diff --git a/swh/objstorage/objstorage.py b/swh/objstorage/objstorage.py index b72214d..68e0340 100644 --- a/swh/objstorage/objstorage.py +++ b/swh/objstorage/objstorage.py @@ -1,278 +1,288 @@ -# Copyright (C) 2015-2017 The Software Heritage developers +# Copyright (C) 2015-2018 The Software Heritage developers # See the AUTHORS file at the top-level directory of this distribution # License: GNU General Public License version 3, or any later version # See top-level LICENSE file for more information import abc from swh.model import hashutil from .exc import ObjNotFoundError ID_HASH_ALGO = 'sha1' ID_HASH_LENGTH = 40 # Size in bytes of the hash hexadecimal representation. DEFAULT_CHUNK_SIZE = 2 * 1024 * 1024 # Size in bytes of the streaming chunks def compute_hash(content): - return hashutil.hash_data( + """Compute the content's hash. + + Args: + content (bytes): The raw content to hash + hash_name (str): Hash's name (default to ID_HASH_ALGO) + + Returns: + The ID_HASH_ALGO for the content + + """ + return hashutil.MultiHash.from_data( content, - algorithms=[ID_HASH_ALGO] - ).get(ID_HASH_ALGO) + hash_names=[ID_HASH_ALGO], + ).digest().get(ID_HASH_ALGO) class ObjStorage(metaclass=abc.ABCMeta): """ High-level API to manipulate the Software Heritage object storage. Conceptually, the object storage offers the following methods: - check_config() check if the object storage is properly configured - __contains__() check if an object is present, by object id - add() add a new object, returning an object id - restore() same as add() but erase an already existed content - get() retrieve the content of an object, by object id - check() check the integrity of an object, by object id - delete() remove an object And some management methods: - get_random() get random object id of existing contents (used for the content integrity checker). Some of the methods have available streaming equivalents: - add_stream() same as add() but with a chunked iterator - restore_stream() same as add_stream() but erase already existing content - get_stream() same as get() but returns a chunked iterator Each implementation of this interface can have a different behavior and its own way to store the contents. """ def __init__(self, *, allow_delete=False, **kwargs): # A more complete permission system could be used in place of that if # it becomes needed super().__init__(**kwargs) self.allow_delete = allow_delete @abc.abstractmethod def check_config(self, *, check_write): """Check whether the object storage is properly configured. Args: check_write (bool): if True, check if writes to the object storage can succeed. Returns: True if the configuration check worked, an exception if it didn't. """ pass @abc.abstractmethod def __contains__(self, obj_id, *args, **kwargs): """Indicate if the given object is present in the storage. Args: obj_id (bytes): object identifier. Returns: True if and only if the object is present in the current object storage. """ pass @abc.abstractmethod def add(self, content, obj_id=None, check_presence=True, *args, **kwargs): """Add a new object to the object storage. Args: content (bytes): object's raw content to add in storage. obj_id (bytes): checksum of [bytes] using [ID_HASH_ALGO] algorithm. When given, obj_id will be trusted to match the bytes. If missing, obj_id will be computed on the fly. check_presence (bool): indicate if the presence of the content should be verified before adding the file. Returns: the id (bytes) of the object into the storage. """ pass def add_batch(self, contents, check_presence=True): """Add a batch of new objects to the object storage. Args: contents (dict): mapping from obj_id to object conetnts Returns: the number of objects added to the storage """ ctr = 0 for obj_id, content in contents.items(): self.add(content, obj_id, check_presence=check_presence) ctr += 1 return ctr def restore(self, content, obj_id=None, *args, **kwargs): """Restore a content that have been corrupted. This function is identical to add but does not check if the object id is already in the file system. The default implementation provided by the current class is suitable for most cases. Args: content (bytes): object's raw content to add in storage obj_id (bytes): checksum of `bytes` as computed by ID_HASH_ALGO. When given, obj_id will be trusted to match bytes. If missing, obj_id will be computed on the fly. """ # check_presence to false will erase the potential previous content. return self.add(content, obj_id, check_presence=False) @abc.abstractmethod def get(self, obj_id, *args, **kwargs): """Retrieve the content of a given object. Args: obj_id (bytes): object id. Returns: the content of the requested object as bytes. Raises: ObjNotFoundError: if the requested object is missing. """ pass def get_batch(self, obj_ids, *args, **kwargs): """Retrieve objects' raw content in bulk from storage. Note: This function does have a default implementation in ObjStorage that is suitable for most cases. For object storages that needs to do the minimal number of requests possible (ex: remote object storages), that method can be overridden to perform a more efficient operation. Args: obj_ids ([bytes]: list of object ids. Returns: list of resulting contents, or None if the content could not be retrieved. Do not raise any exception as a fail for one content will not cancel the whole request. """ for obj_id in obj_ids: try: yield self.get(obj_id) except ObjNotFoundError: yield None @abc.abstractmethod def check(self, obj_id, *args, **kwargs): """Perform an integrity check for a given object. Verify that the file object is in place and that the gziped content matches the object id. Args: obj_id (bytes): object identifier. Raises: ObjNotFoundError: if the requested object is missing. Error: if the request object is corrupted. """ pass @abc.abstractmethod def delete(self, obj_id, *args, **kwargs): """Delete an object. Args: obj_id (bytes): object identifier. Raises: ObjNotFoundError: if the requested object is missing. """ if not self.allow_delete: raise PermissionError("Delete is not allowed.") # Management methods def get_random(self, batch_size, *args, **kwargs): """Get random ids of existing contents. This method is used in order to get random ids to perform content integrity verifications on random contents. Args: batch_size (int): Number of ids that will be given Yields: An iterable of ids (bytes) of contents that are in the current object storage. """ pass # Streaming methods def add_stream(self, content_iter, obj_id, check_presence=True): """Add a new object to the object storage using streaming. This function is identical to add() except it takes a generator that yields the chunked content instead of the whole content at once. Args: content (bytes): chunked generator that yields the object's raw content to add in storage. obj_id (bytes): object identifier check_presence (bool): indicate if the presence of the content should be verified before adding the file. Returns: the id (bytes) of the object into the storage. """ raise NotImplementedError def restore_stream(self, content_iter, obj_id=None): """Restore a content that have been corrupted using streaming. This function is identical to restore() except it takes a generator that yields the chunked content instead of the whole content at once. The default implementation provided by the current class is suitable for most cases. Args: content (bytes): chunked generator that yields the object's raw content to add in storage. obj_id (bytes): object identifier """ # check_presence to false will erase the potential previous content. return self.add_stream(content_iter, obj_id, check_presence=False) def get_stream(self, obj_id, chunk_size=DEFAULT_CHUNK_SIZE): """Retrieve the content of a given object as a chunked iterator. Args: obj_id (bytes): object id. Returns: the content of the requested object as bytes. Raises: ObjNotFoundError: if the requested object is missing. """ raise NotImplementedError diff --git a/swh/objstorage/objstorage_pathslicing.py b/swh/objstorage/objstorage_pathslicing.py index 96f152d..a6fe1b2 100644 --- a/swh/objstorage/objstorage_pathslicing.py +++ b/swh/objstorage/objstorage_pathslicing.py @@ -1,340 +1,340 @@ -# Copyright (C) 2015-2017 The Software Heritage developers +# Copyright (C) 2015-2018 The Software Heritage developers # See the AUTHORS file at the top-level directory of this distribution # License: GNU General Public License version 3, or any later version # See top-level LICENSE file for more information import functools import os import gzip import tempfile import random from contextlib import contextmanager from swh.model import hashutil from .objstorage import (ObjStorage, compute_hash, ID_HASH_ALGO, ID_HASH_LENGTH, DEFAULT_CHUNK_SIZE) from .exc import ObjNotFoundError, Error GZIP_BUFSIZ = 1048576 DIR_MODE = 0o755 FILE_MODE = 0o644 @contextmanager def _write_obj_file(hex_obj_id, objstorage): """ Context manager for writing object files to the object storage. During writing, data are written to a temporary file, which is atomically renamed to the right file name after closing. This context manager also takes care of (gzip) compressing the data on the fly. Usage sample: with _write_obj_file(hex_obj_id, objstorage): f.write(obj_data) Yields: a file-like object open for writing bytes. """ # Get the final paths and create the directory if absent. dir = objstorage._obj_dir(hex_obj_id) if not os.path.isdir(dir): os.makedirs(dir, DIR_MODE, exist_ok=True) path = os.path.join(dir, hex_obj_id) # Create a temporary file. (tmp, tmp_path) = tempfile.mkstemp(suffix='.tmp', prefix='hex_obj_id.', dir=dir) # Open the file and yield it for writing. tmp_f = os.fdopen(tmp, 'wb') with gzip.GzipFile(filename=tmp_path, fileobj=tmp_f) as f: yield f # Then close the temporary file and move it to the right directory. tmp_f.close() os.chmod(tmp_path, FILE_MODE) os.rename(tmp_path, path) @contextmanager def _read_obj_file(hex_obj_id, objstorage): """ Context manager for reading object file in the object storage. Usage sample: with _read_obj_file(hex_obj_id, objstorage) as f: b = f.read() Yields: a file-like object open for reading bytes. """ path = objstorage._obj_path(hex_obj_id) with gzip.GzipFile(path, 'rb') as f: yield f class PathSlicingObjStorage(ObjStorage): """Implementation of the ObjStorage API based on the hash of the content. On disk, an object storage is a directory tree containing files named after their object IDs. An object ID is a checksum of its content, depending on the value of the ID_HASH_ALGO constant (see swh.model.hashutil for its meaning). To avoid directories that contain too many files, the object storage has a given slicing. Each slicing correspond to a directory that is named according to the hash of its content. So for instance a file with SHA1 34973274ccef6ab4dfaaf86599792fa9c3fe4689 will be stored in the given object storages : - 0:2/2:4/4:6 : 34/97/32/34973274ccef6ab4dfaaf86599792fa9c3fe4689 - 0:1/0:5/ : 3/34973/34973274ccef6ab4dfaaf86599792fa9c3fe4689 The files in the storage are stored in gzipped compressed format. Attributes: root (string): path to the root directory of the storage on the disk. bounds: list of tuples that indicates the beginning and the end of each subdirectory for a content. """ def __init__(self, root, slicing, **kwargs): """ Create an object to access a hash-slicing based object storage. Args: root (string): path to the root directory of the storage on the disk. slicing (string): string that indicates the slicing to perform on the hash of the content to know the path where it should be stored. """ super().__init__(**kwargs) self.root = root # Make a list of tuples where each tuple contains the beginning # and the end of each slicing. self.bounds = [ slice(*map(int, sbounds.split(':'))) for sbounds in slicing.split('/') if sbounds ] self.check_config(check_write=False) def check_config(self, *, check_write): """Check whether this object storage is properly configured""" root = self.root if not os.path.isdir(root): raise ValueError( 'PathSlicingObjStorage root "%s" is not a directory' % root ) max_endchar = max(map(lambda bound: bound.stop, self.bounds)) if ID_HASH_LENGTH < max_endchar: raise ValueError( 'Algorithm %s has too short hash for slicing to char %d' % (ID_HASH_ALGO, max_endchar) ) if check_write: if not os.access(self.root, os.W_OK): raise PermissionError( 'PathSlicingObjStorage root "%s" is not writable' % root ) return True def __contains__(self, obj_id): hex_obj_id = hashutil.hash_to_hex(obj_id) return os.path.isfile(self._obj_path(hex_obj_id)) def __iter__(self): """Iterate over the object identifiers currently available in the storage. Warning: with the current implementation of the object storage, this method will walk the filesystem to list objects, meaning that listing all objects will be very slow for large storages. You almost certainly don't want to use this method in production. Return: Iterator over object IDs """ def obj_iterator(): # XXX hackish: it does not verify that the depth of found files # matches the slicing depth of the storage for root, _dirs, files in os.walk(self.root): for f in files: yield bytes.fromhex(f) return obj_iterator() def __len__(self): """Compute the number of objects available in the storage. Warning: this currently uses `__iter__`, its warning about bad performances applies Return: number of objects contained in the storage """ return sum(1 for i in self) def _obj_dir(self, hex_obj_id): """ Compute the storage directory of an object. See also: PathSlicingObjStorage::_obj_path Args: hex_obj_id: object id as hexlified string. Returns: Path to the directory that contains the required object. """ slices = [hex_obj_id[bound] for bound in self.bounds] return os.path.join(self.root, *slices) def _obj_path(self, hex_obj_id): """ Compute the full path to an object into the current storage. See also: PathSlicingObjStorage::_obj_dir Args: hex_obj_id: object id as hexlified string. Returns: Path to the actual object corresponding to the given id. """ return os.path.join(self._obj_dir(hex_obj_id), hex_obj_id) def add(self, content, obj_id=None, check_presence=True): if obj_id is None: obj_id = compute_hash(content) if check_presence and obj_id in self: # If the object is already present, return immediately. return obj_id hex_obj_id = hashutil.hash_to_hex(obj_id) with _write_obj_file(hex_obj_id, self) as f: f.write(content) return obj_id def get(self, obj_id): if obj_id not in self: raise ObjNotFoundError(obj_id) # Open the file and return its content as bytes hex_obj_id = hashutil.hash_to_hex(obj_id) with _read_obj_file(hex_obj_id, self) as f: return f.read() def check(self, obj_id): if obj_id not in self: raise ObjNotFoundError(obj_id) hex_obj_id = hashutil.hash_to_hex(obj_id) try: with gzip.open(self._obj_path(hex_obj_id)) as f: length = None if ID_HASH_ALGO.endswith('_git'): # if the hashing algorithm is git-like, we need to know the # content size to hash on the fly. Do a first pass here to # compute the size length = 0 while True: chunk = f.read(GZIP_BUFSIZ) length += len(chunk) if not chunk: break f.rewind() - checksums = hashutil.hash_file(f, length, - algorithms=[ID_HASH_ALGO]) + checksums = hashutil.MultiHash.from_file( + f, hash_names=[ID_HASH_ALGO], length=length).digest() actual_obj_id = checksums[ID_HASH_ALGO] if obj_id != actual_obj_id: raise Error( 'Corrupt object %s should have id %s' % (hashutil.hash_to_hex(obj_id), hashutil.hash_to_hex(actual_obj_id)) ) except (OSError, IOError): # IOError is for compatibility with older python versions raise Error('Corrupt object %s is not a gzip file' % obj_id) def delete(self, obj_id): super().delete(obj_id) # Check delete permission if obj_id not in self: raise ObjNotFoundError(obj_id) hex_obj_id = hashutil.hash_to_hex(obj_id) try: os.remove(self._obj_path(hex_obj_id)) except FileNotFoundError: raise ObjNotFoundError(obj_id) return True # Management methods def get_random(self, batch_size): def get_random_content(self, batch_size): """ Get a batch of content inside a single directory. Returns: a tuple (batch size, batch). """ dirs = [] for level in range(len(self.bounds)): path = os.path.join(self.root, *dirs) dir_list = next(os.walk(path))[1] if 'tmp' in dir_list: dir_list.remove('tmp') dirs.append(random.choice(dir_list)) path = os.path.join(self.root, *dirs) content_list = next(os.walk(path))[2] length = min(batch_size, len(content_list)) return length, map(hashutil.hash_to_bytes, random.sample(content_list, length)) while batch_size: length, it = get_random_content(self, batch_size) batch_size = batch_size - length yield from it # Streaming methods @contextmanager def chunk_writer(self, obj_id): hex_obj_id = hashutil.hash_to_hex(obj_id) with _write_obj_file(hex_obj_id, self) as f: yield f.write def add_stream(self, content_iter, obj_id, check_presence=True): if check_presence and obj_id in self: return obj_id with self.chunk_writer(obj_id) as writer: for chunk in content_iter: writer(chunk) return obj_id def get_stream(self, obj_id, chunk_size=DEFAULT_CHUNK_SIZE): if obj_id not in self: raise ObjNotFoundError(obj_id) hex_obj_id = hashutil.hash_to_hex(obj_id) with _read_obj_file(hex_obj_id, self) as f: reader = functools.partial(f.read, chunk_size) yield from iter(reader, b'') diff --git a/swh/objstorage/tests/objstorage_testing.py b/swh/objstorage/tests/objstorage_testing.py index 79835f2..698807b 100644 --- a/swh/objstorage/tests/objstorage_testing.py +++ b/swh/objstorage/tests/objstorage_testing.py @@ -1,169 +1,169 @@ -# Copyright (C) 2015-2017 The Software Heritage developers +# Copyright (C) 2015-2018 The Software Heritage developers # See the AUTHORS file at the top-level directory of this distribution # License: GNU General Public License version 3, or any later version # See top-level LICENSE file for more information import time -from swh.model import hashutil from swh.objstorage import exc +from swh.objstorage.objstorage import compute_hash class ObjStorageTestFixture(): def setUp(self): super().setUp() def hash_content(self, content): - obj_id = hashutil.hash_data(content)['sha1'] + obj_id = compute_hash(content) return content, obj_id def assertContentMatch(self, obj_id, expected_content): # noqa content = self.storage.get(obj_id) self.assertEqual(content, expected_content) def test_check_config(self): self.assertTrue(self.storage.check_config(check_write=False)) self.assertTrue(self.storage.check_config(check_write=True)) def test_contains(self): content_p, obj_id_p = self.hash_content(b'contains_present') content_m, obj_id_m = self.hash_content(b'contains_missing') self.storage.add(content_p, obj_id=obj_id_p) self.assertIn(obj_id_p, self.storage) self.assertNotIn(obj_id_m, self.storage) def test_add_get_w_id(self): content, obj_id = self.hash_content(b'add_get_w_id') r = self.storage.add(content, obj_id=obj_id) self.assertEqual(obj_id, r) self.assertContentMatch(obj_id, content) def test_add_big(self): content, obj_id = self.hash_content(b'add_big' * 1024 * 1024) r = self.storage.add(content, obj_id=obj_id) self.assertEqual(obj_id, r) self.assertContentMatch(obj_id, content) def test_add_get_wo_id(self): content, obj_id = self.hash_content(b'add_get_wo_id') r = self.storage.add(content) self.assertEqual(obj_id, r) self.assertContentMatch(obj_id, content) def test_add_get_batch(self): content1, obj_id1 = self.hash_content(b'add_get_batch_1') content2, obj_id2 = self.hash_content(b'add_get_batch_2') self.storage.add(content1, obj_id1) self.storage.add(content2, obj_id2) cr1, cr2 = self.storage.get_batch([obj_id1, obj_id2]) self.assertEqual(cr1, content1) self.assertEqual(cr2, content2) def test_get_batch_unexisting_content(self): content, obj_id = self.hash_content(b'get_batch_unexisting_content') result = list(self.storage.get_batch([obj_id])) self.assertTrue(len(result) == 1) self.assertIsNone(result[0]) def test_restore_content(self): valid_content, valid_obj_id = self.hash_content(b'restore_content') invalid_content = b'unexpected content' id_adding = self.storage.add(invalid_content, valid_obj_id) self.assertEqual(id_adding, valid_obj_id) with self.assertRaises(exc.Error): self.storage.check(id_adding) id_restore = self.storage.restore(valid_content, valid_obj_id) self.assertEqual(id_restore, valid_obj_id) self.assertContentMatch(valid_obj_id, valid_content) def test_get_missing(self): content, obj_id = self.hash_content(b'get_missing') with self.assertRaises(exc.ObjNotFoundError) as e: self.storage.get(obj_id) self.assertIn(obj_id, e.exception.args) def test_check_missing(self): content, obj_id = self.hash_content(b'check_missing') with self.assertRaises(exc.Error): self.storage.check(obj_id) def test_check_present(self): content, obj_id = self.hash_content(b'check_present') self.storage.add(content, obj_id) try: self.storage.check(obj_id) except exc.Error: self.fail('Integrity check failed') def test_delete_missing(self): self.storage.allow_delete = True content, obj_id = self.hash_content(b'missing_content_to_delete') with self.assertRaises(exc.Error): self.storage.delete(obj_id) def test_delete_present(self): self.storage.allow_delete = True content, obj_id = self.hash_content(b'content_to_delete') self.storage.add(content, obj_id=obj_id) self.assertTrue(self.storage.delete(obj_id)) with self.assertRaises(exc.Error): self.storage.get(obj_id) def test_delete_not_allowed(self): self.storage.allow_delete = False content, obj_id = self.hash_content(b'content_to_delete') self.storage.add(content, obj_id=obj_id) with self.assertRaises(PermissionError): self.assertTrue(self.storage.delete(obj_id)) def test_delete_not_allowed_by_default(self): content, obj_id = self.hash_content(b'content_to_delete') self.storage.add(content, obj_id=obj_id) with self.assertRaises(PermissionError): self.assertTrue(self.storage.delete(obj_id)) def test_add_stream(self): content = [b'chunk1', b'chunk2'] _, obj_id = self.hash_content(b''.join(content)) try: self.storage.add_stream(iter(content), obj_id=obj_id) except NotImplementedError: return self.assertContentMatch(obj_id, b''.join(content)) def test_add_stream_sleep(self): def gen_content(): yield b'chunk1' time.sleep(0.5) yield b'chunk2' _, obj_id = self.hash_content(b'placeholder_id') try: self.storage.add_stream(gen_content(), obj_id=obj_id) except NotImplementedError: return self.assertContentMatch(obj_id, b'chunk1chunk2') def test_get_stream(self): content_l = [b'1', b'2', b'3', b'4', b'5', b'6', b'7', b'8', b'9'] content = b''.join(content_l) _, obj_id = self.hash_content(content) self.storage.add(content, obj_id=obj_id) try: r = list(self.storage.get_stream(obj_id, chunk_size=1)) except NotImplementedError: return self.assertEqual(r, content_l) def test_add_batch(self): contents = {} for i in range(50): content = b'Test content %02d' % i content, obj_id = self.hash_content(content) contents[obj_id] = content ret = self.storage.add_batch(contents) self.assertEqual(len(contents), ret) for obj_id in contents: self.assertIn(obj_id, self.storage) diff --git a/swh/objstorage/tests/test_multiplexer_filter.py b/swh/objstorage/tests/test_multiplexer_filter.py index 6e8e4e3..d292d2f 100644 --- a/swh/objstorage/tests/test_multiplexer_filter.py +++ b/swh/objstorage/tests/test_multiplexer_filter.py @@ -1,323 +1,324 @@ -# Copyright (C) 2015-2017 The Software Heritage developers +# Copyright (C) 2015-2018 The Software Heritage developers # See the AUTHORS file at the top-level directory of this distribution # License: GNU General Public License version 3, or any later version # See top-level LICENSE file for more information import random import shutil import tempfile import unittest from string import ascii_lowercase from swh.model import hashutil from swh.objstorage import get_objstorage from swh.objstorage.exc import Error, ObjNotFoundError from swh.objstorage.multiplexer.filter import id_prefix, id_regex, read_only +from swh.objstorage.objstorage import compute_hash def get_random_content(): return bytes(''.join(random.sample(ascii_lowercase, 10)), 'utf8') class MixinTestReadFilter(unittest.TestCase): # Read only filter should not allow writing def setUp(self): super().setUp() self.tmpdir = tempfile.mkdtemp() pstorage = {'cls': 'pathslicing', 'args': {'root': self.tmpdir, 'slicing': '0:5'}} base_storage = get_objstorage(**pstorage) - base_storage.id = lambda cont: hashutil.hash_data(cont)['sha1'] + base_storage.id = compute_hash self.storage = get_objstorage('filtered', {'storage_conf': pstorage, 'filters_conf': [read_only()]}) self.valid_content = b'pre-existing content' self.invalid_content = b'invalid_content' self.true_invalid_content = b'Anything that is not correct' self.absent_content = b'non-existent content' # Create a valid content. self.valid_id = base_storage.add(self.valid_content) # Create an invalid id and add a content with it. self.invalid_id = base_storage.id(self.true_invalid_content) base_storage.add(self.invalid_content, obj_id=self.invalid_id) # Compute an id for a non-existing content. self.absent_id = base_storage.id(self.absent_content) def tearDown(self): super().tearDown() shutil.rmtree(self.tmpdir) def test_can_contains(self): self.assertTrue(self.valid_id in self.storage) self.assertTrue(self.invalid_id in self.storage) self.assertFalse(self.absent_id in self.storage) def test_can_iter(self): self.assertIn(self.valid_id, iter(self.storage)) self.assertIn(self.invalid_id, iter(self.storage)) def test_can_len(self): self.assertEqual(2, len(self.storage)) def test_can_get(self): self.assertEqual(self.valid_content, self.storage.get(self.valid_id)) self.assertEqual(self.invalid_content, self.storage.get(self.invalid_id)) def test_can_check(self): with self.assertRaises(ObjNotFoundError): self.storage.check(self.absent_id) with self.assertRaises(Error): self.storage.check(self.invalid_id) self.storage.check(self.valid_id) def test_can_get_random(self): self.assertEqual(1, len(list(self.storage.get_random(1)))) self.assertEqual(len(list(self.storage)), len(set(self.storage.get_random(1000)))) def test_cannot_add(self): new_id = self.storage.add(b'New content') result = self.storage.add(self.valid_content, self.valid_id) self.assertIsNone(new_id, self.storage) self.assertIsNone(result) def test_cannot_restore(self): result = self.storage.restore(self.valid_content, self.valid_id) self.assertIsNone(result) class MixinTestIdFilter(): """ Mixin class that tests the filters based on filter.IdFilter Methods "make_valid", "make_invalid" and "filter_storage" must be implemented by subclasses. """ def setUp(self): super().setUp() # Use a hack here : as the mock uses the content as id, it is easy to # create contents that are filtered or not. self.prefix = '71' self.tmpdir = tempfile.mkdtemp() # Make the storage filtered self.sconf = {'cls': 'pathslicing', 'args': {'root': self.tmpdir, 'slicing': '0:5'}} storage = get_objstorage(**self.sconf) self.base_storage = storage self.storage = self.filter_storage(self.sconf) # Set the id calculators - storage.id = lambda cont: hashutil.hash_data(cont)['sha1'] + storage.id = compute_hash # Present content with valid id self.present_valid_content = self.ensure_valid(b'yroqdtotji') self.present_valid_id = storage.id(self.present_valid_content) # Present content with invalid id self.present_invalid_content = self.ensure_invalid(b'glxddlmmzb') self.present_invalid_id = storage.id(self.present_invalid_content) # Missing content with valid id self.missing_valid_content = self.ensure_valid(b'rmzkdclkez') self.missing_valid_id = storage.id(self.missing_valid_content) # Missing content with invalid id self.missing_invalid_content = self.ensure_invalid(b'hlejfuginh') self.missing_invalid_id = storage.id(self.missing_invalid_content) # Present corrupted content with valid id self.present_corrupted_valid_content = self.ensure_valid(b'cdsjwnpaij') self.true_present_corrupted_valid_content = self.ensure_valid( b'mgsdpawcrr') self.present_corrupted_valid_id = storage.id( self.true_present_corrupted_valid_content) # Present corrupted content with invalid id self.present_corrupted_invalid_content = self.ensure_invalid( b'pspjljnrco') self.true_present_corrupted_invalid_content = self.ensure_invalid( b'rjocbnnbso') self.present_corrupted_invalid_id = storage.id( self.true_present_corrupted_invalid_content) # Missing (potentially) corrupted content with valid id self.missing_corrupted_valid_content = self.ensure_valid( b'zxkokfgtou') self.true_missing_corrupted_valid_content = self.ensure_valid( b'royoncooqa') self.missing_corrupted_valid_id = storage.id( self.true_missing_corrupted_valid_content) # Missing (potentially) corrupted content with invalid id self.missing_corrupted_invalid_content = self.ensure_invalid( b'hxaxnrmnyk') self.true_missing_corrupted_invalid_content = self.ensure_invalid( b'qhbolyuifr') self.missing_corrupted_invalid_id = storage.id( self.true_missing_corrupted_invalid_content) # Add the content that are supposed to be present self.storage.add(self.present_valid_content) self.storage.add(self.present_invalid_content) self.storage.add(self.present_corrupted_valid_content, obj_id=self.present_corrupted_valid_id) self.storage.add(self.present_corrupted_invalid_content, obj_id=self.present_corrupted_invalid_id) def tearDown(self): super().tearDown() shutil.rmtree(self.tmpdir) def filter_storage(self, sconf): raise NotImplementedError( 'Id_filter test class must have a filter_storage method') def ensure_valid(self, content=None): if content is None: content = get_random_content() while not self.storage.is_valid(self.base_storage.id(content)): content = get_random_content() return content def ensure_invalid(self, content=None): if content is None: content = get_random_content() while self.storage.is_valid(self.base_storage.id(content)): content = get_random_content() return content def test_contains(self): # Both contents are present, but the invalid one should be ignored. self.assertTrue(self.present_valid_id in self.storage) self.assertFalse(self.present_invalid_id in self.storage) self.assertFalse(self.missing_valid_id in self.storage) self.assertFalse(self.missing_invalid_id in self.storage) self.assertTrue(self.present_corrupted_valid_id in self.storage) self.assertFalse(self.present_corrupted_invalid_id in self.storage) self.assertFalse(self.missing_corrupted_valid_id in self.storage) self.assertFalse(self.missing_corrupted_invalid_id in self.storage) def test_iter(self): self.assertIn(self.present_valid_id, iter(self.storage)) self.assertNotIn(self.present_invalid_id, iter(self.storage)) self.assertNotIn(self.missing_valid_id, iter(self.storage)) self.assertNotIn(self.missing_invalid_id, iter(self.storage)) self.assertIn(self.present_corrupted_valid_id, iter(self.storage)) self.assertNotIn(self.present_corrupted_invalid_id, iter(self.storage)) self.assertNotIn(self.missing_corrupted_valid_id, iter(self.storage)) self.assertNotIn(self.missing_corrupted_invalid_id, iter(self.storage)) def test_len(self): # Four contents are present, but only two should be valid. self.assertEqual(2, len(self.storage)) def test_get(self): self.assertEqual(self.present_valid_content, self.storage.get(self.present_valid_id)) with self.assertRaises(ObjNotFoundError): self.storage.get(self.present_invalid_id) with self.assertRaises(ObjNotFoundError): self.storage.get(self.missing_valid_id) with self.assertRaises(ObjNotFoundError): self.storage.get(self.missing_invalid_id) self.assertEqual(self.present_corrupted_valid_content, self.storage.get(self.present_corrupted_valid_id)) with self.assertRaises(ObjNotFoundError): self.storage.get(self.present_corrupted_invalid_id) with self.assertRaises(ObjNotFoundError): self.storage.get(self.missing_corrupted_valid_id) with self.assertRaises(ObjNotFoundError): self.storage.get(self.missing_corrupted_invalid_id) def test_check(self): self.storage.check(self.present_valid_id) with self.assertRaises(ObjNotFoundError): self.storage.check(self.present_invalid_id) with self.assertRaises(ObjNotFoundError): self.storage.check(self.missing_valid_id) with self.assertRaises(ObjNotFoundError): self.storage.check(self.missing_invalid_id) with self.assertRaises(Error): self.storage.check(self.present_corrupted_valid_id) with self.assertRaises(ObjNotFoundError): self.storage.check(self.present_corrupted_invalid_id) with self.assertRaises(ObjNotFoundError): self.storage.check(self.missing_corrupted_valid_id) with self.assertRaises(ObjNotFoundError): self.storage.check(self.missing_corrupted_invalid_id) def test_get_random(self): self.assertEqual(0, len(list(self.storage.get_random(0)))) random_content = list(self.storage.get_random(1000)) self.assertIn(self.present_valid_id, random_content) self.assertNotIn(self.present_invalid_id, random_content) self.assertNotIn(self.missing_valid_id, random_content) self.assertNotIn(self.missing_invalid_id, random_content) self.assertIn(self.present_corrupted_valid_id, random_content) self.assertNotIn(self.present_corrupted_invalid_id, random_content) self.assertNotIn(self.missing_corrupted_valid_id, random_content) self.assertNotIn(self.missing_corrupted_invalid_id, random_content) def test_add(self): # Add valid and invalid contents to the storage and check their # presence with the unfiltered storage. valid_content = self.ensure_valid(b'ulepsrjbgt') valid_id = self.base_storage.id(valid_content) invalid_content = self.ensure_invalid(b'znvghkjked') invalid_id = self.base_storage.id(invalid_content) self.storage.add(valid_content) self.storage.add(invalid_content) self.assertTrue(valid_id in self.base_storage) self.assertFalse(invalid_id in self.base_storage) def test_restore(self): # Add corrupted content to the storage and the try to restore it valid_content = self.ensure_valid(b'ulepsrjbgt') valid_id = self.base_storage.id(valid_content) corrupted_content = self.ensure_valid(b'ltjkjsloyb') corrupted_id = self.base_storage.id(corrupted_content) self.storage.add(corrupted_content, obj_id=valid_id) with self.assertRaises(ObjNotFoundError): self.storage.check(corrupted_id) with self.assertRaises(Error): self.storage.check(valid_id) self.storage.restore(valid_content) self.storage.check(valid_id) class TestPrefixFilter(MixinTestIdFilter, unittest.TestCase): def setUp(self): self.prefix = b'71' super().setUp() def ensure_valid(self, content): - obj_id = hashutil.hash_data(content)['sha1'] + obj_id = compute_hash(content) hex_obj_id = hashutil.hash_to_hex(obj_id) self.assertTrue(hex_obj_id.startswith(self.prefix)) return content def ensure_invalid(self, content): - obj_id = hashutil.hash_data(content)['sha1'] + obj_id = compute_hash(content) hex_obj_id = hashutil.hash_to_hex(obj_id) self.assertFalse(hex_obj_id.startswith(self.prefix)) return content def filter_storage(self, sconf): return get_objstorage('filtered', {'storage_conf': sconf, 'filters_conf': [id_prefix(self.prefix)]}) class TestRegexFilter(MixinTestIdFilter, unittest.TestCase): def setUp(self): self.regex = r'[a-f][0-9].*' super().setUp() def filter_storage(self, sconf): return get_objstorage('filtered', {'storage_conf': sconf, 'filters_conf': [id_regex(self.regex)]})