Summary:
After `arc diff` creates a revision in Phabricator it amends the commit to include a link to the revision in the commit message. For Mercurial this is done with `hg commit --amend --logfile` however this will fail when trying to create a diff for a non-head commit.
This updates `ArcanistMercurialAPI::amendCommit()` to allow amending a non-head commit in two ways, depending on whether `evolve` is in use:
No evolve:
1. Rebasing the current commit onto the current commit's parent, using the new commit message
2. Rebasing all children + descendants of the current commit onto the new resulting commit
3. Stripping the original commit
With evolve:
1. Amend the commit with `hg amend --logfile`
2. Run `hg evolve` to tidy up all commits
Test Plan:
I created 6 commits in a row placing a bookmark at commits 2 `bookmark1`, 4 `bookmark2`, and 6 `bookmark3`, and ensured I had `arc:bookmark` in my base ruleset.
No evolve, non-head changeset:
1. I verified I did not have `evolve` enabled by running `hg debugextensions` and did not see `evolve` in the listed active extensions.
2. I updated to `bookmark1` and modified a file to leave a dirty working state.
3. I ran `arc diff` and when prompted to amend my changes I said "yes", and verified a phab revision was created properly.
4. I checked the status of my repository and verified it was still linear and the bookmarks pointed to the proper commits.
5. I ran `hg log -r bookmark1 --template {desc}` to view the full commit message and verified it contained both `Summary: ...` and `Differential Revision: https://...`.
6. I ran `hg diff -c bookmark1` and verified the changes for that commit included the changes I made in step 2.
No evolve, head changeset:
1. I updated to `bookmark3` which is the head commit and modified a file to leave a dirty working state.
2. I ran `arc diff` and when prompted to amend my changes I said "yes", and verified a phab revision was created properly.
3. I checked the status of my repository and verified it was still linear and all the bookmarks pointed to the proper commits.
4. I ran `hg log -r bookmark3 --template {desc}` to view the full commit message and verified it contained both `Summary: ...` and `Differential Revision: https://...`.
5. I ran `hg diff -c bookmar3` and verified the changes for that commit included the changes I made in step 1.
With evolve:
1. I enabled `evolve` and verified it was enabled by running `hg debugextensions` and saw `evolve` in the listed active extensions.
2. I updated to `bookmark2` and modified a file to leave a dirty working state.
3. I ran `arc diff` and when prompted to amend my changes I said "yes", and verified a phab revision was created properly.
4. I checked the status of my repository and verified it was still linear and all the bookmarks pointed to the proper commits.
5. I ran `hg log -r bookmark2 --template {desc}` to view the full commit message and verfieid it contained both `Summary: ...` and `Differential Revision: https://...`.
6. I ran `hg diff -c bookmark2` and verified the changes for that commit included the changes I made in step 2.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: Korvin, epriestley
Differential Revision: https://secure.phabricator.com/D21686
Summary:
Ref PHI1808. In Mercurial, we must save and restore bookmark state explicitly.
- Save and restore bookmarks.
- Clean up concepts in "arc-ls-markers" slightly, so we don't need separate "isCurrent" and "isActive" flags, hopefully.
I believe the totality of Mercurial state is:
- A (non-bare) working copy points at exactly one commit (which might be the empty/null commit, in an empty repository).
- A working copy has exactly one active branch.
- Each branch has zero or more heads.
- Each head may be closed.
- Each (non-null) commit belongs to exactly one branch.
- Note that the active branch may have zero heads and zero commits which belong to it!
- A working copy has zero or one active bookmark.
To capture this, we now emit:
- A list of branch heads. If a branch head is a working copy commit, that head is flagged as active.
- A list of bookmarks. If a bookmark is the current bookmark, that bookmark is flagged as active.
- A single "branch-state" virtual marker. This covers the case where you have run "hg branch X" to create X, but no objects in the working copy actually correspond to X yet. It also covers the case where you are on a concrete branch, but not any head of that branch.
- A single "commit-state" virtual marker. This always shows the current commit in the working copy.
Test Plan:
- Useful states to test are:
- Empty repository (not all commands currently work here).
- Normal repository, on a bookmark.
- Normal repository, no bookmark.
- "hg up 123" to update to somewhere in history.
- "hg branch X", to start a new branch with no commits.
- Ran "arc branches" and "arc bookmarks" in various states. Saw generally sensible output.
- Ran "arc land --hold ..." in various states against a failing remote. Saw generally sensible output, and saw working properly restored to the original state.
Differential Revision: https://secure.phabricator.com/D21396
Summary:
Ref PHI1805. Under some combination of versions (Python 3.8?), "arc-ls-markers" is running into additional Python runtime issues.
Sprinkle more "b" around to resolve them? Also clean up a couple of plain "arc" issues.
Test Plan:
Landed a change in Mercurial.
Some of this works fine without changes in Python 3.7/2.7 against Mercurial 4.7/5.4, so this may not be exhaustive.
Differential Revision: https://secure.phabricator.com/D21393
Summary:
Ref T13546. See PHI1805. Currently, the "arc-ls-markers" extension doesn't run under Python 3:
- some stuff needs "b'...'" to mark it as a byte string;
- "dict.iteritems()" is gone in Python 3, and "mercurial.pycompat" isn't always available;
- in Python 3, "json" refuses to print byte strings; and
- the compiler caching behavior in Python 3 has changed.
Try to get these things working in the same way under Python 2 and Python 3.
Test Plan:
Ran this command (with `python` as Python 2, locally):
```
$ python /usr/local/bin/hg --config 'extensions.arc-hg=/Users/epriestley/dev/core/lib/arcanist/support/hg/arc-hg.py' arc-ls-markers --
```
...and this command:
```
$ python3 /usr/local/bin/hg --config 'extensions.arc-hg=/Users/epriestley/dev/core/lib/arcanist/support/hg/arc-hg.py' arc-ls-markers --
```
..and saw the same output in both cases (previously, `python3 ...` fataled in various ways).
Maniphest Tasks: T13546
Differential Revision: https://secure.phabricator.com/D21392
Summary:
Ref T9948. Ref T13546. To identify remote branch heads -- not just modified heads -- use "branchmap" like "hg outgoing" does.
I wasn't able to find any other way to get what we want: for example, with a bundlerepo, "peer.heads()" and "peer.changelog.heads()" include local branches which are not present in the remote.
It generally seems difficult (perhaps impossible?) to distinguish between these cases by using "getremotechanges()":
- branch X exists at position Y in both the local and remote;
- branch X exists at positino Y in the local, but not the remote.
In any case, this seems to work properly and //should// do less work than "getremotechanges()" since we don't need to pull as much data over the wire.
Test Plan: Ran "hg arc-ls-markers" in various repositories, got what appeared to be a faithful representation of the remote branch and bookmark state.
Maniphest Tasks: T13546, T9948
Differential Revision: https://secure.phabricator.com/D21349
Summary:
Ref T9948. Ref T13546. This change moves toward a functional "arc land" in Mercurial.
Because of how "bundlerepo.getremotechanges()" works, "hg arc-ls-markers" does not actually list markers in the remote that aren't different from local markers so it's hard to get anywhere with this.
Test Plan: Got somewhat-encouraging output from "arc land" and "hg arc-ls-markers", but too many things are still broken for this to really work yet.
Maniphest Tasks: T13546, T9948
Differential Revision: https://secure.phabricator.com/D21348
Summary: Ref T13546. Ref T9948. It seems challenging to examine a remote in vanilla Mercurial. Provide an "hg arc-ls-remote" command which functions like "git ls-remote" so we can figure out if "--into X" is a bookmark, branch, both, neither, or a branch with multiple heads without mutating the working copy as a side effect.
Test Plan: Ran various "arc land --into ..." commands in a Mercurial working copy, saw apparently-sensible resolution of remote marker names.
Maniphest Tasks: T13546, T9948
Differential Revision: https://secure.phabricator.com/D21343
