Summary: See PHI1466. When an install defines task subtypes, show them on the task graph.
Test Plan:
- On desktop with subtypes defined, column is visible.
- On desktop with subtypes not defined, column is hidden.
- On mobile, column is hidden.
{F6896845}
Differential Revision: https://secure.phabricator.com/D20842
Summary: Depends on D20838. Fixes T13414. Instead of doing coarse diffing with "PhutilEditDistanceMatrix", use hash-and-diff with "DocumentEngine".
Test Plan:
- On a large document (~3K top level blocks), saw a more sensible diff, instead of the whole thing falling back to "everything changed" mode.
- On a small document, still saw a sensible granular diff.
{F6888249}
Maniphest Tasks: T13414
Differential Revision: https://secure.phabricator.com/D20839
Summary: Depends on D20836. Ref T13414. Ref T13425. Ref T13395. Move these to "phabricator/" before trying to improve the high-level diff engine in prose diffs.
Test Plan: Ran "arc liberate", looked at a prose diff (no behavioral change).
Maniphest Tasks: T13425, T13414, T13395
Differential Revision: https://secure.phabricator.com/D20838
Summary: Ref T13410. Fixes T4280. Allows you to put a named anchor into a document explicitly.
Test Plan: Used `{anchor ...}` in Remarkup, used location bar to jump to anchors.
Maniphest Tasks: T13410, T4280
Differential Revision: https://secure.phabricator.com/D20825
Summary:
Depends on D20820. Ref T13410. We currently cut anchor names in the middle, don't support emoji in anchors, and generate relatively short anchors.
Generate slightly longer anchors, allow more unicode, and try not to cut things in the middle.
Test Plan: Created a document with a variety of different anchors and saw them generate more usable names.
Maniphest Tasks: T13410
Differential Revision: https://secure.phabricator.com/D20821
Summary:
Fixes T13392. If you have 17 load balancers in sequence, Phabricator will receive requests with at least 17 "X-Forwarded-For" components in the header.
We want to select the 17th-from-last element, since prior elements are not trustworthy.
This currently isn't very easy/obvious, and you have to add a kind of sketchy piece of custom code to `preamble.php` to do any "X-Forwarded-For" parsing. Make handling this correctly easier.
Test Plan:
- Ran unit tests.
- Configured my local `preamble.php` to call `preamble_trust_x_forwarded_for_header(4)`, then made `/debug/` dump the header and the final value of `REMOTE_ADDR`.
```
$ curl http://local.phacility.com/debug/
<pre>
HTTP_X_FORWARDED_FOR =
FINAL REMOTE_ADDR = 127.0.0.1
</pre>
```
```
$ curl -H 'X-Forwarded-For: 1.1.1.1, 2.2.2.2, 3.3.3.3, 4.4.4.4, 5.5.5.5, 6.6.6.6' http://local.phacility.com/debug/
<pre>
HTTP_X_FORWARDED_FOR = 1.1.1.1, 2.2.2.2, 3.3.3.3, 4.4.4.4, 5.5.5.5, 6.6.6.6
FINAL REMOTE_ADDR = 3.3.3.3
</pre>
```
```
$ curl -H 'X-Forwarded-For: 5.5.5.5, 6.6.6.6' http://local.phacility.com/debug/
<pre>
HTTP_X_FORWARDED_FOR = 5.5.5.5, 6.6.6.6
FINAL REMOTE_ADDR = 5.5.5.5
</pre>
```
Maniphest Tasks: T13392
Differential Revision: https://secure.phabricator.com/D20785
Summary:
Ref T13336. Currently, "bin/storage destroy" destroys every master. This is wonderfully destructive, but if replication fails it's useful to be able to destroy only a replica.
Operate on a single host, and require "--host" to target the operation in cluster mode, so `bin/storage destroy --host dbreplica001` is a useful operation.
Test Plan: Ran `bin/storage destroy` with various flags locally. Will destroy `secure002` and refresh replication.
Maniphest Tasks: T13336
Differential Revision: https://secure.phabricator.com/D20784
Summary:
Depends on D20780. Ref T13403. During initial setup, it's routine to run "bin/config" with a bad database config. We start the stack in "config optional" mode to anticipate this.
However, even in this mode, we may emit warnings if the connection fails in certain ways. These warnings aren't useful; suppress them with "@".
(Possibly this message should move from "phlog()" to "--trace" at some point, but it has a certain amount of context/history around it.)
Test Plan:
- Configured MySQL to fail with a retryable error, e.g. good host but bad port.
- Ran `bin/config set ...`.
- Before: saw retry warnings on stderr.
- After: no retry warnings on stderr.
- (Turned off suppression code artificially and verified warnings still appear under normal startup.)
Maniphest Tasks: T13403
Differential Revision: https://secure.phabricator.com/D20781
Summary: Depends on D20779. Ref T13403. Bad parameters may cause this call to fail without setting an error code; if it does, catch the issue and go down the normal connection error pathway.
Test Plan:
- With "mysql.port" set to "quack", ran `bin/storage probe`.
- Before: wild mess of warnings as the code continued below and failed when trying to interact with the connection.
- After: clean connection failure with a useful error message.
Maniphest Tasks: T13403
Differential Revision: https://secure.phabricator.com/D20780
Summary:
Fixes T13336.
- Prevent `--no-indexes` from being combined with `--for-replica`, since combining these options can only lead to heartbreak.
- In `--for-replica` mode, dump caches too. See discussion in T13336. It is probably "safe" to not dump these today, but fragile and not correct.
- Mark the "MarkupCache" table as having "Cache" persistence, not "Data" persistence (no need to back it up, since it can be fully regenerated from other datasources).
Test Plan: Ran `bin/storage dump` with various combinations of flags.
Maniphest Tasks: T13336
Differential Revision: https://secure.phabricator.com/D20743
Summary:
Fixes T13390. We have some old code which doesn't dynamically select between "utf8mb4" and "utf8". This can lead to dumping utf8mb4 data over a utf8 connection in `bin/storage dump`, which possibly corrupts some emoji/whales.
Instead, prefer "utf8mb4" if it's available.
Test Plan: Ran `bin/storage dump` and `bin/storage shell`, saw sub-commands select utf8mb4 as the client charset.
Maniphest Tasks: T13390
Differential Revision: https://secure.phabricator.com/D20742
Summary:
Ref T13386. If you issue `differential.query` with a large offset (like 3000), it can overheat regardless of policy filtering and fail with a nonsensical error message.
This is because the overheating limit is based only on the query limit, not on the offset.
For example, querying for "limit = 100" will never examine more than 1,100 rows, so a query with "limit = 100, offset = 3000" will always fail (provided there are at least that many revisions).
Not all numbers work like you might expect them to becuase there's also a 1024-row fetch window, but basically small limits plus big offsets always fail.
Test Plan: Artificially reduced the internal window size from 1024 to 5, then ran `differential.query` with `offset=50` and `limit=3`. Before: overheated with weird error message. After: clean result.
Maniphest Tasks: T13386
Differential Revision: https://secure.phabricator.com/D20728
Summary:
Fixes T13384. Currently, the subtype "disabled" configuration is not respected when selecting fields for `ROLE_EDIT`.
The only meaningful caller for `ROLE_EDIT` is transaction validation, but transaction validation should respect fields being disabled by subtype configuration.
Test Plan:
- Added a "required" Maniphest custom field "F", then "disabled" it in a subtype "S".
- Created a task of subtype "S".
- Before: Form submission fails with error "F is required", even though the field is not actually visible on the form and can not be set.
- After: Form submits cleanly and creates the task.
Maniphest Tasks: T13384
Differential Revision: https://secure.phabricator.com/D20726
Summary:
Ref T13373. When you "bin/config set x ..." a value, the success message ("Set x ...") is somewhat ambiguous and can be interpreted as "First, you need to set x..." rather than "Success, wrote x...".
Make the messaging more explicit. Also make this string more translatable.
Test Plan: Ran `bin/config set ...` with various combinations of flags, saw more clear messaging.
Maniphest Tasks: T13373
Differential Revision: https://secure.phabricator.com/D20711
Summary: Fixes T13355. This didn't appear to be a ton of extra work, we just didn't get it for free in the original implementation in D14635.
Test Plan:
- Saw "date" custom fields appear in Conduit API documentation for "maniphest.edit".
- Set custom "date" field to null and non-null values via the API.
{F6666582}
Maniphest Tasks: T13355
Differential Revision: https://secure.phabricator.com/D20690
Summary:
See D20650. Long ago, this got added as "pastebin", but that's the name of another product/company, not a generic term for paste storage.
Rename the database to `phabricator_paste`.
(An alternate version of this patch would rename `phabricator_search` to `phabricator_bing`, `phabricator_countdown` to `phabricator_spacex`, `phabricator_pholio` to `phabricator_adobe_photoshop`, etc.)
Test Plan:
- Grepped for `pastebin`, now only found references in old patches.
- Applied patches.
- Browsed around Paste in the UI without encountering issues.
Reviewers: amckinley
Reviewed By: amckinley
Subscribers: PHID-OPKG-gm6ozazyms6q6i22gyam
Differential Revision: https://secure.phabricator.com/D20661
Summary:
Fixes T13345. See D20650. Currently, `PhabricatorCursorPagedPolicyAwareQuery` does a JOIN against the "title" field so it can apply additional ranking/ordering conditions to the query.
This means that documents with no title (which don't have this field) are always excluded from the result set.
We'd prefer to include them, just not give them any bonus ranking/relevance boost. Use a LEFT JOIN so they get included.
Test Plan:
- Applied D20650 (diff 1), made it use raw `getTitle()` as the document title, indexed a paste with no title.
- Searched for a term in the paste body.
- Before change: no results.
- After change: found result.
{F6601159}
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13345
Differential Revision: https://secure.phabricator.com/D20660
Summary:
Fixes T13342. This does a few different things, although all of them seem small enough that I didn't bother splitting it up:
- Support export of "remarkup" custom fields as text. There's some argument here to export them in some kind of structure if the target is JSON, but it's hard for me to really imagine we'll live in a world some day where we really regret just exporting them as text.
- Support export of "date" custom fields as dates. This is easy except that I added `null` support.
- If you built PHP from source without "--enable-zip", as I did, you can hit the TODO in Excel exports about "ZipArchive". Since I had a reproduction case, test for "ZipArchive" and give the user a better error if it's missing.
- Add a setup check for the "zip" extension to try to avoid getting there in the first place. This is normally part of PHP so I believe users generally won't hit it, I just hit it because I built from source. See also T13232.
Test Plan:
- Added a custom "date" field. On tasks A and B, set it to null and some non-null value. Exported both tasks to Excel/JSON/text, saw null and a date, respectively.
- Added a custom "remarkup" field, exported some values, saw the values in Excel.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13342
Differential Revision: https://secure.phabricator.com/D20658
Summary: See PHI1319. Ref T13291. Bump the remarkup cache version, since the old JIRA / Asana rules may exist in the partial cached representation of remarkup blocks from older versions.
Test Plan: Typed some comments with various formatting, saw remarkup work fine.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13291
Differential Revision: https://secure.phabricator.com/D20619
Summary:
Ref T13328. Currently, we read from `mysqldump` something like this:
```
until (done) {
for (100 ms) {
mysqldump > in-memory-buffer;
}
in-memory-buffer > disk;
}
```
This general structure isn't great. In this use case, where we're streaming a large amount of data from a source to a sink, we'd prefer to have a "select()"-like way to interact with futures, so our code is called after every read (or maybe once some small buffer fills up, if we want to do the writes in larger chunks).
We don't currently have this (`FutureIterator` can wake up every X milliseconds, or on future exit, but, today, can not wake for readable futures), so we may buffer an arbitrary amount of data into memory (however much data `mysqldump` can write in 100ms).
Reduce the update frequency from 100ms to 10ms, and limit the buffer size to 32MB. This effectively imposes an artificial 3,200MB/sec limit on throughput, but hopefully that's fast enough that we'll have a "wake on readable" mechanism by the time it's a problem.
Test Plan:
- Replaced `mysqldump` with `cat /dev/zero` as the source command, to get fast input.
- Ran `bin/storage dump` with `var_dump()` on the buffer size.
- Before change: saw arbitrarily large buffers (300MB+).
- After change: saw consistent maximum buffer size of 32MB.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13328
Differential Revision: https://secure.phabricator.com/D20617
Summary: Ref T13321. The daemons no longer write PID files, so we no longer need to pass any of this stuff to them.
Test Plan: Grepped for affected symbols.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13321
Differential Revision: https://secure.phabricator.com/D20608
Summary:
Fixes T13304. Shell pipes and redirects do not have robust behavior when errors occur. We provide "--compress" and "--output" flags as robust alternatives, but do not currently recommend their use.
- Recommend their use, since their error handling behavior is more robust in the face of issues like full disks.
- If "--compress" is provided but won't work because the "zlib" extension is missing, raise an explicit error. I believe this extension is very common and this error should be rare. If that turns out to be untrue, we could take another look at this.
- Also, verify some flag usage sooner so we can exit with an error faster if you mistype a "bin/storage dump" command.
Test Plan: Read documentation, hit affected error cases, did a dump and spot-checked that it came out sane looking.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13304
Differential Revision: https://secure.phabricator.com/D20572
Summary:
See downstream <https://phabricator.wikimedia.org/T902>. Currently, timezones are rendered with their raw internal names (like `America/Los_Angeles`) which include underscores.
Replacing underscores with spaces is a more human-readable (and perhaps meaningfully better for things like screen readers, although this is pure speculation).
There's some vague argument against this, like "administrators may need to set a raw internal value in `phabricator.timezone` and this could mislead them", but we already give a pretty good error message if you do this and could improve hinting if necessary.
Test Plan: Viewed timezone list in {nav Settings} and the timezone "reconcile" dialog, saw a more-readable "Los Angeles".
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20559
Summary:
Ref T13294. An install is interested in a way to easily answer audit-focused questions like "what edits were made to any Herald rule in Q1 2019?".
We can answer this kind of question with a more granular version of feed that focuses on being exhaustive rather than being human-readable.
This starts a rough version of it and deals with the two major tricky pieces: transactions are in a lot of different tables; and paging across them is not trivial.
To solve "lots of tables", we just query every table. There's a little bit of sleight-of-hand to get this working, but nothing too awful.
To solve "paging is hard", we order by "<dateCreated, phid>". The "phid" part of this order doesn't have much meaning, but it lets us put every transaction in a single, stable, global order and identify a place in that ordering given only one transaction PHID.
Test Plan: {F6463076}
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13294
Differential Revision: https://secure.phabricator.com/D20531
Summary:
See PHI985. I think we pretty much need to start applying language-specific rules, but we can apply at least one more relatively language-agnostic rule: don't match lines which are indented 3+ levels.
In C++, we may have symbols like this:
```
class X {
public:
int m() { ... }
}
```
..but I believe no mainstream language puts symbol definitions 3+ levels deep.
Also clean up some of the tab handling very slightly.
Test Plan: Tests pass, looked at some C++ code and got slightly better (but still not great) matches.
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20479
Summary:
Depends on D20380. Ref T8093. When prototypes are enabled, inject a (hopefully?) no-op proxy into the Git wire protocol.
This proxy decodes "git upload-pack" and allows the list of references to be rewritten, in a similar way to how we already proxy the Subversion protocol to rewrite URIs and proxy the Mercurial protocol to distinguish between read and write operations.
The piece we care about comes at the beginning, and looks like this:
```
<frame-length><ref-hash> <ref-name>\0<server-capabilities>\n
<frame-length><ref-hash> <ref-name>\n
<frame-length><ref-hash> <ref-name>\n
...
<0000>
```
We can add, remove, or modify this section to make it appear that the server has different refs than the refs that exist on disk.
Things I have tried:
- `git ls-remote`
- `git ls-remote` where the server hides some refs.
- `git fetch` where the fetch is a no-op.
Things I have not tried:
- `git fetch` where the fetch is not a no-op.
- Tricking things into doing protocol v2. Or: I tried this, I wasn't successful. In v2, additional "\0" tricks are used to hide data in the capabilities, I think?
- `git ls-remote` where we rewrite/hide the first ref in the list, and need to move the capabilities frame elsewhere.
- `git ls-remote` where the server has no refs at all, or we remove every ref.
So the "interesting" piece of this works, but it almost certainly needs some cleanup to survive interaction with the real world.
Test Plan: See above.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T8093
Differential Revision: https://secure.phabricator.com/D20381
Summary:
Ref T8093. Support dumping the protocol bytes to a side channel logfile, as a precursor to parsing the protocol and rewriting protocol frames to virtualize refs.
The protocol itself is mostly ASCII text so the raw protocol bytes are pretty comprehensible.
Test Plan:
{F6363221}
{F6363222}
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T8093
Differential Revision: https://secure.phabricator.com/D20380
Summary:
Depends on D20411. Ref T13272. Dashboards and panels have new indexes (Ferret and usage edges) that need a rebuild.
For large datasets like commits we have the "activity" flow in T11932, but realistically these rebuilds won't take more than a few minutes on any realistic install so we should be able to just queue them up as migrations.
Let migrations insert a job to basically run `bin/search index --type SomeObjectType`, then do that for dashboards and panels.
(I'll do Herald rules in a followup too, but I want to tweak one indexing thing there.)
Test Plan: Ran the migration, ran `bin/phd debug task`, saw everything get indexed with no manual intervention.
Reviewers: amckinley
Reviewed By: amckinley
Subscribers: PHID-OPKG-gm6ozazyms6q6i22gyam
Maniphest Tasks: T13272
Differential Revision: https://secure.phabricator.com/D20412
Summary:
See PHI1182. Ref T13266. The recent fixes didn't quite cover the case where you have a query, but order by something other than relevance, and page forward.
Refine the tests around building/selecting these columns and paging values a little bit to be more specific about what they care about.
Test Plan:
Executed queries, then went to "Next Page", for:
- query text, non-relevance order.
- query text, relevance order.
- no query text, non-relevance order.
- no query text, relevance order.
Also, made an API call similar to the one in PHI1182.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13266
Differential Revision: https://secure.phabricator.com/D20354
Summary: See PHI1180. Currently, when we failover to a replica, we may not log the failure. Failovers are serious business and bad news, so emit a log even if we are able to connect to the replica.
Test Plan:
Configured a bogus master and a good replica:
```
$ ./bin/mail list-outbound
[2019-03-29 16:26:09] PHLOG: 'Retrying (attempt 1) after connection failure ("AphrontConnectionQueryException", #2002): Attempt to connect to root@127.0.0.2 failed with error #2002: Operation timed out.' at [/Users/epriestley/dev/core/lib/libphutil/src/aphront/storage/connection/mysql/AphrontBaseMySQLDatabaseConnection.php:124]
[2019-03-29 16:26:19] PHLOG: 'Retrying (attempt 2) after connection failure ("AphrontConnectionQueryException", #2002): Attempt to connect to root@127.0.0.2 failed with error #2002: Operation timed out.' at [/Users/epriestley/dev/core/lib/libphutil/src/aphront/storage/connection/mysql/AphrontBaseMySQLDatabaseConnection.php:124]
[2019-03-29 16:26:29] EXCEPTION: (PhutilProxyException) Failed to connect to master database ("local_config"), failing over into read-only mode. {>} (AphrontConnectionQueryException) Attempt to connect to root@127.0.0.2 failed with error #2002: Operation timed out. at [<phutil>/src/aphront/storage/connection/mysql/AphrontBaseMySQLDatabaseConnection.php:362]
<...snip backtrace...>
3945 Voided email rP04f9e72cbd10: Don't subscribe bots implicitly when they act on objects, or when they are…
3946 Voided email rPdf53d72e794c: Allow "Move Tasks to Column..." to prompt for MFA
3947 Voided email rP492b03628f19: Fix a typo in Drydock "Land" operations
3948 Voided email rPb469a5134ddd: Allow "SMTP" and "Sendmail" mailers to have "Message-ID" behavior configured in…
3949 Voided email rPa6fd8f04792d: When performing complex edits, pause sub-editors before they publish to…
...
```
Configured a bogus master and a bogus replica:
```
$ ./bin/mail list-outbound
[2019-03-29 16:26:57] PHLOG: 'Retrying (attempt 1) after connection failure ("AphrontConnectionQueryException", #2002): Attempt to connect to root@127.0.0.2 failed with error #2002: Operation timed out.' at [/Users/epriestley/dev/core/lib/libphutil/src/aphront/storage/connection/mysql/AphrontBaseMySQLDatabaseConnection.php:124]
[2019-03-29 16:27:07] PHLOG: 'Retrying (attempt 2) after connection failure ("AphrontConnectionQueryException", #2002): Attempt to connect to root@127.0.0.2 failed with error #2002: Operation timed out.' at [/Users/epriestley/dev/core/lib/libphutil/src/aphront/storage/connection/mysql/AphrontBaseMySQLDatabaseConnection.php:124]
[2019-03-29 16:27:27] PHLOG: 'Retrying (attempt 1) after connection failure ("AphrontConnectionQueryException", #2002): Attempt to connect to root@127.0.0.3 failed with error #2002: Operation timed out.' at [/Users/epriestley/dev/core/lib/libphutil/src/aphront/storage/connection/mysql/AphrontBaseMySQLDatabaseConnection.php:124]
[2019-03-29 16:27:37] PHLOG: 'Retrying (attempt 2) after connection failure ("AphrontConnectionQueryException", #2002): Attempt to connect to root@127.0.0.3 failed with error #2002: Operation timed out.' at [/Users/epriestley/dev/core/lib/libphutil/src/aphront/storage/connection/mysql/AphrontBaseMySQLDatabaseConnection.php:124]
[2019-03-29 16:27:47] EXCEPTION: (PhabricatorClusterStrandedException) Unable to establish a connection to any database host (while trying "local_config"). All masters and replicas are completely unreachable.
AphrontConnectionQueryException: Attempt to connect to root@127.0.0.2 failed with error #2002: Operation timed out. at [<phabricator>/src/infrastructure/storage/lisk/PhabricatorLiskDAO.php:177]
<...snip backtrace...>
```
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20351
Summary:
Ref T13091. The Ferret "rank" column is a function of the query text and looks something like `SELECT ..., 2 + 2 AS rank, ...`.
You can't apply conditions to this kind of dynamic column with a WHERE clause: you get a slightly unhelpful error like "column rank unknown in where clause". You must use HAVING:
```
mysql> SELECT 2 + 2 AS x WHERE x = 4;
ERROR 1054 (42S22): Unknown column 'x' in 'where clause'
mysql> SELECT 2 + 2 AS x HAVING x = 4;
+---+
| x |
+---+
| 4 |
+---+
1 row in set (0.00 sec)
```
Add a flag to paging column definitions to let them specify that they must be applied with HAVING, then apply the whole paging clause with HAVING if any column requires HAVING.
Test Plan:
- In Maniphest, ran a fulltext search matching more than 100 results, ordered by "Relevance", then clicked "Next Page".
- Before patch: query with `... WHERE rank > 123 OR ...` caused MySQL error because `rank` is not a WHERE-able column.
- After patch: query builds as `... HAVING rank > 123 OR ...`, pages properly, no MySQL error.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13091
Differential Revision: https://secure.phabricator.com/D20298
Summary:
See downstream <https://phabricator.wikimedia.org/T210482>.
On mobile, the task graph can take up most of the screen. Hide it on devices. Keep it on the standalone view if you're really dedicated and willing to rotate your phone or whatever to see the lines.
Test Plan: Dragged window real narrow, saw graph hide.
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20313
Ref T13266. We never page these queries, and previously never reached the
"nextPage()" method. The call order changed recently and this method is now
reachable. For now, just no-op it rather than throwing.
Summary:
Ref T13091. In Differential, if you provide a query and "Sort by: Relevance", we build a query like this:
```
((SELECT revision.* FROM ... ORDER BY rank) UNION ALL (SELECT revision.* FROM ... ORDER BY rank)) ORDER BY rank
```
The internal "ORDER BY rank" is technically redundant (probably?), but doesn't hurt anything, and makes construction easier.
The problem is that the outer "ORDER BY rank" at the end, which attempts to order the results of the two parts of the UNION, can't actually order them, since `rank` wasn't selected.
(The column isn't actually "rank", which //is// selected -- it's the document modified/created subcolumns, which are not.)
To fix this, actually select the fulltext columns into the result set.
Test Plan:
- Ran a non-empty fulltext query in Differential with "Bucket: Required Action" selected so the UNION construction fired.
- Ran normal queries in Maniphest and global search.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13091
Differential Revision: https://secure.phabricator.com/D20297
Summary:
Ref T13091. If you "Order By: Relevance" but don't actually specify a query, we currently raise a bare exception.
This operation is sort of silly/pointless, but it seems like it's probably best to just return the results for the other constraints in the fallback order (usually, by ID). Alternatively, we could raise a non-bare exception here ("You need to provide a fulltext query to order by relevance.")
Test Plan: Queried tasks by relevance with no actual query text.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13091
Differential Revision: https://secure.phabricator.com/D20296
Summary:
Ref T13259. Currently, visiting a page that executes a query with an invalid cursor raises a bare exception that escapes to top level.
Catch this a little sooner and tailor the page a bit.
Test Plan: Visited `/maniphest/?after=335234234223`, saw a nicer exception page.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13259
Differential Revision: https://secure.phabricator.com/D20295
Summary:
Ref T13259. Currently, queries set a flag and return a partial result set when they overheat. This is mostly okay:
- It's very unusual for queries to overheat if they don't have a real viewer.
- Overheating is rare in general.
- In most cases where queries can overheat, the context is a SearchEngine UI, which handles this properly.
In T13259, we hit a case where a query with an omnipotent viewer can overheat: if you have more than 1,000 consecutive commits in the database with invalid `repositoryID` values, we'll overheat and bail out. This is pretty bad, since we don't process everything.
Change this beahvior:
- Throw by default, so this stuff doesn't slip through the cracks.
- Handle the SearchEngine case explicitly ("it's okay to overheat, we'll handle it").
- Make `QueryIterator` disable overheating behavior: if we're iterating over all objects, we want to hit the whole table even if most of it is garbage.
There are some cases where this might cause new exception behavior that we don't necessarily want. For example, in Owners, each package shows "recent commits in this package". If you can't see the first 1,000 recent commits, you'd previously get a slow page with no results. Now you'll probably get an exception.
If these crop up, I think the best approach for now is to deal with them on a case-by-case basis and see how far we get. In the "Owners" case, it might be good to query by repositories you can see first, then query by commits in the package in those repositories. That should give us a better outcome than any generic behavior we could implement.
Test Plan:
- Added 100000 to all repositoryID values for commits on my local install.
- Before making changes, ran `bin/repository rebuild-identities --all --trace`. Saw the script process 1,000 rows and exit silently.
- Applied the first part ("throw by default") and ran `bin/repository rebuild-identities`. Saw the script process 1,000 rows, then raise an exception.
- Applied the second part ("disable for queryiterator") and ran the script again. Saw the script process all 15,000 rows without issues (although none are valid and none actually load).
- Viewed Diffusion, saw appropriate NUX / "overheated" UIs.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13259
Differential Revision: https://secure.phabricator.com/D20294
Summary:
Depends on D20292. Ref T13259. This converts the rest of the `getPagingValueMap()` callsites to operate on internal cursors instead.
These are pretty one-off for the most part, so I'll annotate them inline.
Test Plan:
- Grouped tasks by project, sorted by title, paged through them, saw consistent outcomes.
- Queried edges with "edge.search", paged through them using the "after" cursor.
- Poked around the other stuff without catching any brokenness.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13259
Differential Revision: https://secure.phabricator.com/D20293
Summary:
Depends on D20291. Ref T13259. Move all the simple cases (where paging depends only on the partial object and does not depend on keys) to a simple wrapper.
This leaves a smaller set of more complex cases where we care about external data or which keys were requested that I'll convert in followups.
Test Plan: Poked at things, but a lot of stuff is still broken until everything is converted.
Reviewers: amckinley
Reviewed By: amckinley
Subscribers: PHID-OPKG-gm6ozazyms6q6i22gyam
Maniphest Tasks: T13259
Differential Revision: https://secure.phabricator.com/D20292
Summary:
Ref T13259.
(NOTE) This is "infrastructure/guts" only and breaks some stuff in Query subclasses. I'll fix that stuff in a followup, it's just going to be a larger diff that's mostly mechanical.
When a user clicks "Next Page" on a tasks view and gets `?after=100`, we want to show them the next 100 //visible// tasks. It's possible that tasks 1-100 are visible, but tasks 101-788 are not, and the next visible task is 789.
We load task ID `100` first, to make sure they can actually see it: you aren't allowed to page based on objects you can't see. If we let you, you could use "order=title&after=100", plus creative retitling of tasks, to discover the title of task 100: create tasks named "A", "B", etc., and see which one is returned first "after" task 100. If it's "D", you know task 100 must start with "C".
Assume the user can see task 100. We run a query like `id > 100` to get the next 100 tasks.
However, it's possible that few (or none) of these tasks can be seen. If the next visible task is 789, none of the tasks in the next page of results will survive policy filtering.
So, for queries after the initial query, we need to be able to page based on tasks that the user can not see: we want to be able to issue `id > 100`, then `id > 200`, and so on, until we overheat or find a page of results (if 789-889 are visible, we'll make it there before overheating).
Currently, we do this in a not-so-great way:
- We pass the external cursor (`100`) directly to the subquery.
- We query for that object using `getPagingViewer()`, which is a piece of magic that returns the real viewer on the first page and the omnipotent viewer on the 2nd..nth page. This is very sketchy.
- The subquery builds paging values based on that object (`array('id' => 100)`).
- We turn the last result from the subquery back into an external cursor (`200`) and save it for the next time.
Note that the last step happens BEFORE policy (and other) filtering.
The problems with this are:
- The phantom-schrodinger's-omnipotent-viewer thing isn't explicity bad, but it's sketchy and generally not good. It feels like it could easily lead to a mistake or bug eventually.
- We issue an extra query each time we page results, to convert the external cursor back into a map (`100`, `200`, `300`, etc).
- In T13259, there's a new problem: this only works if the object is filtered out for policy reasons and the omnipotent viewer can still see it. It doesn't work if the object is filtered for some other reason.
To expand on the third point: in T13259, we hit a case where 100+ consecutive objects are broken (they point to a nonexistent `repositoryID`). These objects get filtered unconditionally. It doesn't matter if the viewer is omnipotent or not.
In that case: we set the next external cursor from the raw results (e.g., `200`). Then we try to load it (using the omnipotent viewer) to turn it into a map of values for paging. This fails because the object isn't loadable, even as the omnipotent viewer.
---
To fix this stuff, the new approach steps back a little bit. Primarily, I'm separating "external cursors" from "internal cursors".
An "External Cursor" is a string that we can pass in `?after=X` URIs. It generally identifies an object which the user can see.
An "Internal Cursor" is a raw result from `loadPage()`, i.e. before policy filtering. Usually, (but not always) this is a `LiskDAO` object that doesn't have anything attached yet and hasn't been policy filtered.
We now do this, broadly:
- Convert the external cursor to an internal cursor.
- Execute the query using internal cursors.
- If necessary, convert the last visible result back into an external cursor at the very end.
This fixes all the problems:
- Sketchy Omnipotent Viewer: We no longer ever use an omnipotent viewer. (We pick cursors out of the result set earlier, instead.)
- Too Many Queries: We only issue one query at the beginning, when going from "external" to "internal". This query is generally unavoidable since we need to make sure the viewer can see the object and that it's a real / legitimate object. We no longer have to query an extra time for each page.
- Total Failure on Invalid Objects: we now page directly with objects out of `loadPage()`, before any filtering, so we can page over invisible or invalid objects without issues.
This change switches us over to internal/external cursors, and makes simple cases (ID-based ordering) work correctly. It doesn't work for complex cases yet since subclasses don't know how to get paging values out of an internal cursor yet. I'll update those in a followup.
Test Plan: For now, poked around a bit. Some stuff is broken, but normal ID-based lists load correctly and page properly. See next diff for a more detailed test plan.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13259
Differential Revision: https://secure.phabricator.com/D20291
Summary:
See PHI1063. See PHI1114. Ref T13253. Currently, you can't `bin/worker execute` an archived task and can't `bin/worker retry` a successful task.
Although it's good not to do these things by default (particularly, retrying a successful task will double its effects), there are plenty of cases where you want to re-run something for testing/development/debugging and don't care that the effect will repeat (you're in a dev environment, the effect doesn't matter, etc).
Test Plan: Ran `bin/worker execute/retry` against archived/successful tasks. Got prompted to add more flags, then got re-execution.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13253
Differential Revision: https://secure.phabricator.com/D20246
Summary:
See <https://discourse.phabricator-community.org/t/unhandled-exception-muting-completed-bulk-jobs/2449>. Bulk Jobs have an "edge" table but currently do not support edge transactions. Add support.
This stops "Mute Notifications" from fataling.
The action probably doesn't do what the reporting user expects (it stops edits to the job object from sending notifications; it does not stop the edits the job performs from sending notifications) but I think this change puts us in a better place no matter what, even if we eventually clarify or remove this behavior.
Test Plan: Clicked "Mute Notifications" on a bulk job, got an effect instead of a fatal.
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20226
Summary:
See <https://discourse.phabricator-community.org/t/traceback-rendering-task-query-in-dashboard/2450/>.
It looks like this blames to D19126, which added some more complex constraint logic but overlooked "range" constraints, which are handled separately.
Test Plan:
- Added a custom "date" field to Maniphest with `"search": true`.
- Executed a range query against the field.
Then:
- Before: Warnings about undefined indexes in the log.
- After: No such warnings.
Reviewers: amckinley
Reviewed By: amckinley
Subscribers: jbrownEP
Differential Revision: https://secure.phabricator.com/D20225
Summary:
Ref T5401. Depends on D20201. Add timestamps to worker tasks to track task creation, and pass that through to archive tasks. This lets us measure the total time the task spent in the queue, not just the duration it was actually running.
Also displays this information in the daemon status console; see screenshot: {F6225726}
Test Plan:
Stopped daemons, ran `bin/search index --all --background` to create lots of tasks, restarted daemons, observed expected values for `dateCreated` and `epochArchived` in the archive worker table.
Also tested the changes to `unarchiveTask` by forcing a search task to permanently fail and then `bin/worker retry`ing it.
Reviewers: epriestley
Reviewed By: epriestley
Subscribers: Korvin, epriestley, PHID-OPKG-gm6ozazyms6q6i22gyam
Maniphest Tasks: T5401
Differential Revision: https://secure.phabricator.com/D20200
Summary:
Depends on D20196. See PHI985. When empty, the "moved/copied" gutter currently renders with the same background color as the rest of the line. This can be misleading because it makes code look more indented than it is, especially if you're unfamiliar with the tool:
{F6225179}
If we remove this misleading coloration, we get a white gap. This is more clear, but looks a little odd:
{F6225181}
Instead, give this gutter a subtle background fill in all casses, to make it more clear that it's a separate gutter region, not a part of the text diff:
{F6225183}
Test Plan: See screenshots. Copied text from a diff, added/removed inlines, etc.
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D20197
Summary:
Ref T13161. Ref T12822. See PHI870. Long ago, the web was simple. You could leave your doors unlocked, you knew all your neighbors, crime hadn't been invented yet, and `<th>3</th>` was a perfectly fine way to render a line number cell containing the number "3".
But times have changed!
- In PHI870, this isn't good for screenreaders. We can't do much about this, so switch to `<td>`.
- In D19349 / T13105 and elsewhere, this `::after { content: attr(data-n); }` approach seems like the least bad general-purpose approach for preventing line numbers from being copied. Although Differential needs even more magic beyond this in the two-up view, this is likely good enough for the one-up view, and is consistent with other views (paste, harbormaster logs, general source display) where this technique is sufficient on its own.
The chance this breaks //something// is pretty much 100%, but we've got a week to figure out what it breaks. I couldn't find any issues immediately.
Test Plan:
- Created, edited, deleted inlines in 1-up and 2-up views.
- Replied, keyboard-navigated, keyboard-replied, drag-selected, poked and prodded everything.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13161, T12822
Differential Revision: https://secure.phabricator.com/D20188
