Summary:
`interface`s cannot contain `abstract` methods. This construct will cause a PHP fatal error:
```
Access type for interface method SomeInterface::someMethod() must be omitted
```
Test Plan: Added test cases.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14562
Summary:
`interface` methods cannot contain a body. This construct will cause a fatal error:
```
PHP Fatal error: Interface function X::Y() cannot contain body in /home/josh/workspace/github.com/phacility/arcanist/test.php on line 4
```
Test Plan: Added unit tests.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14561
Summary: A user reported to me that `arc lint` was failing with an error message along the lines of "argument 1 passed to `idx` must be of type array, bool given". I suspect that the user is running an older version of PHP, which means that `array_combine(array(), array())` is returning `false` instead of the expected `array()`. Instead, avoid calling `array_combine` on an empty array.
Test Plan: I don't have PHP 5.3 easily accessible to test this.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley
Differential Revision: https://secure.phabricator.com/D14567
Summary: Return values within constructors are acceptable if they are within a closure or anonymous function.
Test Plan: Added a test case.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14564
Summary:
`abstract` methods cannot contain a body.
```
PHP Fatal error: Abstract function X::Y() cannot contain body in /home/josh/workspace/github.com/phacility/arcanist/test.php on line 4
```
Test Plan: Added unit tests.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14560
Summary:
A class containing `abstract` methods must itself be marked as `abstract`.
```
PHP Fatal error: Class X contains 1 abstract method and must therefore be declared abstract or implement the remaining methods (X::Y) in /home/josh/workspace/github.com/phacility/arcanist/test.php on line 5
```
Test Plan: Added unit tests.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14559
Summary:
Methods cannot be marked as both `abstract` and `private`. This language construct will cause a fatal error:
```
PHP Fatal error: Abstract function X::Y() cannot be declared private in /home/josh/workspace/github.com/phacility/arcanist/test.php on line 4
```
Test Plan: Added unit tests.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14558
Summary: Add a linter rule to detect symbols which are aliased with the same name, such as `use X\Y\Z as Z`. This is unnecessary and is more-simply expressed as `use X\Y\Z`.
Test Plan: Added unit tests.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14557
Summary:
When importing or aliases a symbol with a `use` statement, the leading namespace separator is optional and does not modify the behavior. That is, `use \X` is equivalent to `use X`. As such, the latter syntax should be preferred because it is more concise.
According to the [[http://php.net/manual/en/language.namespaces.importing.php | PHP documentation]]:
> Note that for namespaced names (fully qualified namespace names containing namespace separator, such as `Foo\Bar` as opposed to global names that do not, such as `FooBar`), the leading backslash is unnecessary and not recommended, as import names must be fully qualified, and are not processed relative to the current namespace.
Test Plan: Added test cases.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley
Differential Revision: https://secure.phabricator.com/D14517
Summary:
Instead of blindly assuming that "origin" is the repository that
arcanist should communicate with, use the remote that is configured
for the branch in git.
Test Plan:
Used `arc which` with a branch with no upstream, an
origin/master upstream, and an upstream/master upstream -- the last of
which is being used to create and land this diff.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: joshuaspence, Korvin
Differential Revision: https://secure.phabricator.com/D14530
Summary:
Fixes T9807. We currently run commands like this in some cases:
hg push -r master ''
From T9807, it seems that older Mercurial treated `''` in the same way it would treat no argument, while newer Mercurial does not.
Passing `''` is unusual and not intended.
Test Plan:
From T9807, @cspeckmim confirmed that running this command without the `''` works, and @jgelgens tested the patch itself.
I didn't actually run this code myself, since I don't have Mercurial 3.6.1 installed and the fix seems straightfoward.
Reviewers: chad
Reviewed By: chad
Subscribers: cspeckmim
Maniphest Tasks: T9807
Differential Revision: https://secure.phabricator.com/D14531
Summary:
Landing from a branch that directly tracks origin/master places one in
a detached HEAD state. Instead, examine if there is a local branch of
the name that we landed onto, that also tracks the upstream; if so,
switch to that.
Test Plan:
Made a branch via `git checkout -b testing origin/master`
and tried to `arc land`
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Maniphest Tasks: T9723
Differential Revision: https://secure.phabricator.com/D14420
Summary: This is failing locally with `cppcheck` version 1.69.
Test Plan: Ran `arc unit`.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14524
Summary: Fix the `PHPCompatibilityXHPASTLinterRule` after changes to the way in which #xhpast parses `use` statements. Depends on D14518.
Test Plan: Unit tests
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14519
Summary:
Although it is technically possible to return a value from a PHP constructor, it is rather odd and should be avoided. See some discussion on [[http://stackoverflow.com/q/11904255 | StackOverflow]]. Consider the following example:
```lang=php
class SomeClass {
public function __construct() {
return 'quack';
}
}
```
This doesn't work:
```lang=php, counterexample
echo new SomeClas();
```
This, strangely, does work:
```lang=php
echo id(new SomeClass())->__construct();
```
Test Plan: Added unit tests.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14516
Summary:
Add a linter rule to detect static method calls which //should// reference the `parent` class instead of a hardcoded class reference. For example, consider the following:
```lang=php, counterexample
class SomeClass extends AnotherClass {
public function someMethod() {
AnotherClass::someOtherMethod();
}
}
```
This should instead be written as:
```lang=php
class SomeClass extends AnotherClass {
public function someMethod() {
parent::someOtherMethod();
}
}
```
Test Plan: Added unit tests.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin, epriestley
Differential Revision: https://secure.phabricator.com/D14443
Summary:
Separate XHPAST linter rules in isolation from other rules and formalize the concept of a "linter rule". As the number of XHPAST linter rules grows (we currently have around 80), it becomes increasingly difficult to manage all of the test cases because `ArcanistXHPASTLinterTestCase` currently tests the entire linter (i.e. //all// linter rules) rather than testing individual rules in isolation. See D13534 for a situation in which this is painful. This is particularly bad for third party development because unit tests could break at any time depending on upstream changes.
Basically, in order to facilitate the unit testing of XHPAST linter rules in isolation, I have made the following changes:
# Added a `setRules()` method to `ArcanistXHPASTLinter`. Currently, `ArcanistXHPASTLinter` loads all `ArcanistXHPASTLinterRule` subclasses unconditionally. The `setRules()` method provides a way to override the configured linter rules.
# Formalize the concept of a "linter rule". The `ArcanistXHPASTLinterRule` class was introduced in D10541. I feel that the modularization of `ArcanistXHPASTLinter` has made the linter much more maintainable and easily testable and I intend to extend this concept to other linters, such as `ArcanistTextLinter`.
Test Plan: Ran unit tests.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14010
Summary: Adjusts `ArcanistBraceFormattingXHPASTLinterRule` after changes to the way in which XHPAST parses namespaces. Depends on D14498.
Test Plan: Unit tests are now passing.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14499
Summary: D14037 had the logic slightly wrong and unit test results are now being double rendered.
Test Plan: Ran a unit test with `arc unit -- path/to/test` and saw the results rendered only once.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14495
Summary:
csslint offers some diagnostics related to all the document without any
relevant line number.
In such case, arc lint failed, as setLine expects null or an integer,
but not an empty string.
The 'char' and 'line' attributes in XML report are now optional.
Fixes T9804.
Test Plan: test case to repro the issue added
Reviewers: chad, joshuaspence, #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: Korvin
Maniphest Tasks: T9804
Differential Revision: https://secure.phabricator.com/D14497
Summary: Ref T9131. This doesn't seem to be used... it seems like it is a relic of postponed test results.
Test Plan: N/A
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Maniphest Tasks: T9131
Differential Revision: https://secure.phabricator.com/D14487
Summary: I've been thinking about this for a while... why not just fold `ArcanistPhutilXHPASTLinter` into `ArcanistXHPASTLinter`?
Test Plan: `arc unit`
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D13867
Summary: Ref T8742. As mentioned in D13512. This still needs some work, but looks roughly how I expect it to. Mainly, I want to move the standards stuff to the linter itself rather than the linter rule. I wanted to push this out for some initial feedback though.
Test Plan: This still needs work.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Maniphest Tasks: T8742
Differential Revision: https://secure.phabricator.com/D13942
Summary:
Fixes T9661. Users can construct arbitrarily long chains from the remote, like:
(remote) origin/master -> (local) cascade-a -> (local) cascade-b -> (local) cascade-c -> (local) cascade-d
When a user lands "cascade-d" onto "origin/master", we should pull A, B and C if they aren't ahead of the remote.
If a user lands "cascade-d" onto itself, we should pull A, B, and C if they aren't ahead of the remote, then reset D to the remote.
We also find this chain if the last component of it is connected by the local branch having the same name as the remote branch (typical for "master") instead of an actual connection through tracking brnaches.
Test Plan: See comment below.
Reviewers: chad
Reviewed By: chad
Subscribers: edibiase
Maniphest Tasks: T9661
Differential Revision: https://secure.phabricator.com/D14361
Summary:
Ref T9661. I need to reuse this to fix the complex workflow described in T9661 where we need to follow multiple paths to the upstream and cascade updates across them.
Pull the logic into a separate class to make this easier and less copy/pastey.
This shouldn't change any behavior.
Test Plan: Ran `arc land --preview` from detached head, remote-tracking branch, non-tracking branch, local-tracking branch. Selection of target/remote seemed correct in all cases.
Reviewers: chad
Reviewed By: chad
Subscribers: edibiase
Maniphest Tasks: T9661
Differential Revision: https://secure.phabricator.com/D14360
Summary: Fixes T9660. The behavior for this check wasn't quite right -- we want to check the "source ref" (what we're landing) against the "target onto" (the branch we're landing it onto).
Test Plan:
- Landed from `master` (tracking origin/master). No delete.
- Landed from `feature1` (tracking local/master). Delete.
- Landed from `feature2` (tracking origin/master). Delete.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T9660
Differential Revision: https://secure.phabricator.com/D14358
Summary:
Fixes T9543. Fixes T9658. Ref T3855.
Major functional change is that you can have a sequence of branches like:
origin/master -> notmaster -> feature1
...where they track each other, but you named your local master something else. Currently, we resolve only one level of upstreams, so we try to land onto "notmaster" in this case, which is wrong.
Instead, keep resolving upstreams until we either hit a cycle, don't have another upstream to look at, or find someting in a remote. In this case we'll eventually find "origin/master" and select "origin" as the remote and "master" as the target.
Other minor changes:
- Make this selection process explicit.
- Make the help 3000x longer.
Also fix a bug where we could incorrectly try to tell Differential to update awith `--preview`.
Test Plan:
- Landed from a tag.
- Landed from a tracking branch.
- Landed from an nth-degree tracking branch.
- Tried to land from a local branch with a cycle in upstreams.
- Landed with --remote and --onto.
- Read `arc help land`.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T9658, T3855, T9543
Differential Revision: https://secure.phabricator.com/D14357
Summary:
Ref T3855. Fixes T9537. Fixes T8620. Fixes T4333.
This declares bankruptcy and replaces the entire `arc land` workflow under Git. These are the notable changes:
- (T3855) You can now land from a branch to itself.
- (T3855) We now try to restore the original state very aggressively after any failure, instead of dumping you into the middle of a mess.
- (T9537) You can now land without a local branch.
- ([not actually] T9543) We'll now ignore the local branch if it just happens to be named the same thing as the remote branch but doesn't actually track it.
- (T8620) You can now land from a detached HEAD.
- (T4333) We now preserve the author and author date of whatever you land.
This may need some followup work. In particular:
- The signal handler (that tries to put you in a better place if you ^C in the middle of things) causes ^C to work awkwardly in prompts. This might not be worth it.
- Errors/instructions on push/merge issues might need work.
- I dropped support for `--delete-remote` and `--update-with-blah-blah` because I think these flags aren't worth their complexity.
- I've simplified the update/merge algorithm a bit. It may need some complexity added back in.
- I probably missed a few things because this covers like 200 unique, creative workflows.
- Users might need more guidance on the workflows that drop them in the middle of nowhere if they manage to reach them more often than I think.
Test Plan:
- Used `arc land` to land like at least 15,000 different kinds of changes.
- Landed normally.
- Landed from a branch onto itself.
- Landed from a detached head.
- Landed nothing.
- Landed with no local branch.
- Landed onto made-up branches.
- Landed with bad targets.
- ^C'd things in the middle.
Reviewers: chad
Reviewed By: chad
Subscribers: tycho.tatitscheff
Maniphest Tasks: T3855, T4333, T8620, T9537, T9543
Differential Revision: https://secure.phabricator.com/D14356
Summary:
Fixes T9222. Two issues here:
- First, we currently continue on error. Throw instead. I just swapped us from "phutil_passthru()" to "execx()" since I don't think printing out the "pulling from remote..." status messages is very important, and this makes it easier to raise a useful exception.
- Second, if you have a dirty working copy we currently may try to do some sort of silly stuff which won't work, like prompt you to amend changes. Instead, do a slightly lower-level check and just bail.
Test Plan:
- Ran `arc upgrade` with a dirty working copy and got a tailored, useful error.
- Ran `arc upgrade` with an artificially bad `git pull` command, got a failure with a specific error message.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T9222
Differential Revision: https://secure.phabricator.com/D14317
Summary:
Linters can now use the `version` configuration value to specify the required
version of the external binary. The version number may be prefixed with <, <=,
>, >=, or = to specify the version comparison operator (default: =).
PHP's native `version_compare()` function is used to perform the comparison.
Fixes T4954.
Test Plan: Tested against a sample of external linters.
Reviewers: joshuaspence, epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin, joshuaspence
Projects: #lint
Maniphest Tasks: T4954
Differential Revision: https://secure.phabricator.com/D14298
Summary:
See discussion in D13737. If you're using this linter to match messages which //sometimes// have a character, you can get `""` (empty string) matches when the expression doesn't match. We'll complain about these later.
Instead, cast the matches the expected types.
Test Plan: @csilvers confirmed fix, see D13737.
Reviewers: chad, csilvers
Reviewed By: csilvers
Subscribers: spicyj, csilvers
Differential Revision: https://secure.phabricator.com/D14307
Summary: See D14232. That didn't actually work. It looks like this does.
Test Plan:
- Ran `git commit --author ...` on build server and saw the same failure.
- Ran `git -c ... -c ... commit ...` on build server and saw it work.
Reviewers: chad
Reviewed By: chad
Differential Revision: https://secure.phabricator.com/D14233
Summary:
On `sbuild`, we currently get a failure on this test. Use an explicit `--author` so we can run the test even if `user.email` and `user.name` are not set in global Git config.
```
FAIL ArcanistBundleTestCase::testGitRepository
15 EXCEPTION (CommandException): Command failed with error #128!
16 COMMAND
17 git commit -m 'Mark koan2 +x and edit it.'
18
19 STDOUT
20 (empty)
21
22 STDERR
23
24 *** Please tell me who you are.
25
26 Run
27
28 git config --global user.email "you@example.com"
29 git config --global user.name "Your Name"
30
31 to set your account's default identity.
32 Omit --global to set the identity only in this repository.
33
34 fatal: empty ident name (for <builder@sbuild001.phacility.net>) not allowed
```
Test Plan: Ran `arc unit --everything`. Will verify in production.
Reviewers: chad
Reviewed By: chad
Differential Revision: https://secure.phabricator.com/D14232
Summary:
This is in a similar vein as D14220 and sets a name on linter
messages. This should handle issues from D14165.
Test Plan: Run as many of the changed linters as possible + existing linter tests.
Reviewers: chad, #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14225
Summary: Fixes T9498
Test Plan: Run `arc lint` with errors that get caught and reported by `ArcanistClosureLinter`
Reviewers: joshuaspence, epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Maniphest Tasks: T9498
Differential Revision: https://secure.phabricator.com/D14220
Summary: Ref T5821. Basic idea here is that Harbormaster can run `arc unit --everything --target ...` to get a build target updated.
Test Plan: Ran `arc unit --target ...`, saw web UI update with test results.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T5821
Differential Revision: https://secure.phabricator.com/D14190
Summary:
Fixes T9476. Currently, if you enter no text in "arc:prompt", we abort resolution immediately.
However, this is a bit inconsistent (other rules that fail to resolve are skipped) it is reasonable to put some default rule behind "arc:prompt" so that you can just mash enter to select it. This construction is unusual, but seems fine and sensible to me.
If you're using "arc:prompt" as the last rule, the behavior is the same as before, so this should only make the rule more useful.
Test Plan:
- Ran `arc which --base 'arc:prompt, git:HEAD^'` with and without the patch.
- Without the patch, entering no text at "arc:prompt" failed immediately.
- With the patch, entering no text caused fallback to the "git:HEAD^" rule.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T9476
Differential Revision: https://secure.phabricator.com/D14187
Summary: Ref T9145. Fixes T9316. We now require "name" and "code" has a maximum length (currently, this is 32, but the next diff will raise it to 128).
Test Plan:
- Installed PHPCS.
- Hit both the "name" and "code" issues.
- Applied this patch.
- Got better errors sooner.
Reviewers: chad
Reviewed By: chad
Subscribers: aik099
Maniphest Tasks: T9145, T9316
Differential Revision: https://secure.phabricator.com/D14165
Summary:
Fixes T9455. Depends on D14136. When you have a dirty submodule:
$ nano submodule/file.c # save changes
...we currently ask you to make a commit when you run `arc diff`, which is meaningless and misleading.
Instead, prompt the user separately.
This behavior isn't perfect but I think it's about the best we can do within reason.
Test Plan:
- Ran `arc diff` in a working copy with uncommitted submodule changes only, got new prompt.
- Ran `arc diff` in a working copy with submodule base commit changes only, got old (correct) prompt.
- Ran `arc diff` in a working copy with both, got only old prompt (which is incomplete, but reasonable/meaningful).
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T9455
Differential Revision: https://secure.phabricator.com/D14137
Summary:
Fixes a minor issue from D14034. PHP doesn't like `clone null;`, and if you type a nonsense command like `arc nbrhch` (as I frequently do) we try to `clone null` here.
Instead, just `return null;` if no workflow matches. Clone otherwise.
Test Plan:
- Ran `arc nbrhcanc`
- Ran `arc branch`
Reviewers: BYK, chad
Reviewed By: chad
Differential Revision: https://secure.phabricator.com/D14112
