_length) if content_length is not None else None
progress = Tqdm.tqdm(unit="B", total=total)
for chunk in req.iter_content(chunk_size=1024):
if chunk: # filter out keep-alive new chunks
progress.update(len(chunk))
temp_file.write(chunk)
progress.close()
def _find_latest_cached(url: str, cache_dir: Union[str, Path]) -> Optional[str]:
filename = url_to_filename(url)
cache_path = os.path.join(cache_dir, filename)
candidates: List[Tuple[str, float]] = []
for path in glob.glob(cache_path + "*"):
if path.endswith(".json"):
continue
mtime = os.path.getmtime(path)
candidates.append((path, mtime))
# Sort candidates by modification time, neweste first.
candidates.sort(key=lambda x: x[1], reverse=True)
if candidates:
return candidates[0][0]
return None
class CacheFile:
"""
This is a context manager that makes robust caching easier.
On `__enter__`, an IO handle to a temporarily file is returned, which can
be treated as if it's the actual cache file.
On `__exit__`, the temporarily file is renamed to the cache file. If anything
goes wrong while writing to the temporary file, it will be removed.
"""
def __init__(self, cache_filename: Union[Path, str], mode="w+b") -> None:
self.cache_filename = (
cache_filename if isinstance(cache_filename, Path) else Path(cache_filename)
)
self.cache_directory = os.path.dirname(self.cache_filename)
self.mode = mode
self.temp_file = tempfile.NamedTemporaryFile(
self.mode, dir=self.cache_directory, delete=False, suffix=".tmp"
)
def __enter__(self):
return self.temp_file
def __exit__(self, exc_type, exc_value, traceback):
self.temp_file.close()
if exc_value is None:
# Success.
logger.debug(
"Renaming temp file %s to cache at %s", self.temp_file.name, self.cache_filename
)
# Rename the temp file to the actual cache filename.
os.replace(self.temp_file.name, self.cache_filename)
return True
# Something went wrong, remove the temp file.
logger.debug("removing temp file %s", self.temp_file.name)
os.remove(self.temp_file.name)
return False
# TODO(joelgrus): do we want to do checksums or anything like that?
def get_from_cache(url: str, cache_dir: Union[str, Path] = None) -> str:
"""
Given a URL, look for the corresponding dataset in the local cache.
If it's not there, download it. Then return the path to the cached file.
"""
if cache_dir is None:
cache_dir = CACHE_DIRECTORY
os.makedirs(cache_dir, exist_ok=True)
# Get eTag to add to filename, if it exists.
try:
if url.startswith("s3://"):
etag = _s3_etag(url)
else:
etag = _http_etag(url)
except (ConnectionError, EndpointConnectionError):
# We might be offline, in which case we don't want to throw an error
# just yet. Instead, we'll try to use the latest cached version of the
# target resource, if it exists. We'll only throw an exception if we
# haven't cached the resource at all yet.
logger.warning(
"Connection error occured while trying to fetch ETag for %s. "
"Will attempt to use latest cached version of resource",
url,
)
latest_cached = _find_latest_cached(url, cache_dir)
if latest_cached:
logger.info(
"ETag request failed with connection error, using latest cached "
"version of %s: %s",
url,
latest_cached,
)
return latest_cached
else:
logger.error(
"Connection failed while trying to fetch ETag, "
"and no cached version of %s could be found",
url,
)
raise
filename = url_to_filename(url, etag)
# Get cache path to put the file.
cache_path = os.path.join(cache_dir, filename)
# Multiple processes may be trying to cache the same file at once, so we need
# to be a little careful to avoid race conditions. We do this using a lock file.
# Only one process can own this lock file at a time, and a process will block
# on the call to `lock.acquire()` until the process currently holding the lock
# releases it.
logger.debug("waiting to acquire lock on %s", cache_path)
with FileLock(cache_path + ".lock"):
if os.path.exists(cache_path):
logger.info("cache of %s is up-to-date", url)
else:
with CacheFile(cache_path) as cache_file:
logger.info("%s not found in cache, downloading to %s", url, cache_path)
# GET file object
if url.startswith("s3://"):
_s3_get(url, cache_file)
else:
_http_get(url, cache_file)
logger.debug("creating metadata file for %s", cache_path)
meta = {"url": url, "etag": etag}
meta_path = cache_path + ".json"
with open(meta_path, "w") as meta_file:
json.dump(meta, meta_file)
return cache_path
def read_set_from_file(filename: str) -> Set[str]:
"""
Extract a de-duped collection (set) of text from a file.
Expected file format is one item per line.
"""
collection = set()
with open(filename, "r") as file_:
for line in file_:
collection.add(line.rstrip())
return collection
def get_file_extension(path: str, dot=True, lower: bool = True):
ext = os.path.splitext(path)[1]
ext = ext if dot else ext[1:]
return ext.lower() if lower else ext
def open_compressed(
filename: Union[str, Path], mode: str = "rt", encoding: Optional[str] = "UTF-8", **kwargs
):
if isinstance(filename, Path):
filename = str(filename)
open_fn: Callable = open
if filename.endswith(".gz"):
import gzip
open_fn = gzip.open
elif filename.endswith(".bz2"):
import bz2
open_fn = bz2.open
return open_fn(filename, mode=mode, encoding=encoding, **kwargs)