root / pithos / api / util.py @ 5635f9ef
History | View | Annotate | Download (19.1 kB)
1 |
# Copyright 2011 GRNET S.A. All rights reserved.
|
---|---|
2 |
#
|
3 |
# Redistribution and use in source and binary forms, with or
|
4 |
# without modification, are permitted provided that the following
|
5 |
# conditions are met:
|
6 |
#
|
7 |
# 1. Redistributions of source code must retain the above
|
8 |
# copyright notice, this list of conditions and the following
|
9 |
# disclaimer.
|
10 |
#
|
11 |
# 2. Redistributions in binary form must reproduce the above
|
12 |
# copyright notice, this list of conditions and the following
|
13 |
# disclaimer in the documentation and/or other materials
|
14 |
# provided with the distribution.
|
15 |
#
|
16 |
# THIS SOFTWARE IS PROVIDED BY GRNET S.A. ``AS IS'' AND ANY EXPRESS
|
17 |
# OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
|
18 |
# WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
19 |
# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL GRNET S.A OR
|
20 |
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
21 |
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
22 |
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
|
23 |
# USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
|
24 |
# AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
25 |
# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
|
26 |
# ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
27 |
# POSSIBILITY OF SUCH DAMAGE.
|
28 |
#
|
29 |
# The views and conclusions contained in the software and
|
30 |
# documentation are those of the authors and should not be
|
31 |
# interpreted as representing official policies, either expressed
|
32 |
# or implied, of GRNET S.A.
|
33 |
|
34 |
from functools import wraps |
35 |
from time import time |
36 |
from traceback import format_exc |
37 |
from wsgiref.handlers import format_date_time |
38 |
|
39 |
from django.conf import settings |
40 |
from django.http import HttpResponse |
41 |
from django.utils.http import http_date, parse_etags |
42 |
|
43 |
from pithos.api.compat import parse_http_date_safe |
44 |
from pithos.api.faults import (Fault, NotModified, BadRequest, ItemNotFound, LengthRequired, |
45 |
PreconditionFailed, ServiceUnavailable) |
46 |
from pithos.backends import backend |
47 |
|
48 |
import datetime |
49 |
import logging |
50 |
import re |
51 |
|
52 |
|
53 |
logger = logging.getLogger(__name__) |
54 |
|
55 |
|
56 |
def printable_meta_dict(d): |
57 |
"""Format a meta dictionary for printing out json/xml.
|
58 |
|
59 |
Convert all keys to lower case and replace dashes to underscores.
|
60 |
Change 'modified' key from backend to 'last_modified' and format date.
|
61 |
"""
|
62 |
if 'modified' in d: |
63 |
d['last_modified'] = datetime.datetime.fromtimestamp(int(d['modified'])).isoformat() |
64 |
del(d['modified']) |
65 |
return dict([(k.lower().replace('-', '_'), v) for k, v in d.iteritems()]) |
66 |
|
67 |
def format_meta_key(k): |
68 |
"""Convert underscores to dashes and capitalize intra-dash strings"""
|
69 |
return '-'.join([x.capitalize() for x in k.replace('_', '-').split('-')]) |
70 |
|
71 |
def get_meta_prefix(request, prefix): |
72 |
"""Get all prefix-* request headers in a dict. Reformat keys with format_meta_key()"""
|
73 |
prefix = 'HTTP_' + prefix.upper().replace('-', '_') |
74 |
return dict([(format_meta_key(k[5:]), v) for k, v in request.META.iteritems() if k.startswith(prefix)]) |
75 |
|
76 |
def get_account_meta(request): |
77 |
"""Get metadata from an account request"""
|
78 |
meta = get_meta_prefix(request, 'X-Account-Meta-')
|
79 |
return meta
|
80 |
|
81 |
def put_account_meta(response, meta): |
82 |
"""Put metadata in an account response"""
|
83 |
response['X-Account-Container-Count'] = meta['count'] |
84 |
response['X-Account-Bytes-Used'] = meta['bytes'] |
85 |
if 'modified' in meta: |
86 |
response['Last-Modified'] = http_date(int(meta['modified'])) |
87 |
for k in [x for x in meta.keys() if x.startswith('X-Account-Meta-')]: |
88 |
response[k.encode('utf-8')] = meta[k].encode('utf-8') |
89 |
|
90 |
def get_container_meta(request): |
91 |
"""Get metadata from a container request"""
|
92 |
meta = get_meta_prefix(request, 'X-Container-Meta-')
|
93 |
return meta
|
94 |
|
95 |
def put_container_meta(response, meta): |
96 |
"""Put metadata in a container response"""
|
97 |
response['X-Container-Object-Count'] = meta['count'] |
98 |
response['X-Container-Bytes-Used'] = meta['bytes'] |
99 |
if 'modified' in meta: |
100 |
response['Last-Modified'] = http_date(int(meta['modified'])) |
101 |
for k in [x for x in meta.keys() if x.startswith('X-Container-Meta-')]: |
102 |
response[k.encode('utf-8')] = meta[k].encode('utf-8') |
103 |
response['X-Container-Object-Meta'] = [x[14:] for x in meta['object_meta'] if x.startswith('X-Object-Meta-')] |
104 |
|
105 |
def get_object_meta(request): |
106 |
"""Get metadata from an object request"""
|
107 |
meta = get_meta_prefix(request, 'X-Object-Meta-')
|
108 |
if request.META.get('CONTENT_TYPE'): |
109 |
meta['Content-Type'] = request.META['CONTENT_TYPE'] |
110 |
if request.META.get('HTTP_CONTENT_ENCODING'): |
111 |
meta['Content-Encoding'] = request.META['HTTP_CONTENT_ENCODING'] |
112 |
if request.META.get('HTTP_CONTENT_DISPOSITION'): |
113 |
meta['Content-Disposition'] = request.META['HTTP_CONTENT_DISPOSITION'] |
114 |
if request.META.get('HTTP_X_OBJECT_MANIFEST'): |
115 |
meta['X-Object-Manifest'] = request.META['HTTP_X_OBJECT_MANIFEST'] |
116 |
return meta
|
117 |
|
118 |
def put_object_meta(response, meta): |
119 |
"""Put metadata in an object response"""
|
120 |
response['ETag'] = meta['hash'] |
121 |
response['Content-Length'] = meta['bytes'] |
122 |
response['Content-Type'] = meta.get('Content-Type', 'application/octet-stream') |
123 |
response['Last-Modified'] = http_date(int(meta['modified'])) |
124 |
for k in [x for x in meta.keys() if x.startswith('X-Object-Meta-')]: |
125 |
response[k.encode('utf-8')] = meta[k].encode('utf-8') |
126 |
for k in ('Content-Encoding', 'Content-Disposition', 'X-Object-Manifest'): |
127 |
if k in meta: |
128 |
response[k] = meta[k] |
129 |
|
130 |
def validate_modification_preconditions(request, meta): |
131 |
"""Check that the modified timestamp conforms with the preconditions set"""
|
132 |
if 'modified' not in meta: |
133 |
return # TODO: Always return? |
134 |
|
135 |
if_modified_since = request.META.get('HTTP_IF_MODIFIED_SINCE')
|
136 |
if if_modified_since is not None: |
137 |
if_modified_since = parse_http_date_safe(if_modified_since) |
138 |
if if_modified_since is not None and int(meta['modified']) <= if_modified_since: |
139 |
raise NotModified('Object has not been modified') |
140 |
|
141 |
if_unmodified_since = request.META.get('HTTP_IF_UNMODIFIED_SINCE')
|
142 |
if if_unmodified_since is not None: |
143 |
if_unmodified_since = parse_http_date_safe(if_unmodified_since) |
144 |
if if_unmodified_since is not None and int(meta['modified']) > if_unmodified_since: |
145 |
raise PreconditionFailed('Object has been modified') |
146 |
|
147 |
def validate_matching_preconditions(request, meta): |
148 |
"""Check that the ETag conforms with the preconditions set"""
|
149 |
if 'hash' not in meta: |
150 |
return # TODO: Always return? |
151 |
|
152 |
if_match = request.META.get('HTTP_IF_MATCH')
|
153 |
if if_match is not None and if_match != '*': |
154 |
if meta['hash'] not in [x.lower() for x in parse_etags(if_match)]: |
155 |
raise PreconditionFailed('Object Etag does not match') |
156 |
|
157 |
if_none_match = request.META.get('HTTP_IF_NONE_MATCH')
|
158 |
if if_none_match is not None: |
159 |
if if_none_match == '*' or meta['hash'] in [x.lower() for x in parse_etags(if_none_match)]: |
160 |
raise NotModified('Object Etag matches') |
161 |
|
162 |
def copy_or_move_object(request, src_path, dest_path, move=False): |
163 |
"""Copy or move an object"""
|
164 |
if type(src_path) == str: |
165 |
parts = src_path.split('/')
|
166 |
if len(parts) < 3 or parts[0] != '': |
167 |
raise BadRequest('Invalid X-Copy-From or X-Move-From header') |
168 |
src_container = parts[1]
|
169 |
src_name = '/'.join(parts[2:]) |
170 |
elif type(src_path) == tuple and len(src_path) == 2: |
171 |
src_container, src_name = src_path |
172 |
if type(dest_path) == str: |
173 |
parts = dest_path.split('/')
|
174 |
if len(parts) < 3 or parts[0] != '': |
175 |
raise BadRequest('Invalid Destination header') |
176 |
dest_container = parts[1]
|
177 |
dest_name = '/'.join(parts[2:]) |
178 |
elif type(dest_path) == tuple and len(dest_path) == 2: |
179 |
dest_container, dest_name = dest_path |
180 |
|
181 |
meta = get_object_meta(request) |
182 |
# Keep previous values of 'Content-Type' (if a new one is absent) and 'hash'.
|
183 |
try:
|
184 |
src_meta = backend.get_object_meta(request.user, src_container, src_name) |
185 |
except NameError: |
186 |
raise ItemNotFound('Container or object does not exist') |
187 |
if 'Content-Type' in meta and 'Content-Type' in src_meta: |
188 |
del(src_meta['Content-Type']) |
189 |
for k in ('Content-Type', 'hash'): |
190 |
if k in src_meta: |
191 |
meta[k] = src_meta[k] |
192 |
|
193 |
try:
|
194 |
if move:
|
195 |
backend.move_object(request.user, src_container, src_name, dest_container, dest_name, meta, replace_meta=True)
|
196 |
else:
|
197 |
backend.copy_object(request.user, src_container, src_name, dest_container, dest_name, meta, replace_meta=True)
|
198 |
except NameError: |
199 |
raise ItemNotFound('Container or object does not exist') |
200 |
|
201 |
def get_content_length(request): |
202 |
content_length = request.META.get('CONTENT_LENGTH')
|
203 |
if not content_length: |
204 |
raise LengthRequired('Missing Content-Length header') |
205 |
try:
|
206 |
content_length = int(content_length)
|
207 |
if content_length < 0: |
208 |
raise ValueError |
209 |
except ValueError: |
210 |
raise BadRequest('Invalid Content-Length header') |
211 |
return content_length
|
212 |
|
213 |
def get_range(request, size): |
214 |
"""Parse a Range header from the request
|
215 |
|
216 |
Either returns None, when the header is not existent or should be ignored,
|
217 |
or a list of (offset, length) tuples - should be further checked.
|
218 |
"""
|
219 |
ranges = request.META.get('HTTP_RANGE', '').replace(' ', '') |
220 |
if not ranges.startswith('bytes='): |
221 |
return None |
222 |
|
223 |
ret = [] |
224 |
for r in (x.strip() for x in ranges[6:].split(',')): |
225 |
p = re.compile('^(?P<offset>\d*)-(?P<upto>\d*)$')
|
226 |
m = p.match(r) |
227 |
if not m: |
228 |
return None |
229 |
offset = m.group('offset')
|
230 |
upto = m.group('upto')
|
231 |
if offset == '' and upto == '': |
232 |
return None |
233 |
|
234 |
if offset != '': |
235 |
offset = int(offset)
|
236 |
if upto != '': |
237 |
upto = int(upto)
|
238 |
if offset > upto:
|
239 |
return None |
240 |
ret.append((offset, upto - offset + 1))
|
241 |
else:
|
242 |
ret.append((offset, size - offset)) |
243 |
else:
|
244 |
length = int(upto)
|
245 |
ret.append((size - length, length)) |
246 |
|
247 |
return ret
|
248 |
|
249 |
def get_content_range(request): |
250 |
"""Parse a Content-Range header from the request
|
251 |
|
252 |
Either returns None, when the header is not existent or should be ignored,
|
253 |
or an (offset, length, total) tuple - check as length, total may be None.
|
254 |
Returns (None, None, None) if the provided range is '*/*'.
|
255 |
"""
|
256 |
|
257 |
ranges = request.META.get('HTTP_CONTENT_RANGE', '') |
258 |
if not ranges: |
259 |
return None |
260 |
|
261 |
p = re.compile('^bytes (?P<offset>\d+)-(?P<upto>\d*)/(?P<total>(\d+|\*))$')
|
262 |
m = p.match(ranges) |
263 |
if not m: |
264 |
if ranges == 'bytes */*': |
265 |
return (None, None, None) |
266 |
return None |
267 |
offset = int(m.group('offset')) |
268 |
upto = m.group('upto')
|
269 |
total = m.group('total')
|
270 |
if upto != '': |
271 |
upto = int(upto)
|
272 |
else:
|
273 |
upto = None
|
274 |
if total != '*': |
275 |
total = int(total)
|
276 |
else:
|
277 |
total = None
|
278 |
if (upto and offset > upto) or \ |
279 |
(total and offset >= total) or \ |
280 |
(total and upto and upto >= total): |
281 |
return None |
282 |
|
283 |
if not upto: |
284 |
length = None
|
285 |
else:
|
286 |
length = upto - offset + 1
|
287 |
return (offset, length, total)
|
288 |
|
289 |
def raw_input_socket(request): |
290 |
"""Return the socket for reading the rest of the request"""
|
291 |
server_software = request.META.get('SERVER_SOFTWARE')
|
292 |
if not server_software: |
293 |
if 'wsgi.input' in request.environ: |
294 |
return request.environ['wsgi.input'] |
295 |
raise ServiceUnavailable('Unknown server software') |
296 |
if server_software.startswith('WSGIServer'): |
297 |
return request.environ['wsgi.input'] |
298 |
elif server_software.startswith('mod_python'): |
299 |
return request._req
|
300 |
raise ServiceUnavailable('Unknown server software') |
301 |
|
302 |
MAX_UPLOAD_SIZE = 10 * (1024 * 1024) # 10MB |
303 |
|
304 |
def socket_read_iterator(sock, length=0, blocksize=4096): |
305 |
"""Return a maximum of blocksize data read from the socket in each iteration
|
306 |
|
307 |
Read up to 'length'. If 'length' is negative, will attempt a chunked read.
|
308 |
The maximum ammount of data read is controlled by MAX_UPLOAD_SIZE.
|
309 |
"""
|
310 |
if length < 0: # Chunked transfers |
311 |
data = ''
|
312 |
while length < MAX_UPLOAD_SIZE:
|
313 |
# Get chunk size.
|
314 |
if hasattr(sock, 'readline'): |
315 |
chunk_length = sock.readline() |
316 |
else:
|
317 |
chunk_length = ''
|
318 |
while chunk_length[-1:] != '\n': |
319 |
chunk_length += sock.read(1)
|
320 |
chunk_length.strip() |
321 |
pos = chunk_length.find(';')
|
322 |
if pos >= 0: |
323 |
chunk_length = chunk_length[:pos] |
324 |
try:
|
325 |
chunk_length = int(chunk_length, 16) |
326 |
except Exception, e: |
327 |
raise BadRequest('Bad chunk size') # TODO: Change to something more appropriate. |
328 |
# Check if done.
|
329 |
if chunk_length == 0: |
330 |
if len(data) > 0: |
331 |
yield data
|
332 |
return
|
333 |
# Get the actual data.
|
334 |
while chunk_length > 0: |
335 |
chunk = sock.read(min(chunk_length, blocksize))
|
336 |
chunk_length -= len(chunk)
|
337 |
length += len(chunk)
|
338 |
data += chunk |
339 |
if len(data) >= blocksize: |
340 |
ret = data[:blocksize] |
341 |
data = data[blocksize:] |
342 |
yield ret
|
343 |
sock.read(2) # CRLF |
344 |
# TODO: Raise something to note that maximum size is reached.
|
345 |
else:
|
346 |
if length > MAX_UPLOAD_SIZE:
|
347 |
# TODO: Raise something to note that maximum size is reached.
|
348 |
pass
|
349 |
while length > 0: |
350 |
data = sock.read(min(length, blocksize))
|
351 |
length -= len(data)
|
352 |
yield data
|
353 |
|
354 |
class ObjectWrapper(object): |
355 |
"""Return the object's data block-per-block in each iteration
|
356 |
|
357 |
Read from the object using the offset and length provided in each entry of the range list.
|
358 |
"""
|
359 |
|
360 |
def __init__(self, v_account, v_container, v_object, ranges, size, hashmap, boundary): |
361 |
self.v_account = v_account
|
362 |
self.v_container = v_container
|
363 |
self.v_object = v_object
|
364 |
self.ranges = ranges
|
365 |
self.size = size
|
366 |
self.hashmap = hashmap
|
367 |
self.boundary = boundary
|
368 |
|
369 |
self.block_index = -1 |
370 |
self.block = '' |
371 |
|
372 |
self.range_index = -1 |
373 |
self.offset, self.length = self.ranges[0] |
374 |
|
375 |
def __iter__(self): |
376 |
return self |
377 |
|
378 |
def part_iterator(self): |
379 |
if self.length > 0: |
380 |
# Get the block for the current offset.
|
381 |
bi = int(self.offset / backend.block_size) |
382 |
if self.block_index != bi: |
383 |
try:
|
384 |
self.block = backend.get_block(self.hashmap[bi]) |
385 |
except NameError: |
386 |
raise ItemNotFound('Block does not exist') |
387 |
self.block_index = bi
|
388 |
# Get the data from the block.
|
389 |
bo = self.offset % backend.block_size
|
390 |
bl = min(self.length, backend.block_size - bo) |
391 |
data = self.block[bo:bo + bl]
|
392 |
self.offset += bl
|
393 |
self.length -= bl
|
394 |
return data
|
395 |
else:
|
396 |
raise StopIteration |
397 |
|
398 |
def next(self): |
399 |
if len(self.ranges) == 1: |
400 |
return self.part_iterator() |
401 |
if self.range_index == len(self.ranges): |
402 |
raise StopIteration |
403 |
try:
|
404 |
if self.range_index == -1: |
405 |
raise StopIteration |
406 |
return self.part_iterator() |
407 |
except StopIteration: |
408 |
self.range_index += 1 |
409 |
out = [] |
410 |
if self.range_index < len(self.ranges): |
411 |
# Part header.
|
412 |
self.offset, self.length = self.ranges[self.range_index] |
413 |
if self.range_index > 0: |
414 |
out.append('')
|
415 |
out.append('--' + self.boundary) |
416 |
out.append('Content-Range: bytes %d-%d/%d' % (self.offset, self.offset + self.length - 1, self.size)) |
417 |
out.append('Content-Transfer-Encoding: binary')
|
418 |
out.append('')
|
419 |
out.append('')
|
420 |
return '\r\n'.join(out) |
421 |
else:
|
422 |
# Footer.
|
423 |
out.append('')
|
424 |
out.append('--' + self.boundary + '--') |
425 |
out.append('')
|
426 |
return '\r\n'.join(out) |
427 |
|
428 |
def update_response_headers(request, response): |
429 |
if request.serialization == 'xml': |
430 |
response['Content-Type'] = 'application/xml; charset=UTF-8' |
431 |
elif request.serialization == 'json': |
432 |
response['Content-Type'] = 'application/json; charset=UTF-8' |
433 |
elif not response['Content-Type']: |
434 |
response['Content-Type'] = 'text/plain; charset=UTF-8' |
435 |
|
436 |
if settings.TEST:
|
437 |
response['Date'] = format_date_time(time())
|
438 |
|
439 |
def render_fault(request, fault): |
440 |
if settings.DEBUG or settings.TEST: |
441 |
fault.details = format_exc(fault) |
442 |
|
443 |
request.serialization = 'text'
|
444 |
data = '\n'.join((fault.message, fault.details)) + '\n' |
445 |
response = HttpResponse(data, status=fault.code) |
446 |
update_response_headers(request, response) |
447 |
return response
|
448 |
|
449 |
def request_serialization(request, format_allowed=False): |
450 |
"""Return the serialization format requested
|
451 |
|
452 |
Valid formats are 'text' and 'json', 'xml' if 'format_allowed' is True.
|
453 |
"""
|
454 |
if not format_allowed: |
455 |
return 'text' |
456 |
|
457 |
format = request.GET.get('format')
|
458 |
if format == 'json': |
459 |
return 'json' |
460 |
elif format == 'xml': |
461 |
return 'xml' |
462 |
|
463 |
for item in request.META.get('HTTP_ACCEPT', '').split(','): |
464 |
accept, sep, rest = item.strip().partition(';')
|
465 |
if accept == 'application/json': |
466 |
return 'json' |
467 |
elif accept == 'application/xml' or accept == 'text/xml': |
468 |
return 'xml' |
469 |
|
470 |
return 'text' |
471 |
|
472 |
def api_method(http_method=None, format_allowed=False): |
473 |
"""Decorator function for views that implement an API method"""
|
474 |
def decorator(func): |
475 |
@wraps(func)
|
476 |
def wrapper(request, *args, **kwargs): |
477 |
try:
|
478 |
if http_method and request.method != http_method: |
479 |
raise BadRequest('Method not allowed.') |
480 |
|
481 |
# The args variable may contain up to (account, container, object).
|
482 |
if len(args) > 1 and len(args[1]) > 256: |
483 |
raise BadRequest('Container name too large.') |
484 |
if len(args) > 2 and len(args[2]) > 1024: |
485 |
raise BadRequest('Object name too large.') |
486 |
|
487 |
# Fill in custom request variables.
|
488 |
request.serialization = request_serialization(request, format_allowed) |
489 |
# TODO: Authenticate.
|
490 |
request.user = "test"
|
491 |
|
492 |
response = func(request, *args, **kwargs) |
493 |
update_response_headers(request, response) |
494 |
return response
|
495 |
except Fault, fault:
|
496 |
return render_fault(request, fault)
|
497 |
except BaseException, e: |
498 |
logger.exception('Unexpected error: %s' % e)
|
499 |
fault = ServiceUnavailable('Unexpected error')
|
500 |
return render_fault(request, fault)
|
501 |
return wrapper
|
502 |
return decorator
|