Summary: Fixes T2543. This mode has been a stable prototype for a very long time now; promote it so "--draft" can promote out of "experimental" in Arcanist.
Test Plan: See T2543 for discussion.
Maniphest Tasks: T2543
Differential Revision: https://secure.phabricator.com/D20983
Summary: Ref T13395. Move cache classes, syntax highlighters, other markup classes, and sprite sheets to Phabricator.
Test Plan: Attempted to find any callers for any of this stuff in libphutil or Arcanist and couldn't.
Maniphest Tasks: T13395
Differential Revision: https://secure.phabricator.com/D20977
Summary:
Fixes T13468. See that task for discussion. The older source-rendering code mixes "line number" / "1-based" lists with "block number" / "0-based" lists and then has other bugs which cancel this out.
For block-based diffs, build an explicit block-based mask with only block numbers. This sort of sidesteps the whole issue.
Test Plan: Viewed the diff with the original reproduction case, plus various other block-based diffs, including one-block image diffs, in unified and side-by-side mode. Didn't spot any oddities.
Maniphest Tasks: T13468
Differential Revision: https://secure.phabricator.com/D20959
Summary: Ref T13480. Add an "Author's packages" field to Herald to support writing rules like "if affected packages include X, and author's packages do not include X, raise the alarm".
Test Plan: Wrote and executed rules with the field, saw a sensible field value in the transcript.
Maniphest Tasks: T13480
Differential Revision: https://secure.phabricator.com/D20947
Summary:
Fixes T13443. When a panel raises an exception during edit action generation, it currently escapes to top level. Instead, catch it more narrowly.
Additionally, mark "ChangesetSearchEngine" as not being a suitable search engine for use in query panels. There's no list view or search URI so it can't generate a sensible panel.
Test Plan:
- Added a changeset panel to a dashboard.
- Before: entire dashboard fataled.
- After: panel fataled narrowly, menu fatals narrowly, dashboard no longer permits creation of another Changeset query panel.
Maniphest Tasks: T13443
Differential Revision: https://secure.phabricator.com/D20902
Summary: Ref T13425. When a change adds or removes a block (vs adding or removing a document, or changing a block), we currently try to render `null` as cell content in the unified view. Make this check broader to catch both "no opposite document" and "no opposite cell".
Test Plan: {F7008772}
Subscribers: artms
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20897
Summary: Ref T13440. This feature is used in only one interface which I'm about to rewrite, so throw it away.
Test Plan: Grepped for all affected symbols, didn't find any hits anywhere.
Maniphest Tasks: T13440
Differential Revision: https://secure.phabricator.com/D20882
Summary:
Ref T13425. The changes in D20865 could incorrectly lead to selection of a DocumentEngine that can not generate document diffs if a file was added or removed (for example, when a source file is added).
Move the engine pruning code to be shared -- we should always discard engines which can't generate a diff, even if we don't have both documents.
Test Plan: Viewed an added source file, no more document ref error arising from document engine selection.
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20866
Summary: Ref T13425. When a file (like a Jupyter notebook) is added or removed, we can still render a useful line-by-line diff.
Test Plan:
- Viewed add/modify/remove of Jupyter, source code, and images in 2up/1up mode, everything looked okay.
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20865
Summary:
See PHI1468. Engine selection for diffs is currently too aggressive in trying to find a shared engine and will fall back a shared engine with a very low score, causing all ".json" files to render as Jupyter files.
Only pick an engine as a difference engine by default if it's the highest-scoring engine for the new file.
Test Plan: Viewed ".json" files and ".ipynb" files in a revision. Before, both rendered as Jupyter. Now, the former rendered as JSON and the latter rendered as Jupyter.
Differential Revision: https://secure.phabricator.com/D20850
Summary:
Depends on D20844. Ref T13425. When we line up two blocks and they can be interdiffed (generally: they both have the same type of content), let the Engine interdiff them.
Then, make the Jupyter engine interdiff markdown.
Test Plan: {F6898583}
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20845
Summary:
Depends on D20843. Ref T13425. Add very basic support for "Show Hidden Context", in the form of folding it behind an unclickable shield. This isn't ideal, but should be better than nothing.
Prepare for "intraline" diffs on content blocks.
Fix newline handling in Markdown sections in Jupyter notebooks.
Remove the word "visibile" from the codebase.
Test Plan: {F6898192}
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20844
Summary:
Ref T13425. Some diff checks currently sequence incorrectly:
- When we're rendering block lists, syntax highlighting isn't relevant.
- The "large change" guard can prevent rendering of otherwise-renderable changes.
- Actual errors in the document engine (like bad JSON in a ".ipynb" file) aren't surfaced properly.
Improve sequencing somewhat to resolve these issues.
Test Plan:
- Viewed a notebook, no longer saw a "highlighting disabled" warning.
- Forced a notebook to fail, got a useful inline error instead of a popup dialog error.
- Forced a notebook to have a large number of differences, got a rendering out of it.
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20843
Summary:
Ref T13425. Currently, code inputs and all outputs are grouped into a single block. This is fine for display notebooks but not great for diffing notebooks.
Instead, split source code input into individual lines with one line per block, and each output into its own block.
This allows you to leave actual line-by-line inlines on source, and comment on outputs individually.
Test Plan: {F6888583}
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20840
Summary:
Depends on D20835. Ref T13425. Ref T13414. When a document has a list of content blocks, we may not be able to diff it directly, but we can hash each block and then diff the hashes (internally "diff" also does approximately the same thing).
We could do this ourselves with slightly fewer layers of indirection, but: diff already exists; we already use it; we already have a bunch of abstractions on top of it; and it's likely much faster on large inputs than the best we can do in PHP.
Test Plan: {F6888169}
Maniphest Tasks: T13425, T13414
Differential Revision: https://secure.phabricator.com/D20836
Summary:
Depends on D20834. Ref T13425. After the change from "th" to "td" for accessibility, the algorithm picks which cells it should highlight slightly improperly (it picks too many cells since it can no longer find the line numbers).
Ideally, it would probably highlight //only// the source content, but there isn't an easy way to do this right now. Settle for an incremental improvement for the moment.
Test Plan: Hovered over line numbers, saw a more accurate highlight area.
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20835
Summary: Depends on D20833. Ref T13425. This look like it "just works"?
Test Plan: Left inline comments on a Juptyer notebook. Nothing seemed broken? Confusing.
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20834
Summary:
Depends on D20832. Ref T13425. Emit Jupyter notebooks as diffable blocks with block keys.
No diffing or proper inlines yet.
Test Plan: {F6888058}
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20833
Summary:
Depends on D20831. Ref T13425. As an escape hatch to get out of future DocumentEngine rendering behavior, provide a "View As.." option.
Now I can break DocumentEngine real bad and no one can complain.
Test Plan: Used "View As" to swap document engines for image files.
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20832
Summary:
Depends on D20830. Ref T13425. Have the image engine elect into block rendering, then emit blocks.
This is rough (the blocks aren't actually diffed yet) but image diffs were already pretty rough so this is approximately a net improvement.
Test Plan: Viewed image diffs, saw nothing worse than before.
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20831
Summary:
Ref T13425. Allow DocumentEngines to claim they can produce diffs. If both sides of a change can be diffed by the same document engine and it can produce diffs, have it diff them.
This has no impact on runtime behavior because no upstream engine elects into diff generation yet.
Test Plan: Loaded some revisions, nothing broke.
Maniphest Tasks: T13425
Differential Revision: https://secure.phabricator.com/D20830
Summary:
Fixes T13386. See PHI1391. These constraints largely exist already, but are not yet exposed to Conduit.
Also, tweak some keys to support the underlying query.
Test Plan: Ran `differential.revision.search` queries with the new constraints.
Maniphest Tasks: T13386
Differential Revision: https://secure.phabricator.com/D20730
Summary:
Ref T13385. Currently, if you run `arc diff` in a CWD with more than 255 characters, the workflow fatals against the length of the `sourcePath` database column.
In the long term, removing this property is likely desirable.
For now, truncate long values and continue. This only meaningfully impacts relatively obscure interactive SVN workflows negatively, and even there, "some arc commands are glitchy in very long working directories in SVN" is still better than "arc diff fatals".
Test Plan:
- Modified `arc` to submit very long source paths.
- Ran `arc diff`.
- Before: Fatal when inserting >255 characters into `sourcePath`.
- After: Path truncated at 255 bytes.
Maniphest Tasks: T13385
Differential Revision: https://secure.phabricator.com/D20727
Summary:
See <https://discourse.phabricator-community.org/t/view-task-from-maniphest-e-mail-doesnt-have-url/2827>.
I added "View Task" / "View Commit" buttons recently but the logic for generating URIs isn't quite right. Fix it up.
Test Plan:
- Commented on a task.
- Used `bin/mail show-outbound --id ... --dump-html > out.html` to dump the HTML.
- Previewed the HTML in a browser.
- This time, actually clicked the button to go to the task.
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20586
Summary:
Fixes T13313. The "Download Raw Diff" workflow in Differential currently uses an older way of interacting with Files that doesn't engage the chunk engine and can't handle 8MB+ files.
Update to `IteratorFileUploadSource` -- we're still passing in a single giant blob, but this approach can be chunked.
This will still break somewhere north of 8MB (it will break at 2GB with the PHP string limit if nowhere sooner, since we're putting the entire raw diff in `$raw_diff` rather than using a rope/stream) but will likely survive diffs in the hundreds-of-megabytes range for now.
Test Plan:
- Added `str_repeat('x', 1024 * 1024 * 9)` to the `$raw_diff` to create a 9MB+ diff.
- Configured file storage with no engine explicitly configured for >8MB chunks (i.e., "reasonably").
- Clicked "Download Raw Diff".
- Before: misleading file storage engine error ("no engine can store this file").
- After: large, raw diff response.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13313
Differential Revision: https://secure.phabricator.com/D20579
Summary: Ref T13303. See B22967. This should be "msortv()" but didn't get updated properly.
Test Plan: The system works!
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13303
Differential Revision: https://secure.phabricator.com/D20585
Summary:
Fixes T13300. Currently, if you create a revision and then immediately land it (either using `--draft` or just beating Harbormaster to the punch) it can be stuck in "Draft" forever.
Instead, count landing changes like this as a publishing action.
Test Plan:
- Used `arc diff --hold` to create a revision, then pushed the commit immediately.
- Before change: revision closed, but was stuck in draft.
- After change: revision closed and was promoted out of draft.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13300
Differential Revision: https://secure.phabricator.com/D20565
Summary:
See downstream <https://phabricator.wikimedia.org/T1050>. Some time ago, we added a "View Revision" button to Differential mail. This hasn't created any problems and generally seems good / desirable.
It isn't trivial to just add everywhere since we need a translation string in each case, but at least add it to Maniphest for now. Going forward, we can fill in more applications as they come up.
Test Plan:
Used `bin/mail show-outbound --id <x> --dump-html`:
{F6470461}
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20561
Summary:
Ref T13290. Prior to recent changes, if we parsed some commit C which was associated with a revision R, but R was already closed, we'd skip the whole set of updates because the "close the revision" transaction would fail and we'd throw because we did not `setContinueOnNoEffect()`.
We now continue on no effect so we can get the edge ("commit has revision" / "revision has commit"), since we want it in all cases, but this means we may also apply an extra "Updated revision to reflect committed changes" transaction and new diff. This can happen even if we're careful about not trying to apply this transaction to closed revisions, since two workers may race. (Today, we aren't too careful about this.)
To fix this, just make this transaction no-op itself if the revision is already closed by the time it tries to apply.
This happened on D20451 because a merge commit with the same hash as the last diff was pushed, but it's easiest to reproduce by just running `bin/repository reparse --message <commit>`, which updates related revisions with a new diff every time.
Test Plan:
- Ran `bin/repository reparse --messsage <commit>` several times, on a commit with an associated revision.
- Before: each run attached a new diff and created a new "updated to reflect committed changes" transaction.
- After: repeated runs had no effects.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13290
Differential Revision: https://secure.phabricator.com/D20545
Summary:
Ref T13290. Ref T13291. Now that a full URI is a "mention", the full URI in "Differential Revision: ..." also triggers a mention.
Stop it from doing that, since these mentions are silly/redundant/unintended.
The API here is also slightly odd; simplify it a little bit to get rid of doing "append" with "get + append + set".
Test Plan: Used `bin/repository reparse --publish` to republish commits with "Differential Revision: ..." and verified that the revision PHID was properly dropped from the mention list.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13291, T13290
Differential Revision: https://secure.phabricator.com/D20544
Summary:
See PHI1222. When we publish several transactions to feed at once, we sort them by "action strength" to figure out which one gets to be the title story.
This sort currently uses `msort()`, which uses `asort()`, which is not a stable sort and has inconsistent behavior across PHP versions:
{F6463721}
Switch to `msortv()`, which is a stable sort. Previously, see also T6861.
If all transactions have the same strength, we'll now consistently pick the first one.
This probably (?) does not impact anything in the upstream, but is good from a consistency point of view.
Test Plan:
Top story was published after this change and uses the chronologically first transaction as the title story.
Bottom story was published before this change and uses the chronologically second transaction as the title story.
Both stories have two transactions with the same strength ("create" + "add reviewer").
{F6463722}
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20540
Summary:
Ref T13289. In Maniphest, you can currently search for "Owner: none()" to find tasks with no owner, but there's no way to search for "Reviewers: none()" in Differential right now.
Add support for this, since it's consistent and reasonable and doesn't seem too weird or niche.
Test Plan: Searched for "Reviewers: none()", found revisions with no reviewers. Searched for "Reviewers: alice, none()", "Reviewers: alice", and "Reviewers: <no constraint>" and got sensible results.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13289
Differential Revision: https://secure.phabricator.com/D20537
Summary:
See PHI1210. For certain large inputs, we spend more time than we need to replacing tabs with spaces. Add some fast paths:
- When a line only has tabs at the beginning of the line, we don't need to do as much work parsing the rest of the line.
- When a line has no unicode characters, we don't need to vectorize it to get the right result.
Test Plan:
- Added test coverage.
- Profiled this, got a ~60x performance increase on a 36,000 line 3MB text file.
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20477
Summary:
See PHI1232, which describes a reasonable use case for wanting information about the "draft" ("Hold as Draft / Do Not Auto-Promote") flag.
Also, flesh out "testPlan" and "summary". It's possible these "blob of remarkup" fields might have metadata some day (e.g., a rendered version or a list of PHIDs or something), but we could add more keys, and we already have some other transactions which work like this.
Test Plan: Used "transaction.search" to fetch these transaction types, saw type information and metadata.
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20507
Summary:
Depends on D20469. Ref T13276. See PHI1159. See PHI953. See PHI901.
Allow Herald to detect when "arc land" would (or did) warn users about failed or ongoing builds. This respects the "Warn on Landing" build plan behavior.
To accomplish this:
- When we close a revision, set a "wrong build state" flag if it lands in the wrong build state.
- If the revision is closed when we hit Herald, look for the flag.
- If not (common for push rules, can happen for commit rules if we race against the revision update worker), hit Harbormaster ourselves and check the current state.
Test Plan:
- Wrote a "Require Green" rule.
- Ran it against various commits with various build states (good, not good).
- Fiddled with "Warn on Landing" and saw the effect in rule evaluation.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13276
Differential Revision: https://secure.phabricator.com/D20470
Summary:
Depends on D20468. Ref T13276. See PHI1008.
When a commit or revision "reverts <hash>", and that's the hash of a commit which has a revision, also write a "reverts" edge to the revision.
Test Plan:
Created "reverts X" objects for:
- a revision;
- a commit;
- a commit with a revision (both got marked properly).
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13276
Differential Revision: https://secure.phabricator.com/D20469
Summary:
See PHI1218. When rendering "A vs B", we currently show the properties of diff A without modification.
Instead, take properties from the same place we're taking change details.
See T12664 for a followup.
Test Plan:
- In diff A, removed "+x" from a file.
- In diff B, changed the file but did not remove "+x" from it.
- Diffed B vs A.
- Before change: UI incorrectly shows "+x" removed (both sides incorrect, just showing the change from diff A).
- After change: UI shows 100644 -> null, which is half right.
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20478
Summary:
Depends on D20462. Ref T13276. Currently, the "Message" parser also updates related tasks and revisions when a commit is published.
For PHI1165, which ran into a race with message parsing, I originally believed we needed to separate this logic and lock + yield to avoid the race. D20462 provides what is probably a better approach for avoiding the race.
Still, I think separating these "update related revisions" and "updated related tasks" chunks into separate workers is a net improvement. There may still be some value in doing lock + yield in the future to deal with other issues, and when we occasionally run into problems with pulling a diff out of the repository to update the revision (usually because the diff is too big) this isolates the problem better and allows the commit to import.
I think the only thing to watch out for here is that Herald may now run before the revision and commit are attached to one another. This is fine for all current Herald rules, we just need to be mindful in implementing new rules.
Test Plan: Used `bin/repository reparse --message` on various commits, including commits that close revisions and close tasks.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13276
Differential Revision: https://secure.phabricator.com/D20463
Summary: Depends on D20459. Ref T13276. I'll file a followup to actually destroy the table.
Test Plan:
- Grepped for `TABLE_COMMIT`.
- Ran `bin/storage upgrade -f`, got a clean bill of health.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13276
Differential Revision: https://secure.phabricator.com/D20461
Summary:
Depends on D20458. Ref T13276. Although I'm not thrilled about "needCommitPHIDs()", it has a few callers, including custom fields. Allow "need + attach + get" to survive for now since they're reasonably modern, at least.
However, use edges instead of "TABLE_COMMIT" and require `need...()` + `get...()`, removing the direct `load...()`.
Also remove `RevisionQuery->withCommitPHIDs(...)`, which has no callers.
Test Plan:
- Grepped for `loadCommitPHIDs` (only two hits, the private `RevisionQuery` method).
- Called "differential.getrevision", got commits.
- Viewed a revision, saw "Commits: ...".
- Grepped for `withCommitPHIDs()`, no callers on `RevisionQuery` (some other query classes have methods with this name).
- Called "differential.query", got commits.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13276
Differential Revision: https://secure.phabricator.com/D20459
Summary: Depends on D20457. Ref T13276. Kill all remaining callers to this method and delete it.
Test Plan:
- Grepped for `loadIDsByCommitPHIDs`.
- Viewed blame again to make sure I didn't break it.
- Viewed "History" view for commits with revisions.
- Viewed "Graph" view for commits with revisions.
- Viewed "Merged Commits" table for commits with revisions.
- Viewed "Compare" table for commits with revisions.
- Viewed "Repository" main page history table for commits with revisions.
- Grepped for `linkRevision`.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13276
Differential Revision: https://secure.phabricator.com/D20458
Summary:
See PHI1213. If we don't have identities for "revision X closed by commit Y" stories, just do the plain non-attribution rendering rather than trying to fall back. Falling back won't work since we don't load the data, which should be OK now since identities seem like they're in generally good shape.
(We could probably just throw out the fallback behavior instead, but we can always clean things up later.)
Test Plan: Forced no commit identity on a revision, loaded it, saw reasonable story rendering.
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20460
Summary:
Ref T12164. Ref T13276. Currently, the parsing pipeline copies the author and committer names and PHIDs into the transcaction record as metadata. They are then rendered directly from the metadata.
This makes planned changes to the parsing pipeline (to prevent races when multiple commits matching a single revision are pushed simultaneously) more difficult, and generally won't work with repository identities.
Instead, load the commit and use its identities.
Test Plan: Loaded a revision, saw the same story rendering for a "Closed by commit" story.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13276, T12164
Differential Revision: https://secure.phabricator.com/D20418
Summary:
Depends on D20412. See PHI1147.
- Index the targets of "Add Reviewer", "Add Blocking Reviewer", "Add Auditor", "Add Subscriber", and "Remove Subscriber" Herald rules. My major goal is to get Owners packages. This will also hit projects/users, but we just don't read those edges (for now, at least).
- Add a "Related Herald Rules" panel to Owners Package pages.
- Add a migration to reindex Herald rules for the recent build plan stuff and this, now that such a migration is easy to write.
Test Plan:
Ran migration, verified all rules reindexed.
{F6372695}
Reviewers: amckinley
Reviewed By: amckinley
Subscribers: PHID-OPKG-gm6ozazyms6q6i22gyam
Differential Revision: https://secure.phabricator.com/D20413
Summary:
Ref T13276. This was introduced in D2586 to power a "trigger audits when the committed change does not match the reviewed change" feature. It was removed without ceremony in D15939. Broadly, rebases mean that this sort of feature can't really work like this and this approach is inherently unreliable; see also T182.
This property no longer has readers, and is unlikely to get any in the future since my planned pathway for "committed code must match reviewed code, modulo an automated rebase" is automating the rebase via "Land Revision", not comparing the diff text.
Remove this to simplify the flow of data here so that things in T13276 can be fixed more easily.
Test Plan: Grepped for `vsDiff`, no hits.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13276
Differential Revision: https://secure.phabricator.com/D20395
Summary:
See PHI1175. An install would like to trigger some reminders/guidance if users don't link revisions to JIRA issues.
Expose "JIRA Issue URIs" as a field so Herald can act on the presence or absence of issues.
I'm exposing "JIRA Issue URIs", not a field like "[ Has Jira Issue ][ is true ]", since it's a bit more flexible: you can use a regexp to test against particular `PROJ-123` project prefixes in JIRA, for example.
Test Plan: {F6367696}
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20386
