Summary:
Fixes T1277. The rules we use to figure out the root of the working copy get a bunch of edge cases wrong right now. A particularly troublesome one is when a user has a `/.arcconfig` or `/home/.arcconfig` or similar, which raises a completely useless and confusing error message (T1277).
Rewrite these rules to get all the edge cases correct and do reasonable things in the presence of stray `.arcconfig`. There are a bunch of comments, but basically the algorithm is:
- From the top, go down one directory at a time until we find ".svn", ".git", or ".hg".
- In Subversion, keep going down looking for ".arcconfig". In Git and Mercurial, look for ".arcconfig" only in the same directory.
- Now that we've figured out the VCS root (where the ".vcs" directory is) and the project root (where the ".arcconfig" file is, if it exists), build an identity.
This logic was also spread across three different places. Consolidate it into one and add some logging so we can figure out what's going wrong if users run into trouble.
Test Plan:
- Ran VCS (`arc list`) and non-VCS (`arc help`) commands in Git, Mercurial, and Subversions roots and subdirectories. Also ran them in non-VCS directories. Ran them with and without .arcconfig. All the outputs seemed completely reasonable.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T1277
Differential Revision: https://secure.phabricator.com/D7686
Summary:
Create a new class for them, pass instance around as need.
This looks like it's mostly working, but I'd like to replace the various `new ArcanistConfigurationManager()`
calls with something more suitable.
And maybe get a better name for ArcanistConfigurationManager ("Configuration" is already taken).
Test Plan: arc unit --everything, and then some.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley
CC: Korvin, epriestley, aran, chad
Differential Revision: https://secure.phabricator.com/D7271
Summary:
arc lint was hardcoded for git for amending commits with lint
patches. This enables the same functionality for mercurial.
Test Plan:
Made some changes that would result in a lint patch.
arc diff
Verify that the patches it produces were amended into the commit.
Verified it still works in git as well.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D5716
Summary:
- Added arc.autostash option to have this behaviour off by
default (but configurable on a per-project basis).
- Automatic stashing of changes now informs the user of how
to restore their working directory if Arcanist unexpectedly
terminates.
- Fixed an issue with finalizeWorkingCopy when the workflow
didn't require a clean working copy.
Test Plan:
Test `arc diff` when there are changes in the working
directory; by default it should tell you to stash or commit.
Turn on the arc.autostash option and try again; it should
automatically stash with a message on how to recover, and
it should restore the working directory automatically under
almost any circumstances (other than an unrecoverable error).
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D5385
Summary:
arc diff called 'hg cat' twice for every image binary in the diff.
This turns out to take 1 second per call on a large repo because mercurial
has to parse the manifest every time.
Now arc diff batches up all the files and does only two 'hg cat'
commands. This makes the cost constant relative to the number of
images being uploaded.
Test Plan:
Ran arc diff on a diff with 30 images on both git and hg.
Verified that it was fast and that the images showed up in the web ui.
Reviewers: epriestley
Reviewed By: epriestley
CC: sid0, dschleimer, bos, aran, Korvin
Differential Revision: https://secure.phabricator.com/D5144
Summary:
We assume `git` is available now, but should not. Specifically, if a user runs a working copy operation like `arc list` in an SVN working copy without `git` available, they get this error: P707
We interpret git errors very narrowly; be more liberal in how we interpret them. This assumes users working with `git` will have a functional `git`, but this seems like a reasonable assumption and lets us remove some error text matching code.
Test Plan: Changed `git` to `girt`, ran `arc list`, saw a reasonable exception. Changed back to `git`, saw git detected.
Reviewers: btrahan, vrana
Reviewed By: vrana
CC: svemir, aran
Differential Revision: https://secure.phabricator.com/D4804
Summary:
See D4049, D4096.
- Move commit range storage from Mercurial and Git APIs to the base API.
- Move caching up to the base level.
- Store symbolic name and resolved name separately, so we can re-resolve the correct commit from the symbolic name after dirtying caches.
- Rename `supportsRelativeLocalCommit()` to `supportsCommitRanges()` (old name wasn't very good, and not consistent with new terminology like the `--base` flag).
- Rename `getRelativeCommit()` and `setRelativeCommit()` to `getBaseCommit()` and `setBaseCommit()`.
- Introduce `reloadCommitRange()` and call it from `reloadWorkingCopy()`.
I think this fixes the problem in D4049, and provides a general solution for the class of problems we're running into here, with D4096. Specifically:
- We no longer get dirty caches, as long as you call reloadWorkingCopy() after changing the working copy (or call a method which calls it for you).
- We no longer get order-of-parsing-things problems, because setBaseCommit() reloads the appropriate caches.
- We no longer get nasty effects from calling `requireCleanWorkingCopy()` too early.
Test Plan: This is pretty far-reaching and hard to test. Unit tests; ran various arc commands. :/
Reviewers: vrana
Reviewed By: vrana
CC: aran
Differential Revision: https://secure.phabricator.com/D4097
Summary:
This method is used in three cases:
# For unit tests, to set the range to 'HEAD^' or '.^' in an agnostic way.
# For "amend", to set the range to the commit to be amended (also 'HEAD^' or '.^').
# For "patch" and "upgrade" so we don't fail just because there's an invalid "base" rule somewhere in the config when doing clean-working-copy tests.
For cases (1) and (2), introduce an "arc:this" rule to mean "the current commit". For case (3), remove the call; it is no longer necessary to check the commit range in order to do tests for the working copy state after D4095.
Test Plan: Ran unit tests, "arc upgrade", "arc patch", "arc amend".
Reviewers: vrana
Reviewed By: vrana
CC: aran
Differential Revision: https://secure.phabricator.com/D4096
Summary:
See discussion in D4049.
The getWorkingCopyStatus() method gets called from requireCleanWorkingCopy() in a lot of places, which triggers resolution of the base of the commit range. This is unnecessary; we do not need to examine the base commit in order to determine whether the working copy is dirty or not. This causes problems, in D4049 and elsewhere (we currently have a lot of fluff calls to setDefaultBaseCommit() in workflows which need to call requireCleanWorkingCopy() but do not ever use commit ranges, such as `arc patch`). This is mostly an artifact of SVN, where the "commit range" and "uncommitted stuff in the working copy" are always the same.
- Split the method into two status methods: getUncommittedStatus() (uncommitted stuff in the working copy, required by requireCleanWorkingCopy()) and getCommitRangeStatus() (committed stuff in the commit range).
- Lift caching out of the implementations into the base class.
- Dirty the cache after we commit.
This doesn't do anything useful on its own and creates one caching problem (`commitRangeStatusCache` is not invalidated when the commit range changes because of `setBaseCommit()` or similar) but I wanted to break things apart here. I won't land it until there's a more complete picture.
This creates a minor performance regression in git and hg (we run less stuff in parallel than previously) but all the commands should be disk-bound anyway and the regression should be minor. It prevents a larger regression in `hg` in D4049, and lets us do less work to arrive at common error states (dirty working copy). We can examine perf at the end of this change sequence.
Test Plan: Ran unit tests, various `arc` commands.
Reviewers: vrana
Reviewed By: vrana
CC: aran
Differential Revision: https://secure.phabricator.com/D4095
Summary: I plan to use it in `save_lint.php`.
Test Plan:
$api->getUnderlyingWorkingCopyRevision(); // In Git SVN repo
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D4054
Summary: `svn info` is the slowest command for discovering repository (up to 300 ms) and Subversion is probably the least used repository type with Arcanist. Let's discover it last.
Test Plan:
$ arc lint --trace
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D4039
Summary: Also delete extra newlines.
Test Plan:
$ arc diff # on top of my commit
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Maniphest Tasks: T2025
Differential Revision: https://secure.phabricator.com/D3996
Summary:
There's quite some logic in here:
- It automatically decides whether to create a new commit or amend.
- It partially respects 'default-relative-commit'.
- However if it points to a closed revision then it creates a new commit.
Resolves T2025.
Test Plan:
`arc diff` on:
- Clean committed repository.
- Dirty repository without commit since 'default-relative-commit'.
- Dirty repository with non-revision commit since 'default-relative-commit'.
- Dirty repository with revision commit since 'default-relative-commit'.
- `arc diff HEAD^` on dirty repository on top of closed revision.
Reviewers: epriestley, btrahan
Reviewed By: epriestley
CC: aran, Korvin
Maniphest Tasks: T2025
Differential Revision: https://secure.phabricator.com/D3967
Summary:
This commit doesn't change license of any file. It just makes the license implicit (inherited from LICENSE file in the root directory).
We are removing the headers for these reasons:
- It wastes space in editors, less code is visible in editor upon opening a file.
- It brings noise to diff of the first change of any file every year.
- It confuses Git file copy detection when creating small files.
- We don't have an explicit license header in other files (JS, CSS, images, documentation).
- Using license header in every file is not obligatory: http://www.apache.org/dev/apply-license.html#new.
This change is approved by Alma Chao (Lead Open Source and IP Counsel at Facebook).
Test Plan: Verified that the license survived only in unit tests and LICENSE file.
Reviewers: epriestley, btrahan, edward
Reviewed By: epriestley
CC: aran, Korvin, davidrecordon
Maniphest Tasks: T2035
Differential Revision: https://secure.phabricator.com/D3881
Summary:
- Implement "arc:amended", a base commit DSL rule which always selects HEAD (git) or `.` (hg) if it has "differential revision:" in the commit message. This is unambiguously correct in amend workflows, and can cover holes in other rules like "git:branch-unique(*)".
- Fix a bunch of Mercurial stuff:
- Our use of '.' is wrong, and based on a misunderstanding on my part of the behavior of `hg diff --rev . --rev .`, which means "ignore the second --rev flag", not ". means working directory state". As far as I know there's no explicit way to say "the working copy plus all its changes".
- The `--prune` argument to "hg log" does not support symbolic names like ".^". Use revsets instead.
- Reduce the number of times we need to run `hg branch`.
- We can safely use "." to mean "the working copy revision", and do not need to do "hg --debug id" or similar.
- Generally simplify some of the nonsense in the implementation left over from me having no idea how Mercurial works.
Test Plan:
Ran "arc which" in various scenarios in a mercurial working copy. I //think// I exercised all the changes.
Ran "arc which --base arc:amended" in hg and git working copies without "Differential Revision:" in head/. (no match) and with it (matched head/.).
Reviewers: dschleimer
Reviewed By: dschleimer
CC: Makinde, tido, phleet, aran
Maniphest Tasks: T1233
Differential Revision: https://secure.phabricator.com/D2876
Summary:
New optional mode. If you set 'base' in local, project or global config or pass '--base' to 'arc diff' or 'arc which', it switches to DSL mode.
In DSL mode, lists of rules from args, local, project and global config are resolved, in that order. Rules can manipulate the rule machine or resolve into actual commits. Provides support for some 'arc' rules (mostly machine manipulation) and 'git' rules (symbolic ref and merge-base).
Test Plan:
Ran unit tests. Also:
```$ arc which --show-base --base 'arc:prompt'
Against which commit? HEAD
HEAD
$ arc which --show-base --base 'git:HEAD'
HEAD
$ arc which --show-base --base 'git:fake'
Usage Exception: None of the rules in your 'base' configuration matched a valid commit. Adjust rules or specify which commit you want to use explicitly.
$ arc which --show-base --base 'git:origin/master'
origin/master
$ arc which --show-base --base 'git:upstream'
Usage Exception: None of the rules in your 'base' configuration matched a valid commit. Adjust rules or specify which commit you want to use explicitly.
$ arc which --show-base --base 'literal:derp'
derp
$ arc which --show-base --base 'arc:halt'
Usage Exception: None of the rules in your 'base' configuration matched a valid commit. Adjust rules or specify which commit you want to use explicitly.
$ arc set-config --local base git:origin/master
Set key 'base' = 'git:origin/master' in local config.
$ arc which --show-base
origin/master
$ arc which --show-base --base 'git:HEAD^'
HEAD^
$ arc which --show-base --base 'arc:yield, git:HEAD^'
origin/master
$ arc which --show-base --base 'arc:global, git:HEAD^'
HEAD^
$ arc which --show-base --base 'arc:global, git:merge-base(origin/master)'
3f4f8992fba8d1f142974da36a82bae900e247c0```
Reviewers: dschleimer, vrana
Reviewed By: dschleimer
CC: aran
Maniphest Tasks: T1233
Differential Revision: https://secure.phabricator.com/D2748
Summary:
We currently use the language "relative commit" or "relative local commit" to refer to the head of a commit range. I want to move toward callign this a "base commit", since I think that makes more sense. This moves us a small step in that direction.
The DSL stuff in T1233 also needs access to this stuff, so move it up to the base. This is mostly a bite-sized piece of the change in that diff.
Test Plan: Ran "arc which" in a couple of circumstances, read explanations.
Reviewers: dschleimer, vrana
Reviewed By: dschleimer
CC: aran
Maniphest Tasks: T1233
Differential Revision: https://secure.phabricator.com/D2747
Summary:
Using .arc as the scratch and per-repository configuration directory
has some unfortunate consequenses in the real world. Among other
things, people forget to .gitignore it so it gets checked in.
Test Plan:
the only thing that seems to use this is the relative commit setting
for git. This diff consists of 2 commits, one for the .gitignore and
one for everything else.
Comment out the portion of my .git/config that defines the upstream
for the branch. Run arc diff --only with HEAD^ in
.arc/default-relative-commit. See that .gitignore is not included in
the resultant diff, that .arc no longer exists, and that .git/arc
exists and has HEAD^ in .git/arc/default-relative-commit.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Maniphest Tasks: T1233
Differential Revision: https://secure.phabricator.com/D2725
Summary:
This adds basic support for mutable history for arc diff and arc amend for
mercurial. This is purely opt-in (so it shouldn't affect anyone who doesn't want
this feature) by explicitly setting
"immutable_history" : false
in arc configuration.
This also fixes another instance of weird behaviour for multiple heads - the
first instance was fixed here:
https://github.com/facebook/arcanist/pull/28
Test Plan:
without "immutable_history" turned on)>
When ##arc diff## produces an update diff, it should list only commits that are
ancestors of the current revision - not ones from other heads.
Reviewers: epriestley
CC: csilvers, aran, Koolvin
Differential Revision: https://secure.phabricator.com/D2654
