Summary: `PhutilProxyException` provides the capability to nest exceptions. However, if we throw a `PhutilProxyException` then we currently only display the error message from the top-most exception. Instead, we should print all of the nested exception messages.
Test Plan: Faked an error, saw multiple lines of exception messages.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9697
Summary: If JSHint encounters //too many// errors (by default, more than 50) errors, then it quits prematurely and raises an additional (rather unhelpful) "Too Many Errors" error. We should just ignore this additional error.
Test Plan: Added a test case.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9693
Summary:
The error context (a.k.a. "original text") as provided by `jshint` is not very useful and actually causes `arc lint` to display the lint message incorrectly.
{F169277}
The underlying problem here is that the error context from `jshint` contains the entire line from the input file rather than just the offending source code.
Test Plan: Ran `arc lint -- webroot/rsrc/js/core/behavior-hovercard.js` (in rP) and verified that the output looked reasonable.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9675
Summary: Fixes T5442. `arc lint --everything` currently uses a `FileFinder` to discover paths for linting. A consequence of this is that files that are ignored or otherwise excluded from version control are linted.
Test Plan: Ran `arc lint --everything` in rPHU and noticed that I wasn't prompted to add a trailing newline to `src/.phutil_module_cache`.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Maniphest Tasks: T5442
Differential Revision: https://secure.phabricator.com/D9674
Summary: Consolidate `php_extension_classes.txt` and `php_extension_functions.txt` with `php_compat_info.json`. Given that `php_extension_classes.txt` and `php_extension_functions.txt` are manually generated whereas `php_compat_info.json` is generated automatically, this should make maintenance easier.
Test Plan: Deleted the `src/.phutil_module_cache` file (in each of rARC, rPHU and rP) and ran `arc liberate`.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9662
Summary: As discussed in D9097, the `ArcanistScalaSBTLinter` isn't //really// a linter. Eventually we should add support for a proper Scala linter, but I think that supporting the `ArcanistScalaSBTLinter` as-is is more effort than it is worth.
Test Plan: N/A
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9664
Summary: This `LINT_PHP_53_FEATURES` should have been replaced with `LINT_PHP_COMPATIBILITY` in D9576.
Test Plan: Ran `arc unit`.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9655
Summary: See inline comments on D9576. Also replace `json_decode` with `phutil_json_decode`, for better error handling.
Test Plan: Make sure `arc unit` still works.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9656
Summary:
Ref T5297. Utilize `PhutilJSONParser` to create a native JSON linter (native in that it is pure PHP with no external dependencies).
Since [[https://github.com/Seldaek/jsonlint | JsonLint]] is literally a PHP port of [[https://github.com/zaach/jsonlint | jsonlint]], I figured that we should also deprecate `ArcanistJSONLintLinter`. `ArcanistJSONLinter` //should// behave identically to `ArcanistJSONLintLinter`.
Depends on D9634.
Test Plan: Ran the `ArcanistJSONLintLinter` unit tests.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, Korvin
Maniphest Tasks: T5297
Differential Revision: https://secure.phabricator.com/D9628
Summary: This was broken in D9641. If `$config` is not set, then a `PhutilJSONParserException` will be thrown. The expected behaviour is that it is okay to not explicitly set any configuration.
Test Plan: Ran `arc unit`.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9643
Summary: Depends on D9634. `phutil_json_decode` now throws an exception on invalid JSON.
Test Plan: Ran the unit tests.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9641
Summary: Various tidying up of linting code.
Test Plan: `arc lint` and `arc unit` still pass.
Reviewers: chad, epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9625
Summary: Self explanatory. Also fixed the character offset to start at 1 instead of 0 (because the other linters seem to do this).
Test Plan: `arc unit`
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9626
Summary:
In D9595, we stopped parsing short-form "Differential Revision:" fields in commit messages, and only accept URLs.
I have one of the older style commit messages in my local `arcanist/`, so now we go down this parse failure branch in `arc branch`. This has never worked quite correctly, and if the parse fails we end up with a bad branch dictionary that is missing fields.
Test Plan: Ran `arc branch` in a working copy with an old `Differential Revision:` field at the head of a branch. Before patch: explosions; after patch: works great.
Reviewers: joshuaspence
Reviewed By: joshuaspence
Subscribers: epriestley
Differential Revision: https://secure.phabricator.com/D9620
Summary:
There's some sort of race inside `git` here, where the `git diff-files` command exits with different results some of the time when run in parallel with `git ls-files` or `git diff` (running either command was sufficient to trigger the race).
Run it separately to avoid the race.
I poked around the `git` source a little bit but quickly lost interest given that the issue seems fixed and this workaround is essentially reasonable.
Test Plan: Ran test 20x in a row without failures.
Reviewers: hach-que
Reviewed By: hach-que
Subscribers: epriestley
Differential Revision: https://secure.phabricator.com/D9616
Summary: The output from `arc help --full` is missing a newline character.
Test Plan:
**Before**
```
> arc help --full
--skip-arcconfig
Skip the working copy configuration file
--arcrc-file filename
Use provided file instead of ~/.arcrc.>
```
**After**
```
> arc help --full
--skip-arcconfig
Skip the working copy configuration file
--arcrc-file filename
Use provided file instead of ~/.arcrc.
>
```
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9608
Summary:
This leads to information being lost when others do `arc patch` because the name is used as the email address.
For example:
username = Richard van Velzen
Would give:
'authorName' => null,
'authorEmail' => 'Richard van Velzen'
Test Plan:
ran it through my head a couple of times, and tested it with the common options which all gave the expected result:
'rvanvelzen@company.com',
'Richard van Velzen',
'Richard van Velzen <rvanvelzen@company.com>',
'Richard van Velzen rvanvelzen@company.com',
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9605
Summary: `PhutilConsoleTable` does a better job at aligning columns. We still probably need to do some work to `PhutilConsoleTable` to set a maximum width for a specified column, or elect which columns can be truncated etc but this can probably come later.
Test Plan: {F167687}
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9601
Summary:
Update the `.arclint` file after D9576 and D9590.
Depends on D9576, D9590.
Test Plan: Eyeball it.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9591
Summary: This has been a `TODO` for a few years now... no-one should be using the old-school (non-URL) syntax anymore.
Test Plan: I'm actually not sure if it's okay to drop support for this... please verify.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9595
Summary: Fixes T5385. Provide a flexible means of setting a minimum PHP version for the `ArcanistXHPASTLinter`, instead of relying on `ArcanistXHPASTLinter::LINT_PHP_53_FEATURES` and `ArcanistXHPASTLinter::LINT_PHP_54_FEATURES`.
Test Plan: Fixed up and ran unit tests.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Maniphest Tasks: T5385
Differential Revision: https://secure.phabricator.com/D9576
Summary: This linter rule is very specific to the Phabricator and it is unlikely that it is being used elsewhere. We should move it to `ArcanistPhutilXHPASTLinter`, where other Phabricator-specific linting code lives.
Test Plan: Moved the corresponding test case.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9590
Summary: Ref T4725.
Test Plan: add some files that have unsupported constructs all over the place and run `arc liberate`
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, Korvin
Maniphest Tasks: T4725
Differential Revision: https://secure.phabricator.com/D9585
Summary: Depends on D9586. The progress bar is written to stderr which results in bad stuff happening when you try run `lint` (`PhutilLibraryLinter` calls it)
Test Plan: run `arc lint`, see no more errors
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9587
Summary:
This is useful for wrapper scripts that want to customize arcanist's behavior without affecting the global configuration.
This can be implemented with arcanist_configuration entry in .arcconfig, however it is currently limited to
per-project settings, and this feature makes writing wrapper scripts a little easier.
Test Plan: arc diff --set-config editor=vim (yeah yeah, crappy test case)
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9442
Summary: .. Instead of dots that are hard to count.
Test Plan: remove `.phutil_module_cache` and do `arc liberate`
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9584
Summary: Add a linter which wraps around [[https://github.com/mdevils/node-jscs | JSCS]].
Test Plan: Currently, we can't test this linter with out existing infrastructure. Specifically, `jscs` will only lint `*.js` files. I have done a fair bit of manual testing though.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9565
Summary: This method has been deprecated for a while now (see rARC0d212ccf5a7ac5fc75d2d3036cb3a0bee220fd02). It should be safe to remove.
Test Plan: N/A
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9466
Summary: I can't find an analogous flag for hg.
Test Plan: I don't have any precommit hooks set up on my machine so I can't test this easily, but I think this seems pretty harmless?
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9523
Summary: Allows the `ArcanistXHPASTLinter` to determine whether constants can be used, based on the target PHP version. Also added class methods to the compatibility information, although this isn't used yet (it is small anyway).
Test Plan: Created a test file that contained the `JSON_PRETTY_PRINT` constant. Verified that a linter error was raised.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9571
Summary:
Fixes T5377. The current `scripts/update_compat_info.php` script works for PHP CompatInfo version 2, but doesn't work with the newer version 3.
There are a few breaking changes in version 3 that had to be addressed:
- PHP 5.3 is required. Whilst Arcanist is generally compatible with PHP 5.2, I don't think that having this dependency presents any real issues because it is purely a development tool that is rarely updated.
- [[https://getcomposer.org/ | Composer]] is used for packaging, which makes including the library slightly more complicated. Basically, I had to install `PHP_CompatInfo` globally with Composer (`composer global require "bartlett/php-compatinfo"` and then symlink `~/.composer/vendor` into `externals/includes`.
Test Plan: Compared the `resources/php_compat_info.json` file. There are a bunch of functions/classes that //were// in this file but are no longer, but I think that I've covered the most popular extensions.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, Korvin
Maniphest Tasks: T5377
Differential Revision: https://secure.phabricator.com/D9568
Summary: Just some minor tidying up. I don't think that these methods should ever be overridden in subclasses.
Test Plan: N/A
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9569
Summary: See D9557. Add a unit test to ensure that `src/__phutil_library_map__.php` is up-to-date.
Test Plan: Ran `arc unit`.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9558
Summary: As per the TODO comment, now that `.arclint` is more mature, we should be able to restore PEP8 errors.
Test Plan: N/A
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9473
Summary:
This `ArcanistLesscLinter::getCacheVersion` should have been renamed to `getVersion` as of D8971, but wasn't updated.
Also fix the regex to properly capture version information.
Test Plan: It's pretty difficult to test this stuff.... I have an idea that I will submit a separate diff for.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9475
Summary:
Introducing `--head` caused us to run `git diff base..head` explicitly.
However, we can now hit this workflow:
- We resolve `HEAD` as commit `aaaa1`.
- This is cached.
- We notice dirty working copy changes and prompt the user to amend them to HEAD.
- The user accepts the amend.
- We amend, creating commit `bbbb2`.
- We dirty the commit range and reload the working copy. This //does not// dirty the cache of HEAD.
- We run `git diff`, but it uses the old cached HEAD: `git diff base..aaaa1`.
- This works fine (`aaaa1` still exists, it's just not on any branch) but produces the wrong diff (without amended changes).
To resolve this, implement the "dirty the cache when the range reloads" hook.
Also never try to amend if the user provides `--head`.
Test Plan:
Ran `arc diff --only --trace` in a working copy with a new commit and some uncommitted changes.
- Prior to this change, saw a `git diff base..aaaa1` command and the wrong diff.
- After this change, saw a `git diff base..bbbb2` command and the correct diff.
Reviewers: chad, csilvers, talshiri
Reviewed By: talshiri
Subscribers: epriestley, spicyj
Differential Revision: https://secure.phabricator.com/D9506
Summary:
This adds support for passing range of commits for arc diff. This is useful when you want to submit code reviews for past commits without mucking around with the working copy.
This will probably require changes :)
Test Plan: Tested locally, but totally need to add tests for this
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9369
Summary: This function has been deprecated for a long time (see rARC8150fdf044818c503a588c1b3ef0ddbb93cfa1be). It should be safe to remove it now.
Test Plan: N/A
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9465
Summary: Currently, class and interface names are converted to lowercase when comparing minimum and maximum versions. This is necessary in order to compare with the data from `php_compat_info.json` but causes the lint message to be slightly misleading.
Test Plan:
Linted a test file.
**Before**
```
Error (XHP31) Use Of PHP 5.3 Features
This codebase targets PHP 5.2.3, but `iterator` was not introduced until
PHP 5.3.0.
1 <?php
2
>>> 3 final class FooBar implements Iterator {}
```
**After**
```
Error (XHP31) Use Of PHP 5.3 Features
This codebase targets PHP 5.2.3, but `Iterator` was not introduced until
PHP 5.3.0.
1 <?php
2
>>> 3 final class FooBar implements Iterator {}
```
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9471
Summary: This workflow has been deprecated for a long time now. At some stage, it should be removed.
Test Plan: N/A
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9463
Summary: Fixes T5300. Currently, if a dead symbolic link is linted, all kinds of errors will be thrown by most linters because they will try to read the (non-existent) file contents. Instead, let's not lint symbolic links by default. In the case that the target of a symbolic link is inside the working copy, then it should be being linted anyway.
Test Plan: Created a symbolic link and verified that it wasn't linted (by any linter other than the `ArcanistFilenameLinter`).
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Maniphest Tasks: T5300
Differential Revision: https://secure.phabricator.com/D9448