Page Menu
Home
Software Heritage
Search
Configure Global Search
Log In
Files
F7123273
D2662.diff
No One
Temporary
Actions
View File
Edit File
Delete File
View Transforms
Subscribe
Mute Notifications
Award Token
Flag For Later
Size
7 KB
Subscribers
None
D2662.diff
View Options
diff --git a/swh/storage/cassandra/storage.py b/swh/storage/cassandra/storage.py
--- a/swh/storage/cassandra/storage.py
+++ b/swh/storage/cassandra/storage.py
@@ -17,6 +17,7 @@
Revision, Release, Directory, DirectoryEntry, Content, SkippedContent,
OriginVisit, Snapshot, Origin
)
+from swh.model.hashutil import DEFAULT_ALGORITHMS
from swh.storage.objstorage import ObjStorage
from swh.storage.writer import JournalWriter
@@ -257,7 +258,7 @@
def skipped_content_missing(self, contents):
for content in contents:
if not self._cql_runner.skipped_content_get_from_pk(content):
- yield content
+ yield {algo: content[algo] for algo in DEFAULT_ALGORITHMS}
def directory_add(self, directories: Iterable[Directory]) -> Dict:
directories = list(directories)
diff --git a/swh/storage/in_memory.py b/swh/storage/in_memory.py
--- a/swh/storage/in_memory.py
+++ b/swh/storage/in_memory.py
@@ -41,15 +41,15 @@
class InMemoryStorage:
def __init__(self, journal_writer=None):
- self._contents = {}
- self._content_indexes = defaultdict(lambda: defaultdict(set))
- self._skipped_contents = {}
- self._skipped_content_indexes = defaultdict(lambda: defaultdict(set))
self.reset()
self.journal_writer = JournalWriter(journal_writer)
def reset(self):
+ self._contents = {}
+ self._content_indexes = defaultdict(lambda: defaultdict(set))
+ self._skipped_contents = {}
+ self._skipped_content_indexes = defaultdict(lambda: defaultdict(set))
self._directories = {}
self._revisions = {}
self._releases = {}
@@ -251,45 +251,49 @@
def content_get_random(self):
return random.choice(list(self._content_indexes['sha1_git']))
- def _skipped_content_add(self, contents: Iterable[SkippedContent]) -> Dict:
+ def _skipped_content_add(self, contents: List[SkippedContent]) -> Dict:
self.journal_writer.skipped_content_add(contents)
summary = {
'skipped_content:add': 0
}
- skipped_content_missing = self.skipped_content_missing(
+ missing_contents = self.skipped_content_missing(
[c.to_dict() for c in contents])
- for content in skipped_content_missing:
- key = self._content_key(content, allow_missing=True)
+ missing = {self._content_key(c) for c in missing_contents}
+ contents = [c for c in contents
+ if self._content_key(c) in missing]
+ for content in contents:
+ key = self._content_key(content)
for algo in DEFAULT_ALGORITHMS:
- if content.get(algo):
+ if content.get_hash(algo):
self._skipped_content_indexes[algo][
- content.get(algo)].add(key)
+ content.get_hash(algo)].add(key)
self._skipped_contents[key] = content
summary['skipped_content:add'] += 1
return summary
- def skipped_content_missing(self, contents):
- for content in contents:
- for (key, algorithm) in self._content_key_algorithm(content):
- if algorithm == 'blake2s256':
- continue
- if key not in self._skipped_content_indexes[algorithm]:
- # index must contain hashes of algos except blake2s256
- # else the content is considered skipped
- yield {algo: content[algo]
- for algo in DEFAULT_ALGORITHMS
- if content[algo] is not None}
- break
-
def skipped_content_add(self, content: Iterable[SkippedContent]) -> Dict:
- content = list(content)
now = datetime.datetime.now(tz=datetime.timezone.utc)
content = [attr.evolve(c, ctime=now) for c in content]
return self._skipped_content_add(content)
+ def skipped_content_missing(self, contents):
+ for content in contents:
+ matches = list(self._skipped_contents.values())
+ for (algorithm, key) in self._content_key(content):
+ if algorithm == 'blake2s256':
+ continue
+ # Filter out skipped contents with the same hash
+ matches = [
+ match for match in matches
+ if match.get_hash(algorithm) == key]
+ # if none of the contents match
+ if not matches:
+ yield {algo: content[algo]
+ for algo in DEFAULT_ALGORITHMS}
+
def directory_add(self, directories: Iterable[Directory]) -> Dict:
directories = [dir_ for dir_ in directories
if dir_.id not in self._directories]
@@ -971,17 +975,11 @@
return person
@staticmethod
- def _content_key(content, allow_missing=False):
- """A stable key for a content"""
- return tuple(getattr(content, key, None)
- for key in sorted(DEFAULT_ALGORITHMS))
-
- @staticmethod
- def _content_key_algorithm(content):
+ def _content_key(content):
""" A stable key and the algorithm for a content"""
if isinstance(content, BaseContent):
content = content.to_dict()
- return tuple((content.get(key), key)
+ return tuple((key, content.get(key))
for key in sorted(DEFAULT_ALGORITHMS))
@staticmethod
diff --git a/swh/storage/tests/test_storage.py b/swh/storage/tests/test_storage.py
--- a/swh/storage/tests/test_storage.py
+++ b/swh/storage/tests/test_storage.py
@@ -339,6 +339,37 @@
missing = list(swh_storage.skipped_content_missing([cont, cont2]))
+ assert missing == [
+ {
+ 'sha1': cont['sha1'],
+ 'sha1_git': cont['sha1_git'],
+ 'blake2s256': cont['blake2s256'],
+ 'sha256': cont['sha256']
+ },
+ {
+ 'sha1': cont2['sha1'],
+ 'sha1_git': cont2['sha1_git'],
+ 'blake2s256': cont2['blake2s256'],
+ 'sha256': cont2['sha256']
+ },
+ ]
+
+ actual_result = swh_storage.skipped_content_add([cont, cont, cont2])
+
+ assert 2 <= actual_result.pop('skipped_content:add') <= 3
+ assert actual_result == {}
+
+ missing = list(swh_storage.skipped_content_missing([cont, cont2]))
+
+ assert missing == []
+
+ def test_skipped_content_add_missing_hashes(self, swh_storage):
+ cont = data.skipped_cont
+ cont2 = data.skipped_cont2
+ cont['sha1_git'] = cont2['sha1_git'] = None
+
+ missing = list(swh_storage.skipped_content_missing([cont, cont2]))
+
assert len(missing) == 2
actual_result = swh_storage.skipped_content_add([cont, cont, cont2])
@@ -350,6 +381,29 @@
assert missing == []
+ def test_skipped_content_missing_partial_hash(self, swh_storage):
+ cont = data.skipped_cont
+ cont2 = cont.copy()
+ cont2['sha1_git'] = None
+
+ missing = list(swh_storage.skipped_content_missing([cont, cont2]))
+
+ assert len(missing) == 2
+
+ actual_result = swh_storage.skipped_content_add([cont])
+
+ assert actual_result.pop('skipped_content:add') == 1
+ assert actual_result == {}
+
+ missing = list(swh_storage.skipped_content_missing([cont, cont2]))
+
+ assert missing == [{
+ 'sha1': cont2['sha1'],
+ 'sha1_git': cont2['sha1_git'],
+ 'blake2s256': cont2['blake2s256'],
+ 'sha256': cont2['sha256']
+ }]
+
@pytest.mark.property_based
@settings(deadline=None) # this test is very slow
@given(strategies.sets(
File Metadata
Details
Attached
Mime Type
text/plain
Expires
Wed, Dec 18, 9:59 AM (19 h, 42 m ago)
Storage Engine
blob
Storage Format
Raw Data
Storage Handle
3221865
Attached To
D2662: Fix inconsistent behavior of skipped_content_missing across backends.
Event Timeline
Log In to Comment