Page MenuHomePhabricator

Give "owners.search" a "paths" attachment and a default "owners" value
ClosedPublic

Authored by epriestley on Dec 14 2015, 2:25 PM.
Tags
None
Referenced Files
F14352861: D14775.diff
Thu, Dec 19, 3:01 PM
Unknown Object (File)
Thu, Dec 5, 10:50 AM
Unknown Object (File)
Sun, Dec 1, 12:11 PM
Unknown Object (File)
Sun, Dec 1, 7:11 AM
Unknown Object (File)
Wed, Nov 27, 2:03 AM
Unknown Object (File)
Sat, Nov 23, 2:44 AM
Unknown Object (File)
Fri, Nov 22, 9:01 AM
Unknown Object (File)
Nov 19 2024, 8:48 AM
Subscribers
None

Details

Summary

Ref T9964.

  • Add a "paths" attachment for fetching paths.
  • Always load owners. We will need this to do policy checks in the future, anyway, and this data is not large, is very useful, and is reasonable to load unconditionally.
Test Plan
  • Queried packages via API.
  • Edited packages (paths, owners).
  • Created a package.

Diff Detail

Repository
rP Phabricator
Branch
attach4
Lint
Lint Passed
Unit
Tests Passed
Build Status
Buildable 9548
Build 11407: Run Core Tests
Build 11406: arc lint + arc unit

Event Timeline

epriestley retitled this revision from to Give "owners.search" a "paths" attachment and a default "owners" value.
epriestley updated this object.
epriestley edited the test plan for this revision. (Show Details)
epriestley added a reviewer: chad.
chad edited edge metadata.
This revision is now accepted and ready to land.Dec 14 2015, 4:16 PM
src/applications/paste/engineextension/PhabricatorPasteContentSearchEngineAttachment.php
20

not related?

src/applications/paste/engineextension/PhabricatorPasteContentSearchEngineAttachment.php
20

Sort of, this is adjusting a thing I added in the previous diff, and it's vaaaaaguely related because it's making it consistent with what I did here.

Originally I called the content data, because having two nested content keys seemed a little weird, but then when I got to owners I decided it was fine and better than having every attachment have a generic data key so this is just "I'm going to use descriptive keys instead of generic ones, even if that sometimes means you access result.data[3].attachments.paths.paths[3] in callers, since callers need to be managing that complexity in a reasonable way anyway and this is probably more clear in the long run, and that's not really worse than result.data[3].attachments.paths.data[3]".

This revision was automatically updated to reflect the committed changes.