from time import time
from traceback import format_exc
from wsgiref.handlers import format_date_time
-from binascii import hexlify
+from binascii import hexlify, unhexlify
from django.conf import settings
from django.http import HttpResponse
+from django.utils import simplejson as json
from django.utils.http import http_date, parse_etags
+from django.utils.encoding import smart_str
-from pithos.api.compat import parse_http_date_safe
-from pithos.api.faults import (Fault, NotModified, BadRequest, ItemNotFound, LengthRequired,
- PreconditionFailed, RangeNotSatisfiable, ServiceUnavailable)
+from pithos.api.compat import parse_http_date_safe, parse_http_date
+from pithos.api.faults import (Fault, NotModified, BadRequest, Unauthorized, ItemNotFound,
+ Conflict, LengthRequired, PreconditionFailed, RangeNotSatisfiable,
+ ServiceUnavailable)
from pithos.backends import backend
from pithos.backends.base import NotAllowedError
logger = logging.getLogger(__name__)
-def printable_meta_dict(d):
+def rename_meta_key(d, old, new):
+ if old not in d:
+ return
+ d[new] = d[old]
+ del(d[old])
+
+def printable_header_dict(d):
"""Format a meta dictionary for printing out json/xml.
- Convert all keys to lower case and replace dashes to underscores.
- Change 'modified' key from backend to 'last_modified' and format date.
+ Convert all keys to lower case and replace dashes with underscores.
+ Format 'last_modified' timestamp.
"""
- if 'modified' in d:
- d['last_modified'] = datetime.datetime.fromtimestamp(int(d['modified'])).isoformat()
- del(d['modified'])
+ d['last_modified'] = datetime.datetime.fromtimestamp(int(d['last_modified'])).isoformat()
return dict([(k.lower().replace('-', '_'), v) for k, v in d.iteritems()])
-def format_meta_key(k):
+def format_header_key(k):
"""Convert underscores to dashes and capitalize intra-dash strings."""
-
return '-'.join([x.capitalize() for x in k.replace('_', '-').split('-')])
-def get_meta_prefix(request, prefix):
- """Get all prefix-* request headers in a dict. Reformat keys with format_meta_key()."""
+def get_header_prefix(request, prefix):
+ """Get all prefix-* request headers in a dict. Reformat keys with format_header_key()."""
prefix = 'HTTP_' + prefix.upper().replace('-', '_')
- return dict([(format_meta_key(k[5:]), v) for k, v in request.META.iteritems() if k.startswith(prefix) and len(k) > len(prefix)])
-
-def get_account_meta(request):
- """Get metadata from an account request."""
-
- meta = get_meta_prefix(request, 'X-Account-Meta-')
- return meta
-
-def put_account_meta(response, meta):
- """Put metadata in an account response."""
-
- response['X-Account-Container-Count'] = meta['count']
- response['X-Account-Bytes-Used'] = meta['bytes']
- if 'modified' in meta:
- response['Last-Modified'] = http_date(int(meta['modified']))
+ # TODO: Document or remove '~' replacing.
+ return dict([(format_header_key(k[5:]), v.replace('~', '')) for k, v in request.META.iteritems() if k.startswith(prefix) and len(k) > len(prefix)])
+
+def get_account_headers(request):
+ meta = get_header_prefix(request, 'X-Account-Meta-')
+ groups = {}
+ for k, v in get_header_prefix(request, 'X-Account-Group-').iteritems():
+ n = k[16:].lower()
+ if '-' in n or '_' in n:
+ raise BadRequest('Bad characters in group name')
+ groups[n] = v.replace(' ', '').split(',')
+ if '' in groups[n]:
+ groups[n].remove('')
+ return meta, groups
+
+def put_account_headers(response, meta, groups):
+ if 'count' in meta:
+ response['X-Account-Container-Count'] = meta['count']
+ if 'bytes' in meta:
+ response['X-Account-Bytes-Used'] = meta['bytes']
+ response['Last-Modified'] = http_date(int(meta['modified']))
for k in [x for x in meta.keys() if x.startswith('X-Account-Meta-')]:
- response[k.encode('utf-8')] = meta[k].encode('utf-8')
+ response[smart_str(k, strings_only=True)] = smart_str(meta[k], strings_only=True)
if 'until_timestamp' in meta:
response['X-Account-Until-Timestamp'] = http_date(int(meta['until_timestamp']))
-
-def get_container_meta(request):
- """Get metadata from a container request."""
-
- meta = get_meta_prefix(request, 'X-Container-Meta-')
- return meta
-
-def put_container_meta(response, meta):
- """Put metadata in a container response."""
-
- response['X-Container-Object-Count'] = meta['count']
- response['X-Container-Bytes-Used'] = meta['bytes']
+ for k, v in groups.iteritems():
+ k = smart_str(k, strings_only=True)
+ k = format_header_key('X-Account-Group-' + k)
+ v = smart_str(','.join(v), strings_only=True)
+ response[k] = v
+
+def get_container_headers(request):
+ meta = get_header_prefix(request, 'X-Container-Meta-')
+ policy = dict([(k[19:].lower(), v.replace(' ', '')) for k, v in get_header_prefix(request, 'X-Container-Policy-').iteritems()])
+ return meta, policy
+
+def put_container_headers(response, meta, policy):
+ if 'count' in meta:
+ response['X-Container-Object-Count'] = meta['count']
+ if 'bytes' in meta:
+ response['X-Container-Bytes-Used'] = meta['bytes']
response['Last-Modified'] = http_date(int(meta['modified']))
for k in [x for x in meta.keys() if x.startswith('X-Container-Meta-')]:
- response[k.encode('utf-8')] = meta[k].encode('utf-8')
- response['X-Container-Object-Meta'] = [x[14:] for x in meta['object_meta'] if x.startswith('X-Object-Meta-')]
+ response[smart_str(k, strings_only=True)] = smart_str(meta[k], strings_only=True)
+ l = [smart_str(x, strings_only=True) for x in meta['object_meta'] if x.startswith('X-Object-Meta-')]
+ response['X-Container-Object-Meta'] = ','.join([x[14:] for x in l])
response['X-Container-Block-Size'] = backend.block_size
response['X-Container-Block-Hash'] = backend.hash_algorithm
if 'until_timestamp' in meta:
response['X-Container-Until-Timestamp'] = http_date(int(meta['until_timestamp']))
+ for k, v in policy.iteritems():
+ response[smart_str(format_header_key('X-Container-Policy-' + k), strings_only=True)] = smart_str(v, strings_only=True)
-def get_object_meta(request):
- """Get metadata from an object request."""
-
- meta = get_meta_prefix(request, 'X-Object-Meta-')
+def get_object_headers(request):
+ meta = get_header_prefix(request, 'X-Object-Meta-')
if request.META.get('CONTENT_TYPE'):
meta['Content-Type'] = request.META['CONTENT_TYPE']
if request.META.get('HTTP_CONTENT_ENCODING'):
meta['Content-Disposition'] = request.META['HTTP_CONTENT_DISPOSITION']
if request.META.get('HTTP_X_OBJECT_MANIFEST'):
meta['X-Object-Manifest'] = request.META['HTTP_X_OBJECT_MANIFEST']
- return meta
+ return meta, get_sharing(request), get_public(request)
-def put_object_meta(response, meta, public=False):
- """Put metadata in an object response."""
-
+def put_object_headers(response, meta, restricted=False):
response['ETag'] = meta['hash']
response['Content-Length'] = meta['bytes']
response['Content-Type'] = meta.get('Content-Type', 'application/octet-stream')
response['Last-Modified'] = http_date(int(meta['modified']))
- if not public:
- response['X-Object-Modified-By'] = meta['modified_by']
+ if not restricted:
+ response['X-Object-Modified-By'] = smart_str(meta['modified_by'], strings_only=True)
response['X-Object-Version'] = meta['version']
response['X-Object-Version-Timestamp'] = http_date(int(meta['version_timestamp']))
for k in [x for x in meta.keys() if x.startswith('X-Object-Meta-')]:
- response[k.encode('utf-8')] = meta[k].encode('utf-8')
- for k in ('Content-Encoding', 'Content-Disposition', 'X-Object-Manifest', 'X-Object-Sharing', 'X-Object-Shared-By'):
+ response[smart_str(k, strings_only=True)] = smart_str(meta[k], strings_only=True)
+ for k in ('Content-Encoding', 'Content-Disposition', 'X-Object-Manifest',
+ 'X-Object-Sharing', 'X-Object-Shared-By', 'X-Object-Allowed-To',
+ 'X-Object-Public'):
if k in meta:
- response[k] = meta[k]
+ response[k] = smart_str(meta[k], strings_only=True)
else:
for k in ('Content-Encoding', 'Content-Disposition'):
if k in meta:
md5.update(hash)
meta['hash'] = md5.hexdigest().lower()
-def update_sharing_meta(permissions, v_account, v_container, v_object, meta):
+def update_sharing_meta(request, permissions, v_account, v_container, v_object, meta):
if permissions is None:
return
- perm_path, perms = permissions
+ allowed, perm_path, perms = permissions
if len(perms) == 0:
return
ret = []
meta['X-Object-Sharing'] = '; '.join(ret)
if '/'.join((v_account, v_container, v_object)) != perm_path:
meta['X-Object-Shared-By'] = perm_path
+ if request.user != v_account:
+ meta['X-Object-Allowed-To'] = allowed
+
+def update_public_meta(public, meta):
+ if not public:
+ return
+ meta['X-Object-Public'] = public
def validate_modification_preconditions(request, meta):
"""Check that the modified timestamp conforms with the preconditions set."""
def validate_matching_preconditions(request, meta):
"""Check that the ETag conforms with the preconditions set."""
- if 'hash' not in meta:
- return # TODO: Always return?
+ hash = meta.get('hash', None)
if_match = request.META.get('HTTP_IF_MATCH')
- if if_match is not None and if_match != '*':
- if meta['hash'] not in [x.lower() for x in parse_etags(if_match)]:
- raise PreconditionFailed('Resource Etag does not match')
+ if if_match is not None:
+ if hash is None:
+ raise PreconditionFailed('Resource does not exist')
+ if if_match != '*' and hash not in [x.lower() for x in parse_etags(if_match)]:
+ raise PreconditionFailed('Resource ETag does not match')
if_none_match = request.META.get('HTTP_IF_NONE_MATCH')
if if_none_match is not None:
- if if_none_match == '*' or meta['hash'] in [x.lower() for x in parse_etags(if_none_match)]:
- raise NotModified('Resource Etag matches')
+ # TODO: If this passes, must ignore If-Modified-Since header.
+ if hash is not None:
+ if if_none_match == '*' or hash in [x.lower() for x in parse_etags(if_none_match)]:
+ # TODO: Continue if an If-Modified-Since header is present.
+ if request.method in ('HEAD', 'GET'):
+ raise NotModified('Resource ETag matches')
+ raise PreconditionFailed('Resource exists or ETag matches')
def split_container_object_string(s):
if not len(s) > 0 or s[0] != '/':
def copy_or_move_object(request, v_account, src_container, src_name, dest_container, dest_name, move=False):
"""Copy or move an object."""
- meta = get_object_meta(request)
- permissions = get_sharing(request)
+ meta, permissions, public = get_object_headers(request)
src_version = request.META.get('HTTP_X_SOURCE_VERSION')
try:
if move:
- backend.move_object(request.user, v_account, src_container, src_name, dest_container, dest_name, meta, False, permissions)
+ version_id = backend.move_object(request.user, v_account, src_container, src_name, dest_container, dest_name, meta, False, permissions)
else:
- backend.copy_object(request.user, v_account, src_container, src_name, dest_container, dest_name, meta, False, permissions, src_version)
+ version_id = backend.copy_object(request.user, v_account, src_container, src_name, dest_container, dest_name, meta, False, permissions, src_version)
except NotAllowedError:
raise Unauthorized('Access denied')
- except NameError, IndexError:
+ except (NameError, IndexError):
raise ItemNotFound('Container or object does not exist')
except ValueError:
raise BadRequest('Invalid sharing header')
- except AttributeError:
- raise Conflict('Sharing already set above or below this path in the hierarchy')
-
-def get_int_parameter(request, name):
- p = request.GET.get(name)
+ except AttributeError, e:
+ raise Conflict(json.dumps(e.data))
+ if public is not None:
+ try:
+ backend.update_object_public(request.user, v_account, dest_container, dest_name, public)
+ except NotAllowedError:
+ raise Unauthorized('Access denied')
+ except NameError:
+ raise ItemNotFound('Object does not exist')
+ return version_id
+
+def get_int_parameter(p):
if p is not None:
try:
p = int(p)
return p
def get_content_length(request):
- content_length = request.META.get('CONTENT_LENGTH')
- if not content_length:
- raise LengthRequired('Missing Content-Length header')
- try:
- content_length = int(content_length)
- if content_length < 0:
- raise ValueError
- except ValueError:
- raise BadRequest('Invalid Content-Length header')
+ content_length = get_int_parameter(request.META.get('CONTENT_LENGTH'))
+ if content_length is None:
+ raise LengthRequired('Missing or invalid Content-Length header')
return content_length
def get_range(request, size):
if permissions is None:
return None
+ # TODO: Document or remove '~' replacing.
+ permissions = permissions.replace('~', '')
+
ret = {}
permissions = permissions.replace(' ', '')
if permissions == '':
return ret
for perm in (x for x in permissions.split(';')):
if perm.startswith('read='):
- ret['read'] = [v.replace(' ','') for v in perm[5:].split(',')]
- ret['read'].remove('')
+ ret['read'] = list(set([v.replace(' ','').lower() for v in perm[5:].split(',')]))
+ if '' in ret['read']:
+ ret['read'].remove('')
if '*' in ret['read']:
ret['read'] = ['*']
if len(ret['read']) == 0:
raise BadRequest('Bad X-Object-Sharing header value')
elif perm.startswith('write='):
- ret['write'] = [v.replace(' ','') for v in perm[6:].split(',')]
- ret['write'].remove('')
+ ret['write'] = list(set([v.replace(' ','').lower() for v in perm[6:].split(',')]))
+ if '' in ret['write']:
+ ret['write'].remove('')
if '*' in ret['write']:
ret['write'] = ['*']
if len(ret['write']) == 0:
raise BadRequest('Bad X-Object-Sharing header value')
else:
raise BadRequest('Bad X-Object-Sharing header value')
+
+ # Keep duplicates only in write list.
+ dups = [x for x in ret.get('read', []) if x in ret.get('write', []) and x != '*']
+ if dups:
+ for x in dups:
+ ret['read'].remove(x)
+ if len(ret['read']) == 0:
+ del(ret['read'])
+
return ret
+def get_public(request):
+ """Parse an X-Object-Public header from the request.
+
+ Raises BadRequest on error.
+ """
+
+ public = request.META.get('HTTP_X_OBJECT_PUBLIC')
+ if public is None:
+ return None
+
+ public = public.replace(' ', '').lower()
+ if public == 'true':
+ return True
+ elif public == 'false' or public == '':
+ return False
+ raise BadRequest('Bad X-Object-Public header value')
+
def raw_input_socket(request):
"""Return the socket for reading the rest of the request."""
server_software = request.META.get('SERVER_SOFTWARE')
- if not server_software:
- if 'wsgi.input' in request.environ:
- return request.environ['wsgi.input']
- raise ServiceUnavailable('Unknown server software')
- if server_software.startswith('WSGIServer'):
- return request.environ['wsgi.input']
- elif server_software.startswith('mod_python'):
+ if server_software and server_software.startswith('mod_python'):
return request._req
+ if 'wsgi.input' in request.environ:
+ return request.environ['wsgi.input']
raise ServiceUnavailable('Unknown server software')
MAX_UPLOAD_SIZE = 10 * (1024 * 1024) # 10MB
-def socket_read_iterator(sock, length=0, blocksize=4096):
+def socket_read_iterator(request, length=0, blocksize=4096):
"""Return a maximum of blocksize data read from the socket in each iteration.
Read up to 'length'. If 'length' is negative, will attempt a chunked read.
The maximum ammount of data read is controlled by MAX_UPLOAD_SIZE.
"""
+ sock = raw_input_socket(request)
if length < 0: # Chunked transfers
+ # Small version (server does the dechunking).
+ if request.environ.get('mod_wsgi.input_chunked', None):
+ while length < MAX_UPLOAD_SIZE:
+ data = sock.read(blocksize)
+ if data == '':
+ return
+ yield data
+ raise BadRequest('Maximum size is reached')
+
+ # Long version (do the dechunking).
data = ''
while length < MAX_UPLOAD_SIZE:
# Get chunk size.
data = data[blocksize:]
yield ret
sock.read(2) # CRLF
- # TODO: Raise something to note that maximum size is reached.
+ raise BadRequest('Maximum size is reached')
else:
if length > MAX_UPLOAD_SIZE:
- # TODO: Raise something to note that maximum size is reached.
- pass
+ raise BadRequest('Maximum size is reached')
while length > 0:
data = sock.read(min(length, blocksize))
length -= len(data)
offset < 0 or offset >= size or
offset + length > size]
if len(check) > 0:
- raise RangeNotSatisfiable('Requested range exceeds object limits')
+ raise RangeNotSatisfiable('Requested range exceeds object limits')
ret = 206
+ if_range = request.META.get('HTTP_IF_RANGE')
+ if if_range:
+ try:
+ # Modification time has passed instead.
+ last_modified = parse_http_date(if_range)
+ if last_modified != meta['modified']:
+ ranges = [(0, size)]
+ ret = 200
+ except ValueError:
+ if if_range != meta['hash']:
+ ranges = [(0, size)]
+ ret = 200
if ret == 206 and len(ranges) > 1:
boundary = uuid.uuid4().hex
boundary = ''
wrapper = ObjectWrapper(ranges, sizes, hashmaps, boundary)
response = HttpResponse(wrapper, status=ret)
- put_object_meta(response, meta, public)
+ put_object_headers(response, meta, public)
if ret == 206:
if len(ranges) == 1:
offset, length = ranges[0]
if len(hashmap) == 0:
return hexlify(subhash(''))
if len(hashmap) == 1:
- return hexlify(subhash(hashmap[0]))
+ return hashmap[0]
+
s = 2
while s < len(hashmap):
s = s * 2
- h = hashmap + ([('\x00' * len(hashmap[0]))] * (s - len(hashmap)))
- h = [subhash(h[x] + (h[x + 1] if x + 1 < len(h) else '')) for x in range(0, len(h), 2)]
+ h = [unhexlify(x) for x in hashmap]
+ h += [('\x00' * len(h[0]))] * (s - len(hashmap))
while len(h) > 1:
- h = [subhash(h[x] + (h[x + 1] if x + 1 < len(h) else '')) for x in range(0, len(h), 2)]
+ h = [subhash(h[x] + h[x + 1]) for x in range(0, len(h), 2)]
return hexlify(h[0])
def update_response_headers(request, response):
response['Content-Type'] = 'application/json; charset=UTF-8'
elif not response['Content-Type']:
response['Content-Type'] = 'text/plain; charset=UTF-8'
-
+
+ if not response.has_header('Content-Length') and not (response.has_header('Content-Type') and response['Content-Type'].startswith('multipart/byteranges')):
+ response['Content-Length'] = len(response.content)
+
if settings.TEST:
response['Date'] = format_date_time(time())
def render_fault(request, fault):
if settings.DEBUG or settings.TEST:
fault.details = format_exc(fault)
-
+
request.serialization = 'text'
data = '\n'.join((fault.message, fault.details)) + '\n'
response = HttpResponse(data, status=fault.code)