revert: drop now useless conditional in the backup check Now that we removed the (hopeless) attempt to backup file we knew to be missing in the target changeset, we can stop checking if the file exists in the target changeset.

revert: no backup for `dsadded` set There is only one case where a backup is required in the `dsadded` set, and the current backup mechanism fails to handle it. So we stop trying to do backups at all for now. This will help us to simplify the backup code and finally fix this backup issue.

revert: add more padding in the dispatch list We are going to add more sets and some of them will have longer names. We add padding in a standalone patch for readbility purposes.

revert: add documentation in the dispatch table More sets are coming so documenting the existing ones will help.

revert: add a way for external extensions to prefetch file data This allow extensions that mess with the storage layer (remotefilelog, largefile) to prefetch any data that will be accessed during the revert operation. We are currently fetching more data than theoretically required because the backup code is a bit stupid. Future patches will improve that.

merge with stable

histedit: abort gracefully on --continue/--abort with no state Previously we'd print an ugly message saying that the histedit-state file doesn't exist in the repo.

merge with stable

Added signature for changeset 5dc91146f353

Added tag 3.1.1 for changeset 5dc91146f353

bookmarks: allow pushkey if new equals current Previously a bookmark pushkey would be rejected if the specified 'old' value didn't match the servers 'current' value. This change allows this situation, as long as the 'current' server value equals the 'new' pushkey value already. We are trying to write a hook that forces a server bookmark to move forward, using a changegroup hook. If the user also pushed the bookmark, they would get an error, since they computed their pushkey (old,new) pair before the server moved the bookmark. Long term, bundle2 will let us do this more smartly, but this change seems reasonable for now.

tests: improve test coverage of branch command and existing branches

localrepo: make it possible to pass multiple path elements to join and wjoin This makes join and wjoin behave in the same way as os.path.join. That is, it makes it possible to pass more than one path element to them. Note that the first path element is still required, as it was before this patch, and as is the case for os.path.join.

run-tests: make --interactive work with --view

convert: don't use multi-argument set.update Mumble mumble 2.4 mumble.

error: use docstrings, not bare strings, for error classes

osx: create dmg with installer instead of zip OS X would offer to expand the zip so the (multi file) installer inside it could be run ... but that would leave the expanded zip folder around. Instead, use a .dmg file that automatically will be mounted - that seems more common on OS X. Still, there is two levels of levels of clicking before actually launching the installer. Having a single file installer would be better ... but seems to be hard. A more feasible improvement would be some fancy layout inside the .dmg .

branchmap: pre-filter topological heads before ancestors based filtering We know that topological heads will not be ancestors of anything, so we filter them out to potentially reduce the range of the ancestors computation. On a strongly headed repo this gives humble speedup: from 0.1984 to 0.1629

branchmap: issue a single call to `ancestors` for all heads There is no reason to make multiple calls. This provides a massive speedup for repo with a lot of heads. On a strongly headed repo this gives humble speedup in simple case: from 8.1097 to 5.1051 And massive speedup in other case: from 7.8787 to 0.1984

test-ancestor: add a test for `ancestor` with ancestry within the initset I was wondering if revisions in the initial set that are still ancestors of other elements in the initial set were yielded by `changelog.ancestors`. I now have my answer (they do) and Mercurial has a new test.

bundle2: pull obsmarkers relevant to the pulled set through bundle2 We use the new `obsheads` argument to retrieve them in a bundle2 part at the same time as the changegroup.

getbundle: add `obsmarkers` argument to getbundle This argument triggers the retrieval of all markers relevant to the set of changesets defined by the nodes in `heads`.

pull: use the "cg" argument when pulling a bundle2 We use the `cg` argument to disable the generation of a changegroup part. This is useful to pull information when changesets are already in sync (phases, obsmarkers).

wireprotocol: fix 'boolean' handling The encoding and decoding code were swapped. This is now fixed.

push: only push obsmarkers relevant to the "pushed subset" We should only exchange obsolete markers related to the changesets that are being exchanged. For example, if `A'` is a successor of `A`, we do not want to push the marker if we are not exchanging `A'`. Otherwise `A` would disappear without a successor, leading to confusion for both users and the evolution mechanism. Therefore we now exchange only the markers relevant to the subset of nodes involved in the push (the nodes themselves may be already common but were selected by --rev (or the lack of --rev)). Note that all selected markers are still exchanged on each push. We do not have a discovery protocol for markers in core yet. Such discovery would save us the exchange of markers known on both side.

test-obsolete: sort the output of debugobsolete The set of relevant markers is currently unordered. Therefore the markers will be added in arbitrary order. We sort the list of markers beforehand to ensure stable output for testing.

test-obsolete: change a marker so it is relevant to the exchanged set We are going to only exchange markers relevant to the exchanged changesets. So we need to change this marker to use a known changeset as a successor instead of a precursor.

push: use bundle2 to push obsmarkers when possible

exchange: add a `buildobsmarkerpart` function We'll have to build an obsmarker part for both push and pull. So we build a function to factor out the common part.

obsolete: add a `commonversion` function This function returns the highest common version between the locally known formats and a list of remotely known formats. This is going to be useful to know what format should be used for exchange.