topic: improve the revset used to return name->nodes mapping The 'topic' version already filters public changeset out, and skipping 'changectx' creation will helps performance.

topic: update the changectx method to respect phases The topic is no longer returned once the changeset is no longer mutable (unless explicitly requested).

topic: avoid crash when topic is loaded but not enabled for a repository

compat: update test to match core default state This output changed in 560ceb654180.

topicmap: massive rework Massively rework the way we build and use topicmap. This bring massive performance benefit. Topic map use to be a fully independant thing that we would switch on and off globaly. The caching on disk was broken so the performance were atrocious. Intead, now the topic are inherited from the 'immutable' map. We gave up on storing them on disk for now since the mutable set is usually small enough. The activation is done by hacking new "filter" on the repository and detection when they are one. This is hacky but core is hard to wrap here. Overall this whole wrapping is really scary and we should massage core API to help it.

topic: use the 'topic' revset in namespace This revset performance just got improved.

topic-revset: update the revset to no longer build changectx Atribute access is expensive, and changectx are very expensive to build, so we skip these for better performance. Before: ! wall 0.012195 comb 0.020000 user 0.020000 sys 0.000000 (best of 217) After: ! wall 0.008816 comb 0.010000 user 0.010000 sys 0.000000 (best of 303) (Before this changeset parent) ! wall 0.213261 comb 0.210000 user 0.210000 sys 0.000000 (best of 45)

topic-revset: changectx creation in the revset The current mutability filter is very expensive, we use a more direct and efficent way to do so: Before: ! wall 0.213261 comb 0.210000 user 0.210000 sys 0.000000 (best of 45) After: ! wall 0.012195 comb 0.020000 user 0.020000 sys 0.000000 (best of 217)

topics: add some noise to rewrittent changeset to prevent hash cycle If we have a changeset with topic `x`, we change it's topic to `y`, fine. When we change it's topic back again to `x`, we get the hash of the the obsoleted changeset which had the topic `x` initially. The same happens for few more cases like clearing the topic of a changeset which initially had no topic. This approach is influenced from cmdutil.amend and other commands (rebase, histedit, etc…)

topic-change: update the working copy along when changing topic of '.' This avoids leaving the working copy behind, removing another large parts of issue5441.

topic-change: cleanup the locking and transaction mechanism The previous code did not garanted we would release all lock (and was a bit more complicated than needed.

topics: use a dict for successors of changesets produced by topic change This patch adds a dictionary to store the successors of changesets which got obsoleted when we change it's topic. This is an improvement in the fix for issue5441.

topic: further simplify the clear logic We just set 'topic' to None and the rest should follow.

topics: drop the clean argument from _changetopics() After this commit, None will represent that we want to clean the topic.

topics: rename '--change' flag to '--rev' flag The --change flag was used to read a revset of which topic should be changed. In mercurial we use --rev for this type of flag. Now things will work as follows: `hg topics topicname`: It will set the current topic `hg topics topicname --rev revset`: Change topic of all the revs in the revset to topicname Further patches will try to achieve the following: `hg topics --clear --rev revset`: clear topic from all the revisions in the revset `hg topics --clear`: clear the current topic (Current behavior) `hg topics --rev revset`: show topics on the revisions in the revset

topics: factor out the logic to change topic in a new function It will help to refactor and fix bugs. Moreover we can re-use the logic.

obslog: update the readme This is release not worthy

obslog: add a comment about the final new line of descriptions

obslog: also display description patch with --patch When activating the --patch option, also display the description patch.

obslog: extra patch indentation code The code is simple and we need to reuse it.

obslog: add a patch option Add support for '--patch' option in obslog for a similar effect than "hg log --patch". The patch support is only limited to a few basic cases that will be extended in the future.

obslog: clarify some sorting code I'm not sure why Boris did it, but this belong to the next patch.

obslog: pass directly unfiltered_repo _debugobshistorydisplaynode used to create two times an unfiltered repo, makes _debugobshistoryrevs create only one unfiltered repo and pass it directly.

obslog: remove useless repo arg _debugobshistorydisplaymarker used to takes repo as an argument but it isn't needed anymore.

obslog: small renaming of _debugobshistorysingle Rename _debugobshistorysingle to _debugobshistoryrevs in order to better reflect its purpose.

refactoring: directly use 'ctx.unstable()' in showstack function This is simpler and more efficient.

packaging: flag default as version 6.5.0.dev This helps to reduce confusion when installing intermediate version.

merge: get stable change back into default

obsfate: improve obsfate output Improve obsfate output: * Remove newline and replace it by '; ' * Improve the hybrid object for making join(obsfate, '|') working.

packagin: mark stable branch as 6.4.1.dev This help prevent confusion with installing intermediate version

topics: abort if user wants to show the stack of a non-existent topic

stack: show unstable state for the current revision if it is one (issue5553) Before this patch if the current revision is unstable, hg stack does not show the unstable state. This patch fixes that by showing both current and unstable state for the current revision if that's unstable.

topics: some minute fixes to the documentation which shows up in help I am not a native but these changes are simple and obvious.

topics: make sure we commit on new parents while changing topics (issue5441) While changing topics of a set of linear commits, we used to commit our new changesets with new topic on parents of its predecessor i.e. changeset before the topic change. If the topic of parent was also changed, that parent will become obsolete and hence resulting the cnew commit in unstable state. For a set of linear commits this repeats and we end up in a tree state as mentioned in the bug. This patch fixes the bug by checking whether the parent was obsoleted and if yes, commit on the new parent.

tests: add a test to show issue5441 in test-topic.t The issue will be fixed in the next patch and the test will be helpful to show that.

merge back with stable (test change from stable has been reverted)

Added tag 6.4.0 for changeset e60248f26f92

packaging: prepare version 6.4.0

readme: fix readme for future 6.4.0

test-compat: merge with mercurial-3.9

test-compat: merge with mercurial-4.0

test-compat: merge with mercurial-4.1

compat-test: merge with stable

test-compat: fix previous merge The merge was a bit too heavy handed in dropping some part of the test.

merge with default Integrate new development around obsfate.

obsfate: by default only display the username if differ from the current one This will will keep the output simple in the simple case.

obsfate: display more data about direct prune markers This does not handle all cases, but this should handle the most common ones.

obsfate: split markers fetch from successor set annotation This will help to improve support for pruned changesets.

obsfate: mark successorsetallmarkers public We are about to use it in another module.

obsfate: mark 'preparesuccessorset' as public It is used in another module so it is not really private.

obsfate: use 'split' as verb when appropriate

obsfate: improve prune support in _successorsetverb In the prune case, still have markers (at least the prune markers). If we want to be able to return these markers we need the underlying function to support it.

obsfate: rename 'successorsset' to 'sset' in the loop The old name make it too close to 'successorssets', and the shorter name is not ambiguous in the context anyway.

template: use hex node in the obsmarkers used in the obsfate template This was pointed by Yuya and yuya is right.

template: use hex successors in obsfate This was spotted by Yuya.

template: use hex-node in successors AS pointed by yuya on the list, the raw data for the template should be hex and not binary.

template: update successors doc I'm not very happy about it, but at lease there is a show description now

template: fix precursors documentation

template: used hex-node for the precursors underlying data As pointed by yuya on mercurial-dev we should not use binary node here.

compat-test: merge with mercurial-3.9