Details
- Reviewers
olasd - Group Reviewers
Reviewers - Commits
- rDCIDX8c5a84459d44: Add typing to detect_metadata() and related functions
Diff Detail
- Repository
- rDCIDX Metadata indexer
- Lint
No Linters Available - Unit
No Unit Test Coverage - Build Status
Buildable 30207 Build 47217: Phabricator diff pipeline on jenkins Jenkins console · Jenkins Build 47216: arc lint + arc unit
Event Timeline
Build is green
Patch application report for D8045 (id=28984)
Rebasing onto 1be4e184d4...
Current branch diff-target is up to date.
Changes applied before test
commit eef33fc72d1ec86db15c12ca2eafad03719e0e48
Author: Valentin Lorentz <vlorentz@softwareheritage.org>
Date: Wed Jun 29 11:01:35 2022 +0200
Add typing to detect_metadata() and related functionsSee https://jenkins.softwareheritage.org/job/DCIDX/job/tests-on-diff/285/ for more details.
Build is green
Patch application report for D8045 (id=28987)
Rebasing onto 1be4e184d4...
Current branch diff-target is up to date.
Changes applied before test
commit f7a4bf4e04b3ac4c2fa89cf9b8a5c22e5f0c4d12
Author: Valentin Lorentz <vlorentz@softwareheritage.org>
Date: Wed Jun 29 11:01:35 2022 +0200
Add typing to detect_metadata() and related functionsSee https://jenkins.softwareheritage.org/job/DCIDX/job/tests-on-diff/286/ for more details.
| swh/indexer/metadata.py | ||
|---|---|---|
| 205 | I don't understand this assertion. Is it that the data argument is now deprecated? Also looks to me that with this chunk, this diff is not "just" adding typing. | |
| 218 | same, is this detected_files removal just a typing kinda stuff? looks not obvious to me | |
| 282–284 | ah I see, you moved the behavior here... Some explanations (in the commit message at least) would help following what's going on. | |
Thanks.
I'm slightly surprised by the change of signature for translate_directory_intrinsic_metadata without change to the docstring. AFAICT this function now does the detection itself, so maybe it deserves an update.
| swh/indexer/metadata.py | ||
|---|---|---|
| 205 | by "outright broken", I mean that when leaving the removed branch, dir_ was a dict containing an entries key, instead of being the list itself, as expected by the rest of the code (entry["type"] for entry in dir_) | |
Build is green
Patch application report for D8045 (id=29113)
Rebasing onto 1be4e184d4...
Current branch diff-target is up to date.
Changes applied before test
commit dc035a669329272713b0b977ba2f685501e80ba0
Author: Valentin Lorentz <vlorentz@softwareheritage.org>
Date: Wed Jun 29 11:01:35 2022 +0200
Add typing to detect_metadata() and related functions
This also moves the call to `detect_metadata()` to
`translate_directory_intrinsic_metadata` so type annotations make more
sense; and remove a dead/broken code branch in
`DirectoryMetadataIndexer.index()` that was detected by mypy.See https://jenkins.softwareheritage.org/job/DCIDX/job/tests-on-diff/308/ for more details.
Build is green
Patch application report for D8045 (id=29116)
Rebasing onto 1be4e184d4...
Current branch diff-target is up to date.
Changes applied before test
commit 8c5a84459d4436af7b4b751e48091a99bce0a7d1
Author: Valentin Lorentz <vlorentz@softwareheritage.org>
Date: Wed Jun 29 11:01:35 2022 +0200
Add typing to detect_metadata() and related functions
This also moves the call to `detect_metadata()` to
`translate_directory_intrinsic_metadata` so type annotations make more
sense; and remove a dead/broken code branch in
`DirectoryMetadataIndexer.index()` that was detected by mypy.See https://jenkins.softwareheritage.org/job/DCIDX/job/tests-on-diff/311/ for more details.