Summary:
Add the differential parse cache to the GC. This is the largest object in the
system by a wide margin, I think.
This table is potentially gigantic which is why the script truncates it before
doing a schema change.
Test Plan: Ran the GC daemon, it cleaned up some parse caches.
Reviewed By: jungejason
Reviewers: tuomaspelkonen, jungejason, aran
Commenters: tuomaspelkonen
CC: aran, jungejason, tuomaspelkonen, epriestley
Differential Revision: 620
Summary:
See T303. Enable comment panel haunting.
I hid the preview for the sticky panel, which I think is reasonable?
Test Plan:
https://secure.phabricator.com/file/view/PHID-FILE-64713fa8a7c2a22e5b93/
Reviewed By: broofa
Reviewers: broofa, jungejason, aran, tomo, tuomaspelkonen
CC: aran, broofa
Differential Revision: 615
Summary:
My earlier diff refactored some code without completely
respecting the semantics, sometimes resulting in duplicate field names
returned from differential.getcommitmessage. This fixes that.
Test Plan:
ran "arc diff" with diff causing the bug (commit message
had an empty Revert Plan: field) and verified no duplicate fields
Reviewed By: epriestley
Reviewers: epriestley
CC: aran, dpepper, epriestley
Differential Revision: 610
Summary: Reduce the amount of code duplication here and allow for an override
configuration on the filename.map stuff.
Test Plan: Checked paste, diffusion and differential syntax highlighting and
everything appeared reasonable.
Reviewed By: codeblock
Reviewers: tuomaspelkonen, codeblock, jungejason, aran
CC: aran, codeblock, epriestley
Differential Revision: 601
Summary: I'll clean some of this stuff up in a followup too, but update the
callers to use the new explicit filename-based API.
Test Plan: Looked at paste, Diffusion and Differential.
Reviewed By: tuomaspelkonen
Reviewers: tuomaspelkonen, codeblock, jungejason, aran
CC: aran, tuomaspelkonen
Differential Revision: 600
Summary: These came up in dealing with the diff produced by T271. When a file is
unmodified, don't try to use the "ignore all whitespace" algorithm on it. Also,
detect "changed only by adding or removing trailing whitespace" vs "this file
was not modified" correctly.
Test Plan: Viewed the diff that came out of running 'arc diff' on my T271 mess.
Reviewed By: tuomaspelkonen
Reviewers: alex, jungejason, tuomaspelkonen, aran
CC: aran, tuomaspelkonen
Differential Revision: 548
Summary:
- Add a default list of supported languages to default.conf.php
and make the initial/default value customizable.
- Store a '' in the database to infer the language from the filename/title.
Test Plan:
Tested in my sandbox with pygments enabled and disabled and various
combinations of filename/extension/dropdown selection.
Reviewers:
epriestley
CC:
Differential Revision: 587
Summary:
We already support this (and Facebook uses it) but it is difficult to configure
and you have to write a bunch of code. Instead, provide a simple flag.
See the documentation changes for details, but when this flag is enabled we send
one email with a reply-to like "D2+public+23hf91fh19fh@phabricator.example.com".
Anyone can reply to this, and we figure out who they are based on their "From"
address instead of a unique hash. This is less secure, but a reasonable tradeoff
in many cases.
This also has the advantage over a naive implementation of at least doing object
hash validation.
@jungejason: I don't think this affects Facebook's implementation but this is an
area where we've had problems in the past, so watch out for it when you deploy.
Also note that you must set "metamta.public-replies" to true since Maniphest now
looks for that key specifically before going into public reply mode; it no
longer just tests for a public reply address being generateable (since it can
always generate one now).
Test Plan:
Swapped my local install in and out of public reply mode and commented on
objects. Got expected email behavior. Replied to public and private email
addresses.
Attacked public addresses by using them when the install was configured to
disallow them and by altering the hash and the from address. All this stuff was
rejected.
Reviewed By: jungejason
Reviewers: moskov, jungejason, tuomaspelkonen, aran
CC: aran, epriestley, moskov, jungejason
Differential Revision: 563
Summary: This used to be in the subject but there was a bunch of churn and now
it's nowhere.
Test Plan: Created, updated, and added CCs to a diff.
Reviewed By: moskov
Reviewers: moskov, avitaloliver, jungejason, tuomaspelkonen, aran
CC: aran, moskov
Differential Revision: 567
Summary: The field hints on this interface don't behave correctly. Particularly, when you add yourself as a reviewer you aren't pointed at the issue.
Test Plan: Edited a revision and tried to save invalid changes, including self-reviewership.
Reviewers: moskov, jungejason, tuomaspelkonen, aran
CC:
Differential Revision: 565
explicitly
Summary: You currently have to click through to figure out who got added.
Test Plan:
- Made a comment which added CCs.
- Made a comment which added reviewers.
- Made a comment which added nothing.
- Made a comment which added CCs and reviewers.
Reviewed By: tomo
Reviewers: tomo, jungejason, tuomaspelkonen, aran
CC: aran, tomo
Differential Revision: 562
Summary:
when "arc diff" generates a revision, it attaches a task id
if one is included. However, "arc amend" did not return a task id,
effectively stripping it from the commit message. This diff fixes
that.
NOTE: This is dependent on revision 549 https://secure.phabricator.com/D549
Test Plan:
0. created a custom class to append Facebook task IDs to commit messages and
attached it to the differential.append-commit-message-class config variable
1. created a new diff in the www repot
2. included Task ID: 609350 in the git commit message
3. "arc diff" to generate the revision
4. "arc amend"
5. ensure that the "Task ID:" field remained in the git commit message
Reviewed By: epriestley
Reviewers: dpepper, jungejason, epriestley
CC: aran, epriestley, mgummelt
Differential Revision: 546
Summary:
commit message fields were previously stored as name/value
pairs in an associative array. this resulted in ad hoc code to modify
the structure/rendering of these fields in commit messages. this diff
introduces a new DifferentialCommitMessageField class.
Test Plan:
ran "arc amend" to ensure the commit message still looked good
Reviewed By: epriestley
Reviewers: epriestley
CC: aran, dpepper, epriestley
Differential Revision: 554
Summary:
When a user gets @mentioned in Differential, add them as a CC.
No Maniphest hookup yet since I want to make that one a little more formal.
Depends on D518.
Test Plan:
@mentioned a user and they were added as a CC.
Reviewed By: jungejason
Reviewers: tomo, mroch, jsp, jungejason, aran, tuomaspelkonen
CC: aran, jungejason
Differential Revision: 519
Summary:
We currently have only an "Add reviewers" action, add "Add CCs". This can also
be accomplished less-discoverably with mentions.
Test Plan:
Added reviewers and CCs to revisions. Toggled display between reviewers and CCs.
Reviewed By: jungejason
Reviewers: tomo, mroch, jsp, jungejason, aran, tuomaspelkonen
CC: aran, jungejason
Differential Revision: 521
Summary:Make a new directory, src/infrastructure/markup/remarkup/markuprule/paste/
Make a new class called PhabricatorRemarkupRulePaste in that directory.
Add the rule to DifferentialMarkupEngineFactory.
Test Plan: Created a task in maniphest. Put P1 and P2 in the content.
Created P1 and P2 in Paste. Verified P1 and P2 were highlighted and
linked correctly.
Reviewers:epriestley, codeblock
CC:jungejason
Differential Revision: 539
Summary:
Although I think the recent changes here improved things, the "Unsaved Draft"
language is continuing to confuse new users. Try to find some less-confusing
langauge. Open to suggestions here, too.
Test Plan:
Viewed unsubmitted inline comments.
Reviewed By: jungejason
Reviewers: aran, jungejason, gregprice
CC: aran, epriestley, jungejason
Differential Revision: 501
Summary:
This depends on D513 and D514. Those diffs make the display algorithms safe with
respect to mutating utf8, so we no longer need to repair potentially invalid
utf8 sequences with this hack.
Test Plan:
grepped for calls to this method
Reviewers: jungejason, aran, tuomaspelkonen
Commenters: jungejason
CC: aran, epriestley, jungejason
Differential Revision: 515
Summary:
Differential uses a byte-oriented linewrap algorithm. Instead, use a
character-oriented one which will handle utf-8 properly.
This implies a very slightly performance hit but we only run this code for lines
which need to wrap, and the results get cached. It took about ~2.5ms for the
test file on my machine. I'll keep an eye on it but I think it's currently a
manageable cost.
Test Plan:
Diffed this file: https://secure.phabricator.com/P43
...and got it to render like this:
https://secure.phabricator.com/file/info/PHID-FILE-331ac241bede705b193b/
To do so, I had to disable the un-utf8 block which we can't actually do yet
because of intraline diff, but it shows that once we can get rid of that it
works completely correctly. It will "sort of" work in the meantime (nothing
terrible happens).
Reviewers: jungejason, aran, tuomaspelkonen
CC: aran, epriestley
Differential Revision: 513
Summary:
Replace some more date() calls with locale-aware calls.
Also, at least on my system, the DateTimeZone / DateTime stuff didn't actually
work and always rendered in UTC. Fixed that.
Test Plan:
Viewed daemon console, differential revisions, files, and maniphest timestamps
in multiple timezones.
Reviewed By: toulouse
Reviewers: toulouse, fratrik, jungejason, aran, tuomaspelkonen
CC: aran, toulouse
Differential Revision: 530
Summary:
Provides basic Remarkup support for @mentions. No application integration yet so
these aren't terribly useful until that happens.
Test Plan:
https://secure.phabricator.com/file/view/PHID-FILE-83d68e7af6085ae928df/
Reviewers: tomo, mroch, jsp
Commenters: tomo
CC: aran, tomo, epriestley
Differential Revision: 517
Summary: This depends on D513 and D514. Those diffs make the display algorithms safe with respect to mutating utf8, so we no longer need to repair potentially invalidate utf8 sequences with this hack.
Test Plan: grepped for calls to this method
Reviewers: jungejason, aran, tuomaspelkonen
CC:
Differential Revision: 515
Summary: Differential uses a byte-oriented linewrap algorithm. Instead, use a character-oriented one which will handle utf-8 properly.
This implies a very slightly performance hit but we only run this code for lines which need to wrap, and the results get cached. It took about ~2.5ms for the test file on my machine. I'll keep an eye on it but I think it's currently a manageable cost.
Test Plan: Diffed this file: https://secure.phabricator.com/P43
...and got it to render like this: https://secure.phabricator.com/file/info/PHID-FILE-331ac241bede705b193b/
To do so, I had to disable the un-utf8 block which we can't actually do yet because of intraline diff, but it shows that once we can get rid of that it works completely correctly. It will "sort of" work in the meantime (nothing terrible happens).
Reviewers: jungejason, aran, tuomaspelkonen
CC:
Differential Revision: 513
Summary:
Diffs with missing context don't render properly in the "ignore all whitespace"
algorith, so don't try to use it. These diffs can occur if someone creates a
diff via the web interface, for example, or if they muck around in their copy of
'arc'.
See D473, T246 (a problem with D473), rPe5bb756b5191720 (revert of D473) and
T231.
Test Plan:
Viewed a diff with missing context from the web interface. Verified normal diffs
still rendered with all whitespace ignored.
Reviewed By: fratrik
Reviewers: jungejason, aran, tuomaspelkonen, fratrik
Commenters: jungejason
CC: aran, epriestley, fratrik, jungejason
Differential Revision: 500
Summary:
See comments. I think this will fix the issue, where we end up handling off
garbage to htmlspecialchars() after highlighting a file we've stuck full of \0
bytes.
The right fix for this is to make wordwrap and intraline-diff utf8 aware and
throw this whole thing away. I'll work on that but I think this fixes the
immediate issue.
Test Plan:
diffed the file with a UTF-8 quote in it and got a reasonable render in
Differential
Reviewed By: jungejason
Reviewers: jungejason, aran, tuomaspelkonen
CC: aran, jungejason
Differential Revision: 504
Summary:
See T251. In Gmail, conversations split if you reply to them and the next email
does not "In-Reply-To" your message ID. When an action is triggered by an email,
carry its Message-ID through the stack and use it for "In-Reply-To" and
"References" on the subsequent message.
Test Plan:
Live-patched phabricator.com and replied to a Maniphest thread in Gmail without
disrupting the thread. Locally replied to Maniphest and Differential threads and
verified Message-ID was carried across the reply boundary.
Reviewed By: rm
Reviewers: tcook, jungejason, aran, tuomaspelkonen, rm
CC: aran, epriestley, rm
Differential Revision: 498
Summary:
If you "chmod +x" a file and then generate a diff and click "show 20 lines" on
that diff, you get another janky copy of the property table. Render the property
table only for top-level rendering requests.
This started happening after D409, which fixed the far-more-obvious bug of these
things never showing up. We must have changed the logic at some point since this
side effect was surprising to me. :P
Test Plan:
Created a diff with changes and +x, clicked "show 20 lines".
- Original diff had file property header table thing, showing the +x.
- New context brought in by "show 20 lines" didn't have it anymore.
Reviewed By: jungejason
Reviewers: aran, jungejason, tuomaspelkonen
CC: aran, epriestley, jungejason
Differential Revision: 482
Summary:
create the event
Test Plan:
checked the timeline event was consumed successfully by
facebook daemon when the revision is created from the webUI or arc
command line.
Reviewed By: epriestley
Reviewers: epriestley, aran
CC: aran, epriestley
Differential Revision: 478
Summary:
The owners of a revision are only really the reviewers when the revision is in
NEEDS_REVIEW.
Also build a raw indexed document viewer so you can look at the index of a
document from the web interface.
Finally, reindex revisions when comments are added, not just when the revision
itself is edited.
Test Plan:
Toggled abandon/reclaim on a revision and verified the relationships indexed
properly.
Reviewed By: jungejason
Reviewers: aran, jungejason, tuomaspelkonen
CC: aran, epriestley, jungejason
Differential Revision: 470
Summary:
'Create Diff' with whitespace mode 'ignore-all' is borken, because the
line numbers get mixed up when creating a second diff for the
whitespace changes.
This should be fixed correctly at some point, but currently the
whitespace 'ignore-all' says 'Huge mess' in the comments and I didn't
want to make the mess any bigger.
Test Plan:
Tested that 'Create Diff' showed the diff correctly and a previously
created diff looked correct once the cache was disabled.
Reviewed By: epriestley
Reviewers: epriestley, jungejason
CC: aran, epriestley
Differential Revision: 473
Differential
Summary:
Make some display stuff more consistent.
Test Plan:
Looked at a task and a revision.
Reviewed By: jungejason
Reviewers: aran, jungejason, tuomaspelkonen
CC: aran, jungejason
Differential Revision: 462
Summary:
See attached tasks. See D459 for the ability to merge tasks.
Test Plan:
Looked at posted and unposted inline comments.
Reviewed By: aran
Reviewers: edward, viyer, aran, jungejason, tuomaspelkonen
CC: aran, epriestley
Differential Revision: 461
errors.
Summary:
Make sure reviewers know what they are doing.
Test Plan:
Tested with different diffs that had lint and unit problems.
Reviewed By: epriestley
Reviewers: epriestley, jungejason
CC: grglr, aran, epriestley, tuomaspelkonen
Differential Revision: 432
Summary:
ReviewBoard has a fancier version of this feature that's more granular -- the
keyboard can focus on individual changes. I think that's good and intend to
implement something similar, but this gets us a step closer and gets rid of some
of the bookkeeping stuff like making shortcuts discoverable.
(I have another brnach with Maniphest merging which also uses fatcow icons,
which is why the README seems a little out of context.)
Test Plan:
Used "j" and "k" to jump between changesets. Pressed "?" and got a list of
available shortcuts.
Reviewed By: tuomaspelkonen
Reviewers: aran, jungejason, tuomaspelkonen
CC: moskov, aran, epriestley, tuomaspelkonen
Differential Revision: 412
Summary:
- Make wrap width settable in PHP.
- Dynamically generate max-width based on configurable maximum width.
- Constrain non-diff elements to standard width.
- Provide a configuration setting.
Test Plan:
Set various things to 100 / 120, as far as I could tell everything seemed to
render sensibly? This should have no effect on 80-col changes.
Reviewed By: jdperlow
Reviewers: jdperlow, tuomaspelkonen, jungejason, aran
CC: aran, jdperlow
Differential Revision: 413
Summary:
They currently have "Next", "Previous" and "Reply" links which don't work. Don't
render these links.
Test Plan:
Looked at inline previews, didn't see any silly/nonfunctional links.
Reviewed By: tuomaspelkonen
Reviewers: aran, jungejason, tuomaspelkonen
CC: aran, epriestley, tuomaspelkonen
Differential Revision: 419
Summary:
It was not possible before to update arc unit results for a postponed
test. This change makes it possible. Also the number of postponed tests
are shown in differential.
Let me know if this looks too Facebook specific.
Test Plan:
Tested the conduit call manually from Conduit Console and updated test
results for a diff that had 20 postponed tests.
Reviewed By: jungejason
Reviewers: epriestley, jungejason
Commenters: epriestley
CC: slawekbiel, aran, tuomaspelkonen, jungejason, epriestley
Differential Revision: 416
reviewers as metadata
Summary:
The "Add reviewer" implementation is super lazy right now since I didn't want to
do a schema change. Man up and add a column. I also plan to store "via"
information here (e.g., via email or via mobile).
NOTE: This schema change may take a while since the comment table is pretty big
in Facebook's install.
This needs a little CSS work but I think it's reasonable for now.
Test Plan:
Made comments on revisions and tasks. Added reviewers to a revision, got linked
names instead of a blob of text.
Reviewed By: aran
Reviewers: tomo, jungejason, tuomaspelkonen, aran
Commenters: tomo
CC: aran, tomo, epriestley
Differential Revision: 394
Summary:
- added a new config class for representing the kind of repetition a rule has
(once, every time, first time only)
- added an email action to herald rules for differential to allow someone to get
an email but only the first one
- changed the herald rule ui to allow a user to pick the amount of repetition
Test Plan:
created a test rule and ran it over and over
Reviewed By: epriestley
Reviewers: epriestley, tuomaspelkonen
CC: aran, epriestley, gc3
Revert Plan:
Tags:
- begin *PUBLIC* platform impact section -
Bugzilla: #
- end platform impact -
Differential Revision: 357
Summary:
When a user hits 'cancel' on a 'new', 'edit', or 'reply' operation, add a little
"Changes discarded. __Undo__" insert so they can get their change back. No undo
for delete since there's an explicit prompt. Once this lands we can make
'escape' work again to close dialogs.
This change started feeling really good when I was merging all the duplicate
code and making things more consistent, but by the time I started writing client
rendering it felt gross. I'm not really thrilled with it but I guess it's a step
forward? The feature seems pretty OK in practice. Let me know how much barfing
this causes and I can try to remedy the most acute concerns.
This also fixes a bug where replies always (?) appear on the 'new' side of the
diff (I think?).
Test Plan:
Applied 'new', 'edit', 'delete' and 'reply' operations, pressed 'cancel' and
'okay' in each case, with and without changing text where relevant. All
behaviors seem to conform with expectations, except that canceling out of 'edit'
without changing the text gives you an option to undo when it shouldn't really.
There's no super easy way to get at the original text right now.
Reviewed By: aran
Reviewers: aran, jungejason, tuomaspelkonen
CC: simpkins, aran, epriestley
Differential Revision: 406
Summary:
Historically, we had a bug at some point which caused inline comments to get
associated with changeset 0. Prevent that explicitly. See T108.
Test Plan:
Set "$changeset = 0" in the endpoint and got an exception.
Reviewed By: aran
Reviewers: aran
CC: aran
Differential Revision: 374
Summary:
Never converted this TODO over from XHP.
NOTE: This looks terrible since the CSS didn't make it over, can one of you grab
the rules for .differential-property-table and .property-table-header? If they
aren't still in trunk, try history for html/intern/css/tools/differential/
Test Plan:
Made a property change, looked at it in Differentila.
Reviewed By: aran
Reviewers: jungejason, tuomaspelkonen, aran
CC: codeblock, aran
Differential Revision: 409
Summary:
When whitespace changes between two non-whitespace characters (e.g., in a
string), always treat it as a change.
Test Plan:
Disabled render cache, made internal and external whitespace changes, rendered a
diff, got internal change always marked and external change marked correctly
depending on mode.
Reviewed By: tuomaspelkonen
Reviewers: tuomaspelkonen
CC: aran, tuomaspelkonen, epriestley
Differential Revision: 403
Summary:
The position of the custom rule was incorrect.
Test Plan:
Tested that Facebook task remarkup, Differential remarkup, image macros, quotes,
and random characters were working correctly in comment preview
Reviewed By: jungejason
Reviewers: jungejason
CC: aran, jungejason
Differential Revision: 396
Summary:
I somehow missed this, we render silly nonsense in the comment previews right
now. Don't render these links if we're rendering a preview.
Test Plan:
Looked at comment previews, less nonsense.
Reviewed By: tuomaspelkonen
Reviewers: aran, jungejason, tuomaspelkonen
CC: aran, tuomaspelkonen
Differential Revision: 388
Summary:
See task. Allows users to unsubscribe via email.
Test Plan:
Used mail receiver to unsubscribe from a revision. Tested subscribe/unsubscribe
buttons. Verified "!unsubscribe" appears as an avilable action in email.
Reviewed By: ola
Reviewers: ola, mroch, jungejason, tuomaspelkonen, aran
CC: aran, epriestley, ola
Differential Revision: 385
Summary:
See T178. D372 is the correct fix for this problem, hardcoding .sql3 is not.
Test Plan:
This code should be unreachable after T178 since these files will always be
marked as binary.
Reviewed By: tuomaspelkonen
Reviewers: tuomaspelkonen
CC: elgenie, aran, tuomaspelkonen
Differential Revision: 373
Summary:
Allows you to link to comments with "D123#3" or "T123#3", then adds a pile of JS
to try to make it not terrible. :/
The thing I'm trying to avoid here is when someone says "look at this!
http://blog.com/#comment-239291" and you click and your browser jumps somewhere
random and you have no idea which comment they meant. Since I really hate this,
I've tried to avoid it by making sure the comment is always highlighted.
Test Plan:
Put T1#1 and D1#1 in remarkup and verified they linked properly.
Clicked anchors on individual comments.
Faked all comments hidden in Differential and verified they expanded on anchor
or anchor change.
Reviewed By: aran
Reviewers: aran, tomo, mroch, jungejason, tuomaspelkonen
CC: aran, epriestley
Differential Revision: 383
Summary:
Vendor specific markups are now possible.
Test Plan:
Tested with the Facebook specific tasks markup.
Reviewed By: jungejason
Reviewers: epriestley, jungejason
CC: aran, jungejason
Differential Revision: 349
Summary:
add logging when syntax highlighting parsing throws exception.
Test Plan:
test when exception is thrown with non-php code. I couldn't
create a file to trigger an exception in running pygmentiza, so I
manually threw an exception to test it.
Reviewed By: tuomaspelkonen
Reviewers: tuomaspelkonen, epriestley
CC: aran, tuomaspelkonen, jungejason
Differential Revision: 350
Summary:
Various CSS tweaks and fixes:
- Add remarkup styling to description change views, missed this before.
- Fix CSS so that transactions with only one item (e.g., changed priority)
don't have weird floater underneath them.
- Add more space between transaction items.
- Make default background color lighter and less heavy.
- Use beigey color for comment form in Maniphest.
- Share more CSS between Maniphest and Differential (previews, feedback).
- Move "Leap Into Action" call to Differential, replace Maniphest with
thematically-consistent "Weigh In" (obviously, Maniphest has a nautical theme).
Test Plan:
Browsed Maniphest and Differential in a couple browsers, styling all seems
correct.
Reviewed By: tomo
Reviewers: tomo, aran, jungejason, tuomaspelkonen
CC: anjali, aran, tomo
Differential Revision: 328
Summary:
Some changeset metadata was not being correctly passed between the top-level
parser and the subparser, so it would be lost or incorrect when rendering
headers like "This file was moved from x to y." or rendering certain content
shields, like "the contents of this file were not modified".
Test Plan:
Created a new diff with a file move in it, rendered it, saw "This file was moved
from README to READYOU" correctly.
Reviewed By: aran
Reviewers: tuomaspelkonen, jungejason, grglr, aran
CC: aran, epriestley
Differential Revision: 321
Summary:
Use the new API from D322 to highlight text in parallel in Differential.
Test Plan:
Verified that pygemntize calls started within 20ms of one another in DarkConsole
(also: added a feature to let me do this) instead of running serially.
Reviewed By: aran
Reviewers: jungejason, tuomaspelkonen, aran
CC: aran
Differential Revision: 323
Summary:
This is just fluff to let me mailfilter my local sandbox. Would also allow the
Facebook install to return to "[diff]" if eletuchy is still unhappy about this
change.
Test Plan:
Triggered maniphest/differential emails, had normal prefixes. Overrode prefixes
in my custom config, got sandbox-unique prefixes.
Reviewed By: aran
Reviewers: jungejason, tuomaspelkonen, aran
CC: elgenie, aran
Differential Revision: 291
Phabricator
Summary:
Hook up the last pieces. This shouldn't impact the Facebook install, EXCEPT that
I removed "!accept" and added "!rethink" (plan changes). If you want to continue
supporting !accept, you should override the method in your subclass if you don't
already.
Test Plan:
Used the Mail Receiver test console to send mail to tasks and revisions.
Reviewed By: aran
Reviewers: jungejason, tuomaspelkonen, aran
CC: aran
Differential Revision: 289
Summary:
You can currently attach tasks to revisions from Differential, but not revisions
to tasks from Maniphest. Allow editing from either side.
This logic is kind of tricky but the alternative was massive code duplication.
Test Plan:
Added and removed revisions from maniphest. Added and removed tasks from
differential.
This should have no impact on the Facebook install since none of this is used
there.
Reviewed By: aran
Reviewers: tomo, tuomaspelkonen, jungejason, aran
CC: aran, epriestley
Differential Revision: 288
Summary:
Exception was thrown because there is no getRenderingReference function
for changeset.
Test Plan:
Sandbox loaded and links were working.
Reviewed By: grglr
Reviewers: grglr, epriestley
CC: aran, grglr
Differential Revision: 281
Summary:
Separates changeset IDs from rendering. Now each changeset has a "rendering
reference" which is basically a description of what the ajax endpoint should
render. For Differential, it's in the form "id/vs". For Diffusion,
"branch/path;commit".
I believe this fixes pretty much all of the bugs related to "show more" breaking
in various obscure ways, although I never got a great repro for T153.
Test Plan:
Clicked "show more" in diffusion change and commit views and differential diff,
diff-of-diff, standalone-diff, standalone-diff-of-diff views. Verified refs and
'whitespace' were always sent correctly.
Made inline comments on diffs and diffs-of-diffs. Used "Reply".
Reviewed By: tuomaspelkonen
Reviewers: tuomaspelkonen, jungejason, aran
CC: aran, tuomaspelkonen, epriestley
Differential Revision: 274
Summary:
D254 removed DifferentialReplyHandler::getRevision(), but
is still using it in two places. Correct them.
Test Plan:
send email email handler and verified it works.
Reviewed By: epriestley
Reviewers: tuomaspelkonen, epriestley
CC: aran, epriestley
Blame Revision:
D254
Differential Revision: 277
Summary:
Provide a base PhabricatorMailReplyHandler class which handles the plumbing for
multiplexing email if necessary and supporting public and private reply handler
addressses. DifferentialReplyHandler now extends it, and a new
ManiphestReplyHandler also does.
The general approach here is that we have three supported cases:
- no reply handler, default config, same as what we're doing now
- public reply handler, requires overriding classes but just sets "reply-to"
to some address the install generates and still sends only one email
- private reply handler, provides a default generation mechanism or you can
override it and splits mail apart so we send one to each recipient
Test Plan:
Sent email from Maniphest and Differential with and without
reply-handler-domains set.
Reviewed By: aran
Reviewers: jungejason, tuomaspelkonen, aran
CC: aran, epriestley
Differential Revision: 254
Summary:
Browsing comments was a bit difficult without the possibllity to jump
between comments. These links will make the browsing easier.
Test Plan:
Tested on multiple diffs that the links were working correctly.
Reviewed By: epriestley
Reviewers: epriestley
CC: jungejason, aran, tuomaspelkonen, epriestley
Differential Revision: 266
Summary:
Expanding lines duplicated some lines occasionally, because whitespace
option was different for the original request and the following request.
Test Plan:
Tested that the broken changeset was correct now.
Reviewed By: epriestley
Reviewers: epriestley
CC: jungejason, aran, epriestley
Differential Revision: 263
Summary:
Links to comments were not working because file was hidden after it was deleted.
Test Plan:
Tested that comment anchors were working correctly for deleted files.
Tested that generated files were still hidden.
Reviewed By: epriestley
Reviewers: epriestley
CC: jungejason, aran, epriestley
Differential Revision: 257
Summary:
These didn't get covered when we added square brackets to the rest of the emails
(so gmail can thread them properly).
Test Plan:
Created a local diff, got an email with brackets.
Reviewed By: tuomaspelkonen
Reviewers: rm, tuomaspelkonen, jungejason, aran
CC: aran, epriestley, tuomaspelkonen
Differential Revision: 256
Summary:
Fixes the issue caused by rPa0af5b66437719dba6136579c051982ab275e6a0. Prior to
that patch, isCommentInNewFile() returned $comment->getIsNewFile(). While this
was often the wrong value, it came from the database and was the integer 1 if
true.
After the patch, the function returns 'true' as a boolean, which is passed to JS
and then back to PHP, interpreted as an integer, and evaluates to 0.
To avoid this issue in general, provide an isBool() method on AphrontRequest
which interprets this correctly.
I will also revert the revert of rPa0af5b66437719dba6136579c051982ab275e6a0 when
I land this.
Test Plan:
Clicked "reply" on the right hand side of a diff, got a right-hand-side inline
comment.
Reviewed By: rm
Reviewers: tuomaspelkonen, jungejason, aran, rm
CC: simpkins, aran, epriestley, rm
Differential Revision: 250
Summary:
add email reply handler so that the user can reply to a
differential email to act on the revision. It generates the reply-to
email address, creates email body text with supported commands list, and
handle the action request on the differential revision.
Right now the reply-to handing is disabled in the config file. But a
site using Phabricator can enable it and implement a class
inheriting from DifferentialReplyHandler to enable customized email
handing.
Later we will need to add code to DifferentialMail.php to support
sending separate email to each email recipient to achieve better
security (see D226). The reply-to will be something like
D<revision_id>+<user_id>+<hash>@domain.com. We will create separate task
for it.
Test Plan:
tried comment on a revision from web UI and the email was
sent out as before without any change. When a subclass of
DifferentialReplyHandler is implemented and enabled, email's reply-to is
set and email text is added. Reply to the email with valid command did
create action to the revision.
Reviewed By: epriestley
Reviewers: tuomaspelkonen, epriestley, slawekbiel, dpepper
CC: aran, epriestley, jungejason
Differential Revision: 224
Summary:
Gmail ignores text inside of [square brackets] when deciding what to group
together. This diff does two things to create the right behavior for gmail:
1. put the verb text inside of [square brackets] so different verbs don't
break gmail threading.
2. Add the Diff ID to the email thread, so different diffs with the same name
don't group together.
Furthermore, to aid in distinguishing who is doing what when the from field
can't be spoofed, this diff adds the usename just before the verb. This works
quite well in the english language. For example:
[Differential] [rm requested a review of] D1: [admin] Create arcconfig for
code reviews
[Differential] [rm commented on] D1: [admin] Create arcconfig for code reviews
It's almost like a complete sentence. All it's missing is a period.
Test Plan:
Did it live on my test setup. Received emails with subjects that looked right.
Verified that gmail grouped the emails despite the different actions taking
place (tested: comments, planned changes, request review).
Reviewed By: epriestley
Reviewers: epriestley, jungejason
CC: aran, epriestley, rm
Differential Revision: 251
Summary:
This is a followup to D228. Basically, we use "changeset" (and, implicitly,
changesetID) for way too much stuff now.
One thing it can't possibly capture is the complete, arbitrary mapping between
the left and right sides of the displayed diff and the places we want to store
the left and right side comments. This causes a bunch of bugs; basically adding
inline comments is completely broken in diff-of-diff views prior to this patch.
Make this mapping explicit.
Note that the renderer already passes this mapping to
DifferentialChangesetParser which is why there are no changes outside this file,
I just didn't finish the implementation during the port.
This has the nice side-effect of fixing T132 and several other bugs.
Test Plan:
Made new-file and old-file comments on a normal diff; reloaded page, verified
comments didn't do anything crazy.
Expanded text on a normal diff, made new-file and old-file comments; reloaded
page, verified comments.
Repeated these steps for a previous diff in the same revision; verified
comments.
Loaded diff-of-diffs and verified expected comments appeared. Made new left and
right hand side comments, which almost work, see below.
NOTE: There is still a bug where comments made in the left-display-side of a
diff-of-diffs will incorrectly be written to the right-storage-side of the
right-display-side diff. However, this is an issue with the JS (the PHP is
correct) so I want to pull it out of scope for this patch since I think I need
to fix some other JS stuff too and this improves the overall state of the world
even if not everything is fixed.
Reviewed By: tuomaspelkonen
Reviewers: jungejason, tuomaspelkonen, aran, ola
CC: aran, tuomaspelkonen, epriestley
Differential Revision: 237
Summary:
Point users toward 'arc amend', 'arc commit', 'arc patch' and 'arc export' since
no one is going to read 'arc help'.
There's kind of a tradeoff here where we're wasting a fair amount of UI space
for expert users with the patch/export hints but I think it's probably okay
since there's really no other way to figure out that these features exist.
Note that the "export" command given isn't complete (it needs --git or
--unified), but it will give you a useful error message when you run it, telling
you to specify --git or --unified. If it turns out users get confused by this,
let me know.
Test Plan:
Loaded a revision and looked at it. Faked it into 'accepted' status.
Reviewed By: tuomaspelkonen
Reviewers: tuomaspelkonen, aran, jungejason
CC: aran, tuomaspelkonen, epriestley
Differential Revision: 242
Summary:
Add links to the 'standalone view' to grab the raw source text. I think this
operation is rare enough that it's okay to hide it like this. I changed
'Standalone View' to 'View Standalone / Raw' to improve discoverability.
This also fixes the broken Standalone View links in Diffusion by no longer
rendering them.
Test Plan:
viewed old and new sources for a changeset
Reviewed By: tuomaspelkonen
Reviewers: tuomaspelkonen, jungejason, aran
CC: aran, tuomaspelkonen
Differential Revision: 243
Summary:
Provides support for per-user x per-object unique reply-to email addresses, plus
SMTP integration.
This does not actually make Phabricator use these in outbound email.
Test Plan:
Used test console to validate in-Phabricator routing and handling.
Piped emails into the "mail_handler.php" script to validate mail parsing.
Configured sendmail and sent mail to Phabricator.
Technically I haven't conducted all parts of this test on the same machine since
I lost the will to configure more SMTP servers after configuring phabricator.com
Reviewed By: jungejason
Reviewers: jungejason, tuomaspelkonen, aran
CC: aran, epriestley, jungejason
Differential Revision: 226
Summary:
DifferentialChangesetParser currently takes the Changeset object to mean a bunch
of different and mutually conflicting things implicitly:
- Changeset ID is used to access the render cache.
- Changeset ID is also used to tell the ajax endpoint what to render when
clicking "show more".
- Changeset object has the actual changes.
- Changeset ID and "oldChangesetID" are used to choose where to show inline
comments and how to attach new ones.
This indirectly causes a bunch of problems, like T141 and T132. Move toward
making all these separate things explicit. I want to have the changeset object
only mean the actual changes to display.
Test Plan:
Looked at changesets and verified the render cache was accessed correctly (and
not accessed in other cases).
Reviewed By: tuomaspelkonen
Reviewers: jungejason, tuomaspelkonen, aran
CC: aran, epriestley, tuomaspelkonen
Differential Revision: 228
Summary:
This mode was fixed in D174 by grglr and is the best mode for almost all changes
once nonbroken, so make it the default.
This is also the mode which takes advantage of the render cache.
Test Plan:
Loaded a revision, got "Ignore All" as the default mode, and revision rendering
wasn't silly/broken.
Reviewed By: tuomaspelkonen
Reviewers: grglr, jungejason, tuomaspelkonen, aran
CC: aran, tuomaspelkonen
Differential Revision: 229
Summary:
Previously, Remarkup allowed you to paste in an image URI and get an inline
image. However, it did this by hotlinking the image which isn't so hot in an
open source product.
Restore this feature, but use image proxying instead. The existing image macro
code does most of the work.
There is a mild security risk depending on the network setup so I've left this
default-disabled and made a note about it. It should be safe to enable for
Facebook.
Test Plan:
Pasted in image and non-image links, got reasonable behavior. Verified proxying
appears to work. Verified that file:// shenanigans produce 400.
Reviewed By: tuomaspelkonen
Reviewers: aran, jungejason, tuomaspelkonen
Commenters: cpiro
CC: aran, cpiro, tuomaspelkonen
Differential Revision: 214
Summary:
The old query was effectively
SELECT DISTINCT revision.* FROM `differential_revision` revision
JOIN `differential_relationship` relationship ON
(relationship.revisionID = revision.id
AND relationship.objectPHID in
('PHID-USER-a113b9ae4ee9524d0a20'))
OR revision.authorPHID = 'PHID-USER-a113b9ae4ee9524d0a20'
LEFT JOIN `differential_viewtime` viewtime ON
viewtime.viewerPHID in ('PHID-USER-a113b9ae4ee9524d0a20')
AND viewtime.objectPHID = revision.phid
AND GREATEST(1304022277, IFNULL(viewtime.viewTime, 0)) <
revision.dateModified
ORDER BY dateModified DESC;
I'm not a db performance expert but it looks like the problem is that we
have to scan all revisions
mysql> EXPLAIN SELECT DISTINCT revision.* FROM `differential_revision`
revision JOIN `differential_relationship` relationship ON
(relationship.revisionID = revision.id AND relationship.objectPHID
in ('PHID-USER-a113b9ae4ee9524d0a20')) OR revision.authorPHID =
'PHID-USER-a113b9ae4ee9524d0a20' LEFT JOIN `differential_viewtime`
viewtime ON viewtime.viewerPHID in ('PHID-USER-a113b9ae4ee9524d0a20')
AND viewtime.objectPHID = revision.phid AND GREATEST(1304022277,
IFNULL(viewtime.viewTime, 0)) < revision.dateModified ORDER BY
dateModified DESC;
+----+-------------+--------------+-------+--------------------+------------+---------+-------+--------+------------------------------------+
| id | select_type | table | type | possible_keys | key |
key_len | ref | rows | Extra |
+----+-------------+--------------+-------+--------------------+------------+---------+-------+--------+------------------------------------+
| 1 | SIMPLE | revision | ALL | PRIMARY,authorPHID | NULL |
NULL | NULL | 254127 | Using temporary; Using filesort |
| 1 | SIMPLE | viewtime | ref | PRIMARY | PRIMARY |
66 | const | 17 | Distinct |
| 1 | SIMPLE | relationship | index | PRIMARY,objectPHID | objectPHID |
72 | NULL | 966900 | Using where; Using index; Distinct |
+----+-------------+--------------+-------+--------------------+------------+---------+-------+--------+------------------------------------+
The new query is a lot faster
mysql> EXPLAIN SELECT revs.* FROM ( (SELECT revision.* FROM
`differential_revision` revision WHERE revision.authorPHID in
('PHID-USER-a113b9ae4ee9524d0a20')) UNION (SELECT revision.* FROM
`differential_revision` revision JOIN differential_relationship rel WHERE
rel.revisionId = revision.Id AND rel.objectPHID =
'PHID-USER-a113b9ae4ee9524d0a20')) as revs LEFT JOIN `differential_viewtime`
viewtime ON viewtime.viewerPHID = 'PHID-USER-a113b9ae4ee9524d0a20' AND
viewtime.objectPHID = revs.phid WHERE GREATEST(1304022277,
IFNULL(viewtime.viewTime, 0)) < revs.dateModified ORDER BY revs.dateModified;
+----+--------------+------------+--------+--------------------+------------+---------+-----------------------------------------+------+--------------------------+
| id | select_type | table | type | possible_keys | key |
key_len | ref | rows | Extra
|
+----+--------------+------------+--------+--------------------+------------+---------+-----------------------------------------+------+--------------------------+
| 1 | PRIMARY | <derived2> | ALL | NULL | NULL |
NULL | NULL | 3021 | Using filesort
|
| 1 | PRIMARY | viewtime | ref | PRIMARY | PRIMARY |
66 | const | 17 | Using where
|
| 2 | DERIVED | revision | ref | authorPHID | authorPHID |
67 | | 1040 | Using where
|
| 3 | UNION | rel | ref | PRIMARY,objectPHID | objectPHID |
66 | | 3822 | Using where; Using
index |
| 3 | UNION | revision | eq_ref | PRIMARY | PRIMARY |
4 | phabricator_differential.rel.revisionID | 1 |
|
| NULL | UNION RESULT | <union2,3> | ALL | NULL | NULL
| NULL | NULL | NULL |
|
+----+--------------+------------+--------+--------------------+------------+---------+-----------------------------------------+------+--------------------------+
Test Plan:
Loaded differential updates with new query, made sure page loaded quickly. Ran
the query from the command-line, it took about .4 seconds.
Reviewed By: Girish
Reviewers: tuomaspelkonen, jungejason, Girish
Commenters: btrahan
CC: aran, btrahan, elynde, Girish
Differential Revision: 181
Summary:
The 'All Revisions and Reviews' Query takes about 2 seconds when I run
it from the mysql command-line:
SELECT revision.*
FROM `differential_revision` revision LEFT JOIN
`differential_relationship` relationship
ON revision.id = relationship.revisionID
AND relationship.relation = 'revw'
WHERE revision.authorPHID in ('PHID-USER-a113b9ae4ee9524d0a20') OR
relationship.objectPHID in ('PHID-USER-a113b9ae4ee9524d0a20')
GROUP BY revision.id ORDER BY dateModified DESC
2419 rows in set (2.05 sec)
This takes about 0.1-0.2 seconds. Just dug into this because I guess
phabricator is haven't a bunch of mysql timeouts.
I don't know what the hell I'm doing; this is just faster
Test Plan:
Loaded 'All Revisions and Reviews' in sandbox
http://phabricator.dev1577.snc6.facebook.com/differential/filter/related/
Made sure it had same results as the version in prod
https://phabricator.fb.com/differential/filter/related/
Still slow to generate all that html
Reviewed By: epriestley
Reviewers: epriestley, aran, tuomaspelkonen, jungejason
CC: aran, epriestley
Differential Revision: 182
Summary:
This isn't terribly elegant but it solves the problem without loss of
generality. We can pursue a more finessed solution later if it seems prudent.
Test Plan:
Created a revision matched by a blanket herald rule, and then commented on it.
Comment email had X-Herald-Rules header in it.
Reviewed By: aran
Reviewers: aran, tuomaspelkonen, jungejason
CC: aran
Differential Revision: 218
Summary:
The diff view page should point to the revision, if the diff has
already been attached to one. The form to select the revision was also
removed in this case.
Let's me know if it should be possible to reattach a diff to a different
revision.
Test Plan:
Tested that a new diff created with '--preview' option was not attached to
any revision. After attaching the diff manually, made sure that the diff view
page
showed the link to the revision correctly.
Reviewed By: epriestley
Reviewers: jungejason, epriestley
CC: aran, epriestley
Differential Revision: 216
Summary:
I only actually enabled it in Remarkup previously.
Test Plan:
Created a python diff, got syntax highlighted.
Reviewed By: aran
Reviewers: aran, tuomaspelkonen, jungejason
CC: aran, epriestley
Differential Revision: 202
Summary:
I pretty shortsightedly made sending a side effect of save() in the case that a
server is configured for immediate sending. Move this out, make it explicit, and
get rid of all the tangles surrounding it.
The web tool now ignores the server setting and only repsects the checkbox,
which makes far more sense.
Test Plan:
Sent mails from Maniphest, Differential, and the web console. Also ran all the
unit tests. Verified headers from Maniphest.
Reviewed By: rm
Reviewers: aran, rm
CC: tuomaspelkonen, rm, jungejason, aran
Differential Revision: 200
Summary:
Amazon SES does not allow us to set a Message-ID header, which means
that threads are incorrect in Mail.app (and presumably other applications
which respect In-Reply-To and References) because the initial email does not
have anything which attaches it to the rest of the thread. To fix this, never
rely on Message-ID if the mailer doesn't support Message-ID.
(In the Amazon SES case, Amazon generates its own Message-ID which we can't
know ahead of time).
I additionally used all the Lisk isolation from the other tests to make this
testable and wrote tests for it.
I also moved the idea of a thread ID lower in the stack and out of
DifferentialMail, which should not be responsible for implementation details.
NOTE: If you push this, it will cause a one-time break of threading for
everyone using Outlook since I've changed the seed for generating Thread-Index.
I feel like this is okay to avoid introducing more complexity here.
Test Plan:
Created and then updated a revision, messages delivered over Amazon
SES threaded correctly in Mail.app. Verified headers. Unit tests.
Reviewed By: rm
Reviewers: aran, tuomaspelkonen, jungejason, rm
Commenters: aran
CC: aran, rm, epriestley
Differential Revision: 195
Summary:
DifferentialRevision stores this field as a dictionary but the
Editor incorrectly passed it to Herald as a raw value array. Ideally the
property should be called unsubscribedDict or something but I'm increasingly
thinking we're going to centralize subscriptions for Adjutant or some similar
system so I'm disinclined to pursue the schema change just yet. I provided an
explicit raw-value-oriented API, at least.
Test Plan:
With two accounts, A and B: created universal Herald CC rule with
user B, created a revision with user A, had user B unsubscribe, had user A
update the revision. User B was not resubscribed.
Reviewed By: aran
Reviewers: jungejason, tuomaspelkonen, aran
CC: ola, aran, epriestley
Differential Revision: 188
Summary:
Thanks to erling for the report. This was XSSable, although you could
only get yourself.
Test Plan:
Made a comment like "</textarea><h1>" and edited it before and after
the patch. Proper behavior with this patch.
Reviewed By: aran
Reviewers: erling, jungejason, tuomaspelkonen, aran
CC: aran
Differential Revision: 187
Summary:
If you don't have a custom renderer, this variable does not get set
and emits a warning when you try to read it.
Test Plan:
Loaded page before and after change, warnings went away (lines
145 and 154 in old file).
Reviewed By: aran
Reviewers: tuomaspelkonen, aran
CC: jungejason, aran, epriestley
Differential Revision: 186
Summary:
Fixed buggy and incomplete logic for handling IGNORE ALL mode properly.
A subparser is used to parse the non-ws-ignoring changeset while a
ws-ignoring changeset is handed off to the original parser. At a later
step, the original parser queries the subparser for its lines of text
(which are formatted properly due to being in non-ws-ignoring mode) and
uses them to replace the text in the ws-ignoring diff.
Task ID: 549940
Test Plan:
-turn off caching temporarily (the cached view is still indented
improperly)
-visit http://phabricator.dev1943.facebook.com/D242591
-note aligned, but not completely highlighted, indentation right above the
comments complaining about
indentation issues
Reviewed By: aran
Reviewers: tuomaspelkonen, jungejason, aran
Commenters: epriestley
CC: aran, epriestley, grglr
Differential Revision: 174
Summary:
This adds a new view to differential called Updates.
The high-level goal of Updates is to enabled differential to be
effectively used without email notifications. I've tried doing things
like automatically deleting differential emails where I'm in the 'to'
line since they show up on the main diffential page but then there's
always the chance an important diff flies by without me seeing it. Also,
sometimes someone comments on a diff post-commit but differential
doesn't surface those diffs.
I re-created a test db on my devserver using mysqldump to get data on
revs > 230000 so I would have some test data. We need to add a simple
viewtime table but I didn't want to do that in production. Here's the
table:
CREATE TABLE differential_viewtime (
viewerPHID varchar(64) not null,
objectPHID varchar(64) not null,
viewTime int unsigned not null,
PRIMARY KEY (viewerPHID, objectPHID)
);
Issues:
-Once we turn this on, all diffs will be 'unviewed'. What do you think
about a 'Clear All' button or something?
-Maybe we should add a pager
This feature would be insanely useful, let me know what you think.
Test Plan:
Loaded Updates in my sandbox
http://phabricator.dev1577.snc6.facebook.com/differential/filter/updates/
Clicked a diff, then went back, made sure diff disappeared from Updates
list
Reviewed By: tuomaspelkonen
Reviewers: epriestley, jungejason, tuomaspelkonen
Commenters: epriestley
CC: epriestley, elynde, tuomaspelkonen
Differential Revision: 169
Summary:
Take Tuomas's first diff. Sorry, my bad on diff feedback.
Test Plan:
Load D240353 without any options and make sure 'ignore-trailing'
selected.
Reviewed By: tuomaspelkonen
Reviewers: tuomaspelkonen
CC: epriestley, aran, tuomaspelkonen
Differential Revision: 172
Summary:
'ignore-all' is very confusing, especially for Python, because the indentation
is incorrect.
Test Plan:
Tested loading D240353 without any options and made sure 'ignore-trailing' was
selected and the diff looked correct.
Reviewed By: aran
Reviewers: jungejason, simpkins, aran, dpepper
Commenters: dpepper
CC: grglr, aran, dpepper, tuomaspelkonen
Differential Revision: 171
