Summary:
- We used to have connection-level caching, so we needed getTransactionKey() to make sure there was one transaction state per real connection. We now cache in Lisk and each Connection object is guaranteed to represent a real, unique connection, so we can make this a non-static.
- I kept the classes separate because it was a little easier, but maybe we should merge them?
- Also track/implement read/write locking.
- (The advantage of this over just writing LOCK IN SHARE MODE is that you can use, e.g., some Query class even if you don't have access to the queries it runs.)
Test Plan: Can you come up with a way to write unit tests for this? It seems like testing that it works requires deadlocking MySQL if the test is running in one process.
Reviewers: vrana, btrahan
Reviewed By: vrana
CC: aran
Differential Revision: https://secure.phabricator.com/D2398
Summary:
Several problems:
- With fpm-warmup, 'PhabricatorAccessLog' is always loaded, even if it hasn't actually initialized. Use a global instead (barf). I'll fix this when I refactor index.php, hopefully soon.
- The 'POST' check isn't sufficient in Firefox for HTML5 uploads -- not 100% sure why, maybe it encodes post bodies differently? I added an additional '__file__' requirement, and will add this param to GET on all file uploads in a future diff.
See discussion in D2381.
Test Plan: Uploaded files with Firefox via drag-and-drop without various mysterious errors.
Reviewers: vrana, btrahan, jungejason
Reviewed By: vrana
CC: aran
Differential Revision: https://secure.phabricator.com/D2405
Summary:
The highlighting is distracting according to Nick Shrock and others.
Real designer, Lee Byron, helped me with this.
It also gives us unagressive target for jumping to the source line in future.
Another feature I will probably implement is highlighting also the source of copies/moves.
I will use the right side of the left column for it.
Test Plan:
Hover copied notifier.
Hover coverage notifier.
I've also checked that this doesn't break our super-flaky old/new code JavaScript detector.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin, leebyron, schrockn
Differential Revision: https://secure.phabricator.com/D2403
Summary: We currently try to do "app login" for all OAuth providers, but not all of them support it in a meaningful way. Particularly, it always fails for Google.
Test Plan: Ran google diagnostics on a working config, no longer got a diagnostic failure.
Reviewers: btrahan, vrana, csilvers
Reviewed By: csilvers
CC: aran
Differential Revision: https://secure.phabricator.com/D2377
Summary: Primarily for @csilvers who has 92 million repositories or something. This is a touch hacky, but movitated by pragmatism.
Test Plan:
- Ran "repository.create" to create repositories, "repository.query" to list them.
- Tested most or maybe all of the error conditions, probably.
Reviewers: btrahan, vrana, csilvers
Reviewed By: csilvers
CC: aran
Differential Revision: https://secure.phabricator.com/D2396
Summary: Generally moves us toward having a sane approach to transaction handling.
Test Plan: See test case, which fails before this patch and passes afterwards.
Reviewers: vrana, btrahan, jungejason
Reviewed By: vrana
CC: aran
Differential Revision: https://secure.phabricator.com/D2394
Summary:
It saves some time on non-highlighting generated and other not interesting code.
The code is quite complex (300 lines methods) so I'm not sure if everything is moved correctly.
P.S. I hope that moved code detector will work...
Test Plan:
Display generated file with all whitespace, verify that it is not highlighted.
Display normal file.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Maniphest Tasks: T1134
Differential Revision: https://secure.phabricator.com/D2358
Summary:
I wanted to point someone on a file uploaded to Phabricator and the normal link is just too long.
I guess that this also improves security. Because pointing someone to the file directly reveals the secret key used in /data/ and it can be served without auth?
We already use `{F123}` so there will be no conflicts in future because we wouldn't want to reuse it for something else.
I promote the link on /file/ - it adds one redirect but I think it's worth it. I also considered making the link from the File ID column but there are already too many links (with some duplicity).
Test Plan:
/file/
/F123 (redirect)
/F9999999999 (404)
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2380
Summary:
I will use it for highlighting users which are not currently available.
Maybe I will also use it in the nagging tool.
I don't plan creating a UI for it as API is currently enough for us.
Maybe I will visualize it at /calendar/ later.
I plan creating `user.deletestatus` method when this one will be done.
Test Plan:
`storage upgrade`
Call Conduit `user.addstatus`.
Verify DB.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2382
Summary: I have a patch which makes uploads all fancy and adds progress bars, but document the landscape first since it's quite complicated.
Test Plan: Generated, read docs. Configured `storage.upload-size-limit` to various values.
Reviewers: btrahan, vrana
Reviewed By: vrana
CC: aran
Maniphest Tasks: T875
Differential Revision: https://secure.phabricator.com/D2381
Summary: This script is not used since D976.
Test Plan:
grep init_env
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2379
lighttpd could support rules like this, to add efficiency:
# Aggressively cache all static files
$HTTP["url"] =~ "\.(jpg|gif|png|css|js|htc)" {
expire.url = ( "" => "access 1 years" )
}
# Compress files for faster transfer
compress.filetype = (
"text/plain",
"text/html",
"text/javascript",
"text/css",
"text/xml"
)
compress.cache-dir = <would need to set to something>?
I don't know if that is necessary or useful. Probably not a good idea
at this point, where the code is changing so rapidly: a 1 year cache
of javascript code could cause trouble. And i think the default
lighttpd.conf already compresses text/html, text/plain, text/css, and
application/x-javascript by default, so we're ok there (could add
text/javascript and text/xml, I guess).
Summary: We currently make a ludicrously gigantic permission request to do Google auth (read/write access to the entire address book), since I couldn't figure out how to do a more narrowly tailored request when I implemented it. @csilvers pointed me at some much more sensible APIs; we can now just ask for user ID, name, and email address.
Test Plan: Created a new account via Google Oauth. Linked/unlinked an existing account. Verified diagnostics page still works correctly. Logged in with a pre-existing Google account created with the old API (to verify user IDs are the same through both methods).
Reviewers: btrahan, vrana, csilvers, Makinde
Reviewed By: csilvers
CC: aran
Differential Revision: https://secure.phabricator.com/D2378
Summary: I will need it for nagging tool.
Test Plan:
None yet.
Please suggest me how to create a testing database (I need to insert some data in the table). I guess that it is now possible?
There is also probably some bug in `arc unit` - `setEnvConfig()` is not called before `getEnvConfig()` resulting in fatal error.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2376
Summary:
We will need it for two purposes:
- Status tool.
- Nagging tool - @aran suggested using "3 business days" and I don't want it to fall on New Year's Eve or such.
I don't plan working on any interface for editing this as this kind of data should be always imported.
Test Plan:
`bin/storage upgrade`
`scripts/calendar/import_us_holidays.php`
/calendar/
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2375
Summary: This allows writing inline comments and reduces different behavior between normal and very large diffs.
Test Plan:
Verify that normal diff works.
Verify that very large diff works.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2361
Summary:
This adds support to differential fields to display warnings before a revision
gets accepted. Since lint and unit are differential fields, the code for their
warnings was moved into their respective field specification classes, so there
is only one code path for warnings (lint, unit, or custom).
Test Plan:
Select 'Accept' on a revision with lint/unit warnings and see messages appear
like they used to. Change it back to 'Comment' and they go away. Repeat with
a revision without lint/unit warnings and see no warnings appear. Checked
darkconsole to see no errors due to this.
Reviewers: jungejason, epriestley, vrana
Reviewed By: epriestley
CC: aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2363
Summary: NOTE: `renderViewOptionsDropdown()` adds unnecessary parameters to URL but the link just redirects anyway.
Test Plan:
Show Raw File (Left and Right) in SVN and Git.
Verify also Added and Deleted files.
Reviewers: epriestley, aran
Reviewed By: epriestley
CC: Koolvin
Differential Revision: https://secure.phabricator.com/D2370
Summary:
- This is only slightly useful for updating Differential, since DiffQuery (vs RawDiffQuery) already gets you most of what you need. The only thing is that DiffQuery returns the diff for one path only right now(and the SVN version is very "special"). Should be easy to fix in the Git/HG cases at least, though (or maybe just use RawDiffQuery to avoid the SVN mess).
- Added a "download raw diff" link.
Test Plan: Viewed Diffusion and raw commits for SVN, Mercurial and Git repositories.
Reviewers: vrana, btrahan, jungejason
Reviewed By: vrana
CC: aran
Differential Revision: https://secure.phabricator.com/D2350
Summary:
- Unit tests can request storage fixtures.
- We build one fixture across all tests in the process, which can quickstart (takes roughly 1s to build, 200ms to destroy for me). This is a one-time cost for running an arbitrary number of fixture-based tests.
- We isolate all the connections inside transactions for each test, so individual tests don't affect one another.
Test Plan: Ran unit tests, which cover the important properties of fixtures.
Reviewers: btrahan, vrana, jungejason, edward
Reviewed By: btrahan
CC: aran, davidreuss
Maniphest Tasks: T140
Differential Revision: https://secure.phabricator.com/D2345
Summary:
When choosing a verb to show with a closed differential revision, choose the
verb based on the upstream vcs, not the vcs used to create the diff, since these
are not the same thing. I also updated the documentation for the next step for
an accepted diff for the case where the local vcs and backing vcs aren't the
same (since arc land doesn't work for those).
Test Plan:
Loaded a committed diff and an accepted diff from fbcode and www to check that
they show the correct thing.
Reviewers: jungejason, epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Maniphest Tasks: T1118
Differential Revision: https://secure.phabricator.com/D2360
Summary: Only inlines were indexed (contrary to what comment claims).
Test Plan: Index one revision, check database.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2359
Summary:
I think this improves things, let me know if you have feedback.
Also addresses T840.
Test Plan: See screenshots...
Reviewers: vrana, btrahan, jungejason
Reviewed By: vrana
CC: aran, zeeg
Maniphest Tasks: T840
Differential Revision: https://secure.phabricator.com/D2357
Summary: basically by validating we have good user data when we set the user data.
Test Plan: simulated a failure from a phabricator on phabricator oauth scenario. viewed ui that correctly told me there was an error with the provider and to try again.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Maniphest Tasks: T1077
Differential Revision: https://secure.phabricator.com/D2337
Summary: 'cuz it looks dumb to use a URI slug
Test Plan: viewed a post liked the title
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2338
Summary:
Before: {F10754}
After: {F10753}
Test Plan:
View revision with lint warnings and unit errors.
Click on Details.
Click on Details.
Click on Details.
Click on Details.
Reviewers: asukhachev, epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2351
Summary:
- Currently, connections are responsible for connection caching. However, I want unit tests to be able to say "throw away the entire connection cache" with storage fixtures, and this is difficult/impossible when connections are responsible for the cache.
- The only behavioral change is that previously we would use the same connection for read-mode and write-mode queries. We'll now establish two connections. No installs actually differentiate between the modes so it isn't particularly relevant what we do here. In the long term, we should probably check the "w" cache before building a new "r" connection, so transactional code which involves reads and writes works (we don't have any such code right now).
Test Plan: Loaded pages, verified only one connection was established per database. Ran unit tests.
Reviewers: btrahan, vrana, jungejason, edward
Reviewed By: vrana
CC: aran
Maniphest Tasks: T140
Differential Revision: https://secure.phabricator.com/D2342
Summary: Allow the default namespace to be set in configuration, so you can juggle multiple copies of sandbox test data or whatever.
Test Plan: Changed default namespace, verified web UI and "storage" script respect it.
Reviewers: btrahan, vrana, jungejason
Reviewed By: vrana
CC: aran
Maniphest Tasks: T345
Differential Revision: https://secure.phabricator.com/D2341
Summary:
Also reduce the memory usage a little bit (before increasing it again).
I use the same CSS class as for the copied code.
Test Plan: Parsed 100 diffs and checked about 10 of them - looks good.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2339
Summary:
This addresses three issues with the current patch management system:
# Two people developing at the same time often pick the same SQL patch number, and then have to go rename it. The system catches this, but it's silly.
# Second/third-party developers can't use the same system to manage auxiliary storage they may want to add.
# There's no way to build mock databases for unit tests that need to do reads.
To resolve these things, you can now name your patches whatever you want and conflicts are just merge conflicts, which are less of a pain to fix than filename conflicts.
Dependencies are now a DAG, with implicit dependencies created on the prior patch if no dependencies are specified. Developers can add new concrete subclasses of `PhabricatorSQLPatchList` to add storage management, and define the dependency branchpoint of their patches so they apply in the correct order (although, generally, they should not depend on the mainline patches, presumably).
The commands `storage upgrade --namespace test1234` and `storage destroy --namespace test1234` will allow unit tests to build and destroy MySQL storage.
A "quickstart" mode allows an upgrade from scratch in ~1200ms. Destruction takes about 200ms. These seem like fairily reasonable costs to actually use in tests. Building from scratch patch-by-patch takes about 6000ms.
Test Plan:
- Created new databases from scratch with and without quickstart in a separate test namespace. Pointed the webapp at the test namespaces, browsed around, everything looked good.
- Compared quickstart and no-quickstart dump states, they're identical except for mysqldump timestamps and a few similar things.
- Upgraded a legacy database to the new storage format.
- Destroyed / dumped storage.
Reviewers: edward, vrana, btrahan, jungejason
Reviewed By: btrahan
CC: aran, nh
Maniphest Tasks: T140, T345
Differential Revision: https://secure.phabricator.com/D2323
Summary:
Splits out the SQL changes. These are most of the changes, but primarily mechanical:
- Moved "initialize.sql" to "0000.legacy.sql" and partially reverted to an older version, such that patches 0000 + 000 + 001 + ... + 137 put us in the right state when applied sequentially.
- Removed "create database" commands from all SQL. These are handled by separate DB patches now, so we have the data to do operations like "storage databases" (list databases) and "storage destroy" (drop databases).
- Removed "phabricator_" namespace from all SQL, and replaced with "{$NAMESPACE}_" token so we can namespace databases.
- Shortened some column lengths so patches apply correctly if originally created as InnoDB; also a few similar tweaks elsewhere.
Test Plan: See D2323 for discussion and test plan.
Reviewers: edward, vrana, btrahan, jungejason
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T140, T345
Differential Revision: https://secure.phabricator.com/D2329
Summary:
- Show README on the repository screen.
- Move README to the bottom of the page for both repository and browse screens.
- Support "README.rainbow".
Test Plan: Looked at repository, browse screens. Made a "README.rainbow".
Reviewers: btrahan, vrana, jungejason
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T1104
Differential Revision: https://secure.phabricator.com/D2336
