obshashrange: rework saving branching to be clearer and safer

sqlcache: protect read query too Some error (like locked database) can even happens when doing readonly operation. So we protect them too. At that point, it seems that pysqlite3 is not the right tool for this job.

sqlite: fast path when nothing to save If we have no new data to save, we should try to do anything. Doing something concurrently is hard enough. If we can avoid doing it at all that would be great.

sqlcache: initialize meta table last It turns out that pysqlite commit after each table creation. This commit is an hopeless attemps to make the caches more concurrent friendly.

obshashrange: always check in base cachekey against the recorded one Ignoring the check in the none case is just disaster waiting to happens.

stablerangecache: fix output in the drifted case The node were not hex, and the field had different order in the pair. This is now fixed.

sqlcache: also ignore integrity error integrity error can happens when multiple client tries to warm similar cache. Given we leave multiple hole in the cache that can be warmed by anyone, this is harder to avoid that we would like. We simply catch the error in this case. Someone will warm the missing entry later.

sqlcache: cache OperationError when saving This is a cache, so we should not crash when trying to operate on it. OperationError can happens when the database is seen as readonly or locked.

sqlcache: passe better connection option These new options should help with handling transaction consistency and database access on high load.

contrib: introduce a small script to stress tests obsolescence exchange Tester have been reporting some error that seems to originate from concurrent access/update to the cache. This script will be useful to reproduce these situations locally.

test-compat: merge mercurial-4.4 into mercurial-4.3

test: adjust glob in previous changeset I forgot to glob the time.

branching: merge with stable

packaging: mark as development version This avoid confusion.

Added tag 8.1.1 for changeset e7abf863e113

packaging: prepare release 8.1.1 Topic is untouched so we keep the previous version number.

branching: merge with stable

test-compat: merge mercurial-4.5 into mercurial-4.4

test-compat: merge mercurial-4.6 into mercurial-4.5

test-compat: merge stable into mercurial-4.6

changelog: add a line about improved prune documentation

changelog: reorder content

prune: more human-friendly documentation The previous help string generally had a highly technical language. I've tried my best to address that.

prune: improve help first line The previous help string for prune didn't mention successors in the first line. Note: Pierre-Yves David changed the original "hidden" to "obsolete" as it is better concept to put forward. (a bit more specific and with more consequence than just "hidden"). We need a good `hg help obsolete/obsolescence` result. We'll try to get this in the next version.

prune: try to clarify the "multi" revisions case I hope this highlight the use case for all three options.

prune: move help about biject after the one about split and fold The --biject flag seems a less common use case.

prune: no longer consider dropping --biject Field testing show it save people one wrong call often enough that it is worth keeping.

prune: more human-friendly documentation for --biject

prune: use ``-s/--succ`` in documentation Extracted by Pierre-Yves David

branching: merge with stable

compat: import the module dagutil module in evolve Discovery code is core have been rewritten to no longer depends on dagutil and the module have been removed. This break evolve's obsmarker discovery (older version of the discovery use the same concept).. A quick work around is to import the dagutil module in evolve and use that for now.

evolve: delete evolvestate after aborting interrupted `next --evolve` We didn't use to delete the evolvestate after aborting an interrupted `next --evolve`. This patch makes sure we do that. The test changes demonstrate the fix and we add a CHANGELOG entry about the fix.

tests: check that we delete evolvestate correctly after abort/stop This patch adds test to check whether we delete the evolvestate correctly after aborting or stopping an interrupted evolve which is caused because of `next --evolve`.

changelog: update with issue5927 fix

evolve: set the second parent if we get conflicts while evolving This patch makes sure we set the second parent correctly if we run into conflicts while evolving. Also, we need to make sure to reset to one parent before committing.

evolvecmd: import mercurial.node module as nodemod We use node as a variable name at few places and the conflicts with variable name are not good. Moreover this practice is followed in core mercurial also.

tests: demonstrate the interrupted evolve does not set p2 This patch adds a test to demonstrate issue5927 which is about evolving not setting p2 correctly in interruption.

cleanup: remove unused import

cleanup: remove old comment that lost all meaning

template: remove compatibility layer for `showobsolete` The keyword exist in core since 4.2

exthelper: drop compatibility layer for `command` We dropped support for 4.2.

topic: drop compat layer for `command` We dropped support for 4.2

legacy: drop compat layer for "command"

drophack: drop compat layer for "command"

compat: drop compatibility layer for updatecaches The `repo.updatecaches` method is support since Mercurial 4.2

branching: merge with stable

obshashrange: force saving of stablesort and firstmerge cache For some reason the check in repo is not working for stable sort. We also force firstmerge to be saved as test show it was missing.

test: add a test about cache warming This test reveal that multiple cache are not saved to disk after discovery (firstmerge, stablesort).

stablerange: save stablesort cache alongside the stablerange one This one does not introduce inconsistency but helps performance. G: changed hgext3rd/evolve/stablerangecache.py

obshashrange: always save stable range cache alongside the obshashrange one If the on disk date cover different area, the invalidation of affected range will misbehave.

obshashrange: clear in-memory cache alongside the sqlite one Clearing on disk data in nice, but we also need to remove in memory one or we'll get incorrect results.

branching: merge with fixes on stable

changelog: summarize the recent improvement

obshashrange: re-enabled more selective pruning of affected range

obshashrange: fix computation of affected ranges The computation of impacted nodes and associated revs is fully reworked. In addition,we introduce multiple new tests covering cases that were previous wrongly handled.

obshashrange: add more validation output to tests We are about to update the logic and add more tests.

obshashrange: correctly detect changeset directly affected by prune Before this changesets, parent of standard obsmarkers were wrongly considered affected and pruned changeset were wrongly not considered affected.

obshashrange: do not search for affected ranges above the highest we have It is a fast way to know we don't have an affected range for an affected revision.

obshashrange: do not search for affected stable range cache is unavailable Before this changeset we where resetting in all cases, and then looking for affected ranges. In addition is the stable range were not available, the functions silently returned no ranges affected. Now, do one or the other depending of the availability of the stable range cache. In practice we always do a broad reset because the code detecting affected changeset is currently buggy.

stablerange: build closure a bit less inefficiently The new way make me a bit less sad than the old one.

discovery: make sure repository wrapping happens in the right order Otherwise we may end up in situation were cache are not warmed in the right order, crashing.

branching: merge with stable

changelog: add an entry about the next --evolve fix

changelog: update 8.1.0 release date

next: fix topic constraint in the --evolve case We were using the wrong variable.

next: use set for membership testing There are no reason to use a list here.

test: adapt to core output change on unknown command CORE-TEST-OUTPUT-UPDATE: 5199c5b6fd29

compat: use new style template keyword declaration

test: fix previous changeset output The previous changeset got published prematurely.

obshashrange: add a test for `hg debugupdatecache` It seem to be properly filling the caches.

branching: merge with stable

templatekw: do not override in-core {obsolete} keyword The extension version uses old API, which will be deprecated in Mercurial 4.8.

packaging: mark version as development version This avoid confusion when installing non-tagged version.

Added tag 8.1.0 for changeset 2c5d79c6459c

packaging: prepare version

branching: merge stable into default

test-compat: merge mercurial-4.4 into mercurial-4.3

test-compat: merge mercurial-4.5 into mercurial-4.4

test-compat: merge mercurial-4.6 into mercurial-4.5

test-compat: merge stable into mercurial-4.6

compat: fix obslog compatiblity with 4.3 Due to the way of how successor sets are retrieved in new Mercurial version, retrieving it again in 4.3 way would be a performance hit and complexify the code too much. Don't retrieve markers of successor sets which makes obslog filtered output more limited in 4.3.

compat: add a fixed version of mergecopies for hg<=4.3 In hg<=4.3, copies._fullcopytracing() does not exist and we have to wrap copies.mergecopies() to fix the AssertionError coming from core.

test-compat: merge mercurial-4.5 into mercurial-4.4

test-compat: merge mercurial-4.6 into mercurial-4.5

test-compat: merge stable into mercurial-4.6

compat: make compatible call to ui.edit for hg <= 4.3 This is similar to what is already happening elsewhere.

compat: ignore extra status feature for hg <= 4.3 The feature does not exist in 4.3 and below.

format: fix indentation level Previous change got pushed without the fix by mistake.

test-compat: merge mercurial-4.6 into mercurial-4.5

test-compat: merge stable into mercurial-4.6

compat: disable empty topic warning on `hg phases` for hg <=4.5 This is annoying to support for lower version.

compat: restore compatibility with Mercurial <= 4.5 The __init__ arguments changed a lot in 4.6 and 4.7, lets do be as transparent as possible

test-compat: merge stable into mercurial-4.6

test-compat: merge stable into mercurial-4.6

branching: merge stable back into default

compat: improve detection of Mercurial 4.6 fixed case where AttributeError can be thrown when importing wireproto

cleanup: use NotImplementedError instead of NotImplemented Caught by pyflakes.

CHANGELOG: add some entries and remove duplicated entries

branching: merge default into stable The stable branch of Mercurial core now contains Mercurial 4.7 so evolve branch policy requires this merge. The @ bookmark is in the right location, so people doing clone will get to the latest release.

branching: merge with the other head

test-compat: backed out changeset 8fe8152ff071 These output change were for 4.7 only.

test-compat: backed out changeset 41fc764bf28c These output change are for mercurial-4.7 only

test-compat: backed out changeset dda5b2134b32 These change were for Mercurial 4.7 only.

test-compat: backout change to 1e103c7f7663

compat: use older API for older version This API changed in 4.7, we need to adjust.

branching: create mercurial-4.6 branch for updated test output This branch will be used to check for 4.6 compatibility.

builddependencies: extract function for getting successors The code no longer makes a difference between single- and mult-successor cases and just needs a set (or generator, really) of successors to iterate over. Let's extract that into a function.

builddependencies: share code between single- and multi-successor cases The two cases now have more similar structure ("if succ in revs: dependencies[r].add(succ)"), so let's share the code so it doesn't start drifting apart again.

builddependencies: build inverse dict from forward dict It's error-prone to keep "dependencies" and "rdependencies" in sync (we don't do it correctly when there are multiple successors or a node). It's easier to just create "rependencies" from "dependencies" after it's complete.

builddependencies: remove a use of defaultdict I don't see much reason to make "rdependencies" be a defaultdict when "dependencies" is not. It's easy to initialize each entry ourselves.

branching: merge with stable

builddependencies: don't remove found deps when divergence found (issue5946) It seems obviously wrong to not keep any dependencies for a revision that we had already found (for p1) if we run into divergence (from p2). It also happens to fix issue5946 :)

builddependencies: don't add dependency on revision outside input set This was already handled in the single-successor case, but had been missed in the multiple-successors case. Note that there seems to be a copy of builddependencies() in the topics extension. I don't use topics more than I have to, so I'll let someone else fix that code.

builddependencies: consider all divergent successors We were only considering one. In test-evolve-abort-conentdiv.t:165, the input revs were {5, 8} and dependency dict was {8: set([]), 5: set([10])}, which is a little weird (10 is not in the input set). It still worked because the callers used it (they seemed to only care if there were *any* dependencies). This patch fixes the issue by considering all successors. That means the dependency dict will be {8: set([]), 5: set([8, 10])} after this patch. The next patch will remove the 10 from that set.

tests: add test for issue5946 The test case is a copy of the one for 5833. The only difference is that the merge parents are recorded in the opposite order (and that the test is truncated because it fails).

obslog: check filtered output with folded changesets

obslog: check filtered output with splitted changesets

obslog: check filtered output with pruned changesets

obslog: check filtered output with amended changesets

tests: add a content-divergent obshistory test

tests: extract the obshistory amend then fold test case into a separate file

tests: extract the content-divergence test case into a separate file

tests: extract the obshistory fold test case into a separate file

tests: extract the obshistory lots of split test case into a separate file

tests: extract the obshistory split test case into a separate file

tests: extract the obshistory prune test case into a separate file

tests: extract the obshistory amend test case into a separate file

tests: extract obshistory setup in a separate file We are gonna soon explode the obshistory test file into several, extract the setup to reuse it as much as possible.

obslog: add a new flag to filter out non-local nodes

compat: drop compatibility hack for mercurial <4.3 We no longer support version prior 4.3.

compat: temporarily move copies fix to compat.py This patch moves a fixed version of copies._fullcopytracing() to compat.py as this was not fixed in core before release and also we need to do this anyway for the older versions of mercurial. This fix is that when base changeset is not ancestor of any of the merging commits, we don't throw an error. The phabricator differential which will fix this in core is https://phab.mercurial-scm.org/D3896. This fixes the traceback we get in previous changeset while relocating one of the content-divergent changeset.

evolve: continue conflicted relocation content-divergence This patch adds logic to continue an conflicted content-divergence resolution which was interrupted while relocating one of the divergent commits. The tests shows the there is some mishandling of copies internally which will be fixed in next patch.

topic: use self instead of repo in topicrepo methods Referring to repo here was somehow preventing it from being garbage-collected (important in hgweb, where currently every request gets a new repo).

evolvecmd: introduce fn to be wrapped by extensions resolving content-div The function added will be used by extension by wrapping it and adding logic to merge data introduced by them. For example: topic extension can wrap this function and add logic to merge topic names.

evolvecmd: move remaining logic to merge commit desc to existing utility fn This was a bit awful because we had a utility function and still we did some of merging outside of that fn. So let's move the remaining parts there.

amend: allow aborting an `amend --patch` by saving an empty file (issue5925) This patch fixes issue5925 by allow aborting a `hg amend --patch` call by saving an empty patch file. This is similar to aborting of `hg amend` when commit message is empty. Test changes demonstrate the fix.

tests: add a test to demonstrate issue5925 The issue will be fixed in next patch.

test: fix possible output change from 4.7 Change in debug timing can display these extra messages.

state: add the get() method that evolvecmd:1836 expects

rewing: prevent rewind in case of uncommitted changes rewind can update, which means merge and troubles. We might relax this limitation in the future.

rewind: default to rewinding the current stack

rewing: add the ability to rewind "from" revisions

evolve: exclude "identical" precursors from the precursors set I'm not sure it is the right thing to do, but it simply thing a lot for now. Otherwise, using 'precursors' to find rewind direction become too complex. This is going to give problematic result in case of prune. Since there are no other precursors to follow.

rewind: move revision selection into its own function This will help making the logic more advanced.

rewind: automatically rewind entire stack We now rewind the full stack, avoiding creating orphans. A `--exact` flag is added to force rewinding only the explicitly specified changesets.

rewind: add a test about rewinding top of stack For now this rewind is creating an orphan. We add the test to check the behavior in that case.

rewind: use rewinded parent when creating multiple changesets Use use the rewinded parent instead of creating orphans based on identical, but obsolete, changesets.

rewind: add a test for rewinding a fold The code already have basic support for case. However the result create orphan that will need to be improved. However, we keep this changeset simple.

rewind: add a test for rewinding a split There are no code change needed but it is worth testing on its own.

rewind: update the working copy if it gets obsoleted If the working copy parent is rewinded, follow the rewind.

rewind: add a message about obsolete changeset Not the best output, but useful to have the data. We can improve that later.

rewind: obsolete latest successors unless instructed otherwise To prevent the creations of divergence, we automatically mark the successors set of the rewinded changeset as obsoleted by the rewind result. A new flag `--as-divergence` is added to disable this behavior.

rewind: test rewinding an amended changesets We are now able to rewind a simple amended changeset. For now, its current successors has not been obsoleted and we created divergence.

rewind: add a message about the rewinded changesets This helps to understand what happened.

rewind: add a proto version of the command This first version focus on rewinding a single revision without care for the consequence.

stablesort: make sure heads are processed in sorted order changeset 72621094505f will fix some ordering issue in the `heads` revset so we stabilise the output in advance.

evolve: move the bookmarks also when updating to successor (issue5923) When we are on an obsolete node and does `hg evolve`, it updates to it's sucessor, however it does not move the bookmark to the successor. This patch adds logic to make sure we move the bookmark too. Test change demonstrates the fix.

tests: add test to demonstrate issue5923 This will be fixed in upcoming patch.

changelog: add an entry about the new `evolve.interrupted` section

evolve: make one missed error to align other error messages Previous patches missed this error to change it to the new format. Let's align this with all other messages raised by `hg evolve` command.

evolve: raise error.InterventionRequired instead of Abort when conflicts error.InterventionRequired is a better error to raise in such cases. Thanks to marmoute for noticing this.

changelog: add a changelog entry about divergence resolution We made a lots of progress there.

evolve: create resolved cset on successors of current parents While resolving content-divergence, sometimes, the parent can be obsolete and have a successor. We should consider the successors of the parents instead of obsolete parents as that will do something like orphan+content-divergence resolution both at the same time which is powerful and results in a single stack while resolving content-divergence of stacks. The test changes demonstrates that we can now successfully evolve a content divergent stack.

tests: fix a test case to not create conflicts This is the test case where we are testing about divergence resolution of stack of patches. If the stack is as follows: C1 C2 | | B1 B2 | | A1 A2 \/ base And A1-A2, B1-B2 and C1-C2 are content-divergent, and merging A1-A2 results in conflicts. Now if you resolve conflicts and continue resoling B1-B2, the conflicts of A1-A2 will still be there. We need somehow to understand the work we has done and not create conflicts again. To test the evolution functionality for now, we make sure conflicts does not occur.

evolve: introduce a resolutionparent variable and store it in evolvestate This patch introduces a resolutionparent variable which will be the parent of final changeset after resolving content-divergent changesets. Till now, the parent is always the parent of one of the divergent node, but now in upcoming cases, it can be a successor of both the parents of divergent nodes.

evolve: introduce a dirstatedance() fn to fix dirstate after parent change This patch introduces a dirstatedance() function which will be used after fixing the dirstate when we change parents using repo.dirstate.setparents(). Look at docs added for more details.

evolve: create a new commit instead of amending one of the divergents This patch changes the behavior of evolve command while resolving content-divergence to create a new commit instead of amending one of the divergent ones. In past, I have made this change, backed out this change and now today again I am doing this change, so let's dive in some history. Using cmdrewrite.amend() was never a good option as that requires hack to delete the evolvestate and also gives us less control over things. We can't make the commit on top of different parents as that of content-divergent ones. Due to all these, I first made this change to create a new commit instead of amending one. But, after few days, there was flakiness observed in the tests and turned out that we need to do some dirstate dance as repo.dirstate.setparents() does not always fix the dirstate. That flakiness was a blocker for progress at that time and we decided to switch to amend back so that we can have things working with some hacks and we can later fix the implementation part. Now, yesterday while tackling resolving content-divergence of a stack which is as follows: C1 C2 | | B1 B2 | | A1 A2 \/ base where, A1-A2, B1-B2, C1-C2 are content-divergent with each other. Now we can resolve A1-A2 very well because they have the same parent and let's say that resolution leads to A3. Now, we want to resolve B1-B2 and make the new resolution commit on top of A3 so that we can end up something like: C3 | B3 | A3 | base however, amending one of the divergent changesets, it's not possible to create a commit on a different parent like A3 here without some relocation. We should prevent relocation as that may leads to some conflicts and should change the parent before committing. So, looking ahead, we can't move with using amend as still using that we will need some relocation hacks making code ugly and prone to bad behaviors, bugs. Let's change back to creating a new commit so that we can move forward in a good way. About repo.dirstate.setparents() not setting the dirstate, I have researched yesterday night about how we can do that and found out that we can use cmdrewrite._uncommitdirstate() here. Expect upcoming patches to improve the documentation of that function. There are lot of test changes because of change in hash but there is no behavior change. The only behavior change is in test-evolve-abort-contentdiv.t which is nice because creating a new commit helps us in stripping that while aborting. We have a lot of testing of content-divergence and no behavior change gives enough confidence for making this change. I reviewed the patch carefully to make sure there is no behavior change and I suggest reviewer to do the same.

evolve: always order the revs to be evolved The current code only order the revs if we are resolving orphans and not in other case. The ordering is important when we pass --all and tackle multiple occurence of the same instability. Let's order the revs for each instability. There are enough test coverage of phase-divergence and content-divergence that I feel confident doing it because none of the test changed. The change in test-evolve-content-divergent.t demonstrates the indeed ordering helped.

tests: add test of resolution of content-divergent stacks This patch adds a test of resolution of content-divergence stacks where a series of changesets get mutated as two different series resulting in whole content-divergent stacks. As the tests shows, we are not processing the changesets in right order which needs to be fixed first.

evolve: improve error messages when conflicts occur This patch improves the error messages when conflicts occur. First, we drop the line 'evolution failed', that is not the best line we can show and evolution didn't failed, it's just interrupted by the conflicts and when user will run `hg evolve --continue`, things will be fine. I still remember when I first saw 'evolution failed', I got a bit scare as am I in a recoverable position or not. So let's drop this scary line. Second, we replace the error messages to say `resolve conflicts and see help-topic`. The help topic was added recently and documents all the three flags very well. Addition of tests also showed that all the three flags works fine with all the three instability type. So we should advertise them more. Third, we now raise the error with our error message rather than raising MergeFailure and having evolution related text in hint or stderr above. This increase the focus on the error message we want to show. After this patch, I think error messages by evolve in case of conflicts will be same in every case.

evolve: point to `hg help evolve.interrupted` while resolving content-div Recent test additions shows that --abort and --stop works well while resolving content-divergence and we should advertise them. This patch does so by pointing to evolve.interrupted help topic.

tests: add test about aborting content-divergence when multiple parents This patch adds tests for --abort flag of hg evolve command when we are resolving content-divergence which are on different parents and parent of one of them is the gca of both of them. The tests shows that we are doing good and we should advertise the flag during conflicts.

CHANGELOG: add an entry about the --update flag Since --update is still the default, we make the entry about --no-update.

evolve: show unfinished information in `hg status -v` (issue5886) `hg status -v` shows information about unfinished states or `hg status` shows it when `commands.status.verbose=True` is set. Before this patch, information about interrupted evolve does not show up because evolve does not resgiters in cmdutil.STATES. This patch make sure we register evolve related information there. There has been an idea to unify cmdutil.unfinishedstates and cmdutil.STATES which needs some work and the resulting state does looks a bit complicated. So till the time we unify them, let's fix this bug by registering in cmdutil.STATES.

style: drop a (commented) debug print Left-over from 4dcf87849f9d.

tests: update tests output with changes in d0abd7949ea3 in core hg CORE-TEST-OUTPUT-UPDATE: d0abd7949ea3 CORE-TEST-OUTPUT-UPDATE: 1e9c357d3ddf

branching: merge with stable Merging with the new 8.0.1 release.

Added tag 8.0.1 for changeset 0887c30255a1

packaging: prepare for 8.0.1 release

changelog: add a couple of missing entry

evolve: strip the extra obsmarkers if any on `hg evolve --abort` Previous patches started storing some obsmarkers which needs to stripped during `evolve --abort` which can't be stripped by stripping the new evolved changesets because they point to a node which existed before the evolution was performed. This patch adds logic to strip those obsmarkers by finding their indices in the obsstore and then stripping them.

evolve: store the obsmarkers to be deleted in evolvestate Sometimes, we don't create a new node and obsolete one of the old node in favour of an existing node. When user calls, `hg evolve --abort`, we need to strip that obsmarker to go back in the same old state. The obsmarker created for new node are stripped when the new nodes are stripped. This patch start storing such obsmarkers in the evolvestate so that we can delete them later if required. Right now we just store successsor and predecessor information, we might need a better obsmarker serialization techinque in future to make it more robust.

evolve: store obsmarkers to be deleted in evolvestate There are operations which don't create a new node rather they just obsolete one of the unstable node in favour of existing one. In such cases if a further resolution leads to conflicts and user calls `evolve --abort` or `evolve --stop`, we might need to delete that obsmarker. So let's store such obsmarkers in the evolvestate.

evolve: factor out logic to merge branches in separate function This patch moves the logic to merge branches while resolving content-divergence to a separate function. This makes code clear and better to understand.

tests: add more test of --stop while resolving content-divergence This tests the --stop flag when conflicts occur while relocating one of the divergent changesets.

evolve: strip the relocation commit on `hg evolve --stop` While resolving content-divergence on multiple parents, we relocate one of the commit on the parent of another one and then merge the divergent changesets. Merging can leads to conflicts, and if user does `hg evolve --stop`, we need to strip that relocated changeset too! This patch does that. Test changes demonstrates the fix.

evolve: declare updated and pctx variables in stopevolve() They will used in next patch.

evolve: store the old other divergent node after relocation in evolvestate When the content-divergent changes are on different parents and we relocate of the node to the parent of another one, we get a new other-divergent node. We still need to store the older other-divergent in evolvestate so that we can use that information.

evolve: factor out logic for --stop flag in separate function This will make code clean

tests: add test for --stop flag when resolving multiparents content-div This patch adds a test for --stop flag when we are in an interrupted content-divergent resolution which have multiple parents. Conflicts happen we merge the divergent changesets after relocation. The test shows that we need to strip the relocated changeset.

evolve: move logic to continue interrupted content-div resolution to new fn Each of the other type has defined their own function to continue/complete the interrupted resolution. We take out the logic for content-div resolution too in a separate function to make the top level function looks clean.

evolve: rename help topic evolve.interrupted-evolve to evolve.interrupted `evolve.interrupted-evolve` has the word evolve twice. Now that this help topic is a part of evolve command, evolve.interrupted will mean the interrupted evolve command only.

packaging: update tested-with statements in evolve and topic

CHANGELOG: mention about 4.6.1 compatibility

test-compat: merge mercurial-4.4 into mercurial-4.3

test-compat: merge mercurial-4.5 into mercurial-4.4

test-compat: merge stable into mercurial-4.5

utility: use ui.interactive() instead of checking config value Before this patch, the rev selection prompt which pops up in case of ambiguity to ask user to select a rev used to check the config value of ui.interactive which can is by default set to None. We should use ui.interactive() instead which is more smart in such cases. Thanks to martinvonz for suggesting this.

CHANGELOG: add couple of entries about things fixed We are preparing for a release!

next-prev: always respect the --merge flag An earlier bug fix series which started respecting commands.update.check introduced a bug in which --merge was not respected sometimes. This patch fixes that behavior by passing updatecheck as 'none' to hg.updatetotally() if --merge is passed which is similar to what `hg update` does.

next-prev: add tests showing --merge not respected in some cases This patch adds tests which show that --merge is not respected when you set commands.update.check to any value. This is a bug I wrote few days ago and will fix it in next changeset. I like this process of writing bugs and fixing them.

evolve: store the evolvestate if relocation results in conflicts We missed storing the evolvestate on disk when relocation lead to conflicts. We need to store it to grab information to continue or abort or stop the evolve.

tests: add more test of resolution of content-divergent changesets This patch adds test of resolution of content-divergent changesets when relocation is required and it leads to conflicts. The test shows that we are not doing the right thing. This will be fixed in upcoming patches.

evolvestate: store info about relocation while resolving content-divergence While resolving content-divergence when they are on different parents, we relocate one of the node to parent of another one and this can lead to conflicts sometimes. So we need to store information about that state in evolvestate so that we can use that when user does `hg evolve --continue` or `--stop` or `--abort`. Now that we make sure, we store the new other node in evolvestate, we fix the behavior of handling the relocated commit and merge the right changesets.

evolve: factor out logic to merge content divergent csets in separate fn This will help us in reusing the logic if we need to while doing evolve --continue.

evolve: commit the transaction if conflicts occur while merging content-div Yes, let's commit the transaction in case conflicts occur. Yes, this is what unshelve does and this is one of the reasons we don't like unshelve. Previous patches added support for resolving content-divergence when they are on different parents with parent of one being the gca. In such cases, we relocate one of the divergent commit to the parent of another one. All the relocation stuff and merging divergent changeset stuff happens in a single transaction, so if there are conflicts while merging, we abort and the transaction rollsback and our relocated commit is not applied after abort. We don't want to process the relocation because that can lead to conflicts and we will have dirty wdir because of resolving conflicts. So, we commit the transaction when merging results in conflicts to make sure if relocation happened, we commit that. This fixes the absence of relocation commit found in previous patch and uncover a new bug about handling of relocated commit. Upcoming patch will fix it.

tests: add tests for resolving content-divergence when different parents This patch adds more tests for cases when we try to resolve content-divergence when they are on different parents. The last test shows the we are somehow missing the relocated commit in interrupted transaction. This will be fixed in next patch.

evolve: implement resolution of content-divergence when on differet parents This patch implements the basic version of resolution of content-divergence changesets when they are on different parents but one of the parent is gca of both the different parents. The functionality first relocates the divergent changeset which was left behind and then resolves the content-divergence like it resolves in normal cases. This is a very basic implementation because it still does not work on interrupted evolution. Test changes in this patch shows the basic functionality working. The output of dry-run is also required to be tweaked in such cased. Upcoming patches will add more tests and will improve the implementation to work during conflicts too.

tests: add tests for resolving content-divergence with parent change This patch adds a new test file test-evolve-content-divergence.t which tests the resolution of content divergence when the divergent changesets does not have the same parents.

CHANGELOG: add an entry about fixing issue5877

evolve: pop up editor if conflicts occur while merging commit messages Last patch added support for merging commit messages while resolving content divergence. In case of conflicts we fallback to one of the divergent commit messages. After this patch, we will pop up an editor where user has to resolve the conflicts in the commit messages and then continue. Tests are added for this. We need to fix handling in test-sharing.t to use HGEDITOR to fix conflicts and pass the new commit message. Future improvement can be that we should respect HGMERGE while merging commit descriptions too.

evolve: merge commit messages while resolving content-divergence (issue5877) While resolving content-divergence, there can be divergence in commit messages too which should be resolved. This patch uses the hg's internal merge API's to merge commit messages of divergent changesets with predecessor as the base. In case of conflicts, we fallback to divergent's commit message which should be changed in future. We should pop up the editor with the commit message and the conflict markers and let user fix that. Tests are added for the functionality added.

evolve: move if-else block under the right block the index related if-else should be only executed when all three base, divergent and other have different branch names. I coded this last night and mistakenly mis-indented it. Caught when I was adding more tests.

tests: make sure we don't result in commit messages conflict Upcoming patches will introduce functionality of merging commit messages while resolving content divergence which can lead to conflicts. We don't want to test the conflicts scenario here, so let's make sure there are no conflicts of commit messages in this test.

tests: don't use mkcommit shell function in test-divergent.t Upcoming series will start merging commit descriptions while resolving content-divergence. This tests are here to test the basics of content-divegrence resolution and we should not be testing much of commit description merge here. Replacing mkcommit function with normal hg calls will help us decide on the commit messages ourselves and prevent conflicts while merging commit messages.

evolve: remove duplicated documentation File saving failure…

evolve: show `hg help evolve.interrupted-evolve` in error when conflicts The help list various flags which you can use during interrupted evolve resolution and description about what they do. This will be very helpful for user on how to continue an interrupted evolve.

evolve: add a help topic `Interrupted-Evolve` This help topic contains description on how the various flags work during interrupted evolve. This help topic will be included in error message during interrupted evolve in next patch.

branching: merge with stable

changelog: fix version number error

evolve: return the new node formed only as the replacement sometimes, the the other content divergent changeset is obsoleted in favour of the local divergent changeset in which case we should not store that as a replacement because `hg evolve --abort` will strip that changeset. Let's make sure we return the newnode only if formed. The test results shows that we are still not 100% accurate as we need to strip an obsmarker.

tests: add more test of --abort flag for content-divergence resolution This patch adds test when we resolve multiple content-divergence changesets at once and resolution fo first content-divergence does not form new commits, resolution of second one leads to conflicts. As the output shows, we are wrongly stripping a changeset here, which should be fixed in next patch.

evolve: prompt user for branch selection while resolving content-divergence When resolving content-divergence, there can be case when both the content-divergent changesets are on different named branches. In that case, we are not sure which branch should the resolution commit should be on. This patch adds logic to prompt user to choose the branch which the resolution commit should be on and then creates the new resolution commit on that branch. This also adds a new test file which have tests for the branch selection feature we have added.

evolve: pass the base revision in _completecontentdivergent() We need the base revision to merge commit messages, branches and more stuff. Let's make sure we have that by passing into the function which completes the divergence resolution.

evolve: store the base of content-divergents in evolvestate We need the base changeset for merging commit messages, branches. Let's store it in evolvestate.

evolve: use cmdrewrite.amend() instead of creating a new commit This patch backs out changeset 88601e1cd5d8. We replaced cmdrewrite.amend() with creating new commit because using amend(), a high level function was not good and creating a new commit gave us more control over things. **HOWEVER** to get more control, you need to get arrange all the pieces together perfectly. Creating a new commit, the existing logic has some bugs around dirstate handling, the `repo.dirstate.setparents()` call especially. Sometimes this fixes the dirstate correctly sometimes not. As Pierre-Yves David said, "Having a "clumsy" implementation with the right behavior is usually a good first step. So having an amend based implementation seems like a good first step.", let's take a step back and get back using amend so that we can have a correct basic implementation which is bug free.

corecompat: update the tests output to show phase update summary eb9835014d20 in core added support to transactions to show a summary on the phase changed during that transaction. Due to this, couple of tests output needs to be updated. This patch does that. While realeasing evolve, this patch should be backed out on branches where hg<=4.6. CORE-TEST-OUTPUT-UPDATE: eb9835014d20

CHANGELOG: add an entry about issue 5808 being fixed

next-prev: replace the error hint to mention about `--merge` flag

tests: add tests for conflicts while respecting commands.update.check This patch adds tests to check that we don't run into conflicts while respecting commands.update.check. The error message now does not show the hint about --merge option which we should fix.

next: respect `commands.update.check` config option in `hg next` (issue5808) `commands.update.check` config option determines what level of checking should be done when we update to another revision. When we set this config option to `noconflict`, it updates your working directory to a specified revision only when there will be no conflicts which is a very nice behavior. This patch adds logic to respect the config option so that we can have that nice behavior with `hg next` too. It turns out, we just needed to use hg.updatetotally() and add some cases to check config option. Right now, we only respect the config if value is set to `noconflict`. Otherwise we fallback to the default behavior. This is a part of fixing issue5808. Test changes demonstrate the fix.

prev: respect `commands.update.check` config option in `hg prev` (issue5808) `commands.update.check` config option determines what level of checking should be done when we update to another revision. When we set this config option to `noconflict`, it updates your working directory to a specified revision only when there will be no conflicts which is a very nice behavior. This patch adds logic to respect the config option so that we can have that nice behavior with `hg prev` too. It turns out, we just needed to use hg.updatetotally() and add some cases to check config option. Right now, we only respect the config if value is set to `noconflict`. Otherwise we fallback to the default behavior. This is a part of fixing issue5808. Test changes demonstrate the fix.

tests: add test demonstrating issue5808 This patch adds test showing that `next` and `prev` does not respect the `commands.update.check` config option, particularly when set to `noconflict`.

next: bail out early in `hg next --evolve` if dirty wdir This patch adds logic to bail out early when --evolve is passed with next command and the working directory is dirty. This will help us in preventing the message saying `do you want --merge` which is useless and misleading when `--evolve` is passed.

tests: add test of `hg next --evolve` with dirty working directory I was testing some output change to `hg next --evolve` in dirty working directory and found that we don't have any test for that.

CHANGELOG: add an entry about fixing issue5897

evolve: fix `hg abort` when hg next --evolve leads to conflicts (issue5897) Before this patch, if you do `hg next --evolve` and end up having conflicts and then you try `hg abort`, it will throw a KeyError. This patch fixes that.

tests: add a test demonstrating issue5897 When conflicts are caused due to `hg next --evolve`, `hg evolve --abort` breaks! This patch adds test to demonstrate that. This will be fixed in next patch.

evolve: add an `--update` flag to `hg evolve` command This patch adds a new `--update` flag which defaults to True. The update flag updates the working directory to the head of evolved revisions after `hg evolve` command. This flag can be used to not update your working directory after `hg evolve` which has been one of the biggest feature request from fig team at Google. To not update your working directory, the user can do `hg evolve --no-update`. If the working directory get obsoleted by evolve command, it will update to the successor. This patch also adds some basic tests for the flag introduced.

tests: add tests for testing --abort while content-divergence resolution This patch adds a new test file with a very basic test of `--abort` on an interrupted evolve which was resolving content-divergence. Upcoming patches will add more rigourous tests.

evolve: fix storage of replacements and temprevs node in evolvestate The _completephaseivergent() was not respecting rules on how to store the new node ids created in evolvestate. The rules are: * store the temporary relocated commits in temprevs * store the final new commit in replacements This patch make sure we follow the above rules and also use the return value returned by _resolvephasedivergence(). The tests change demonstrates that we are all good now.

tests: add more tests related to aborting while resolving phase-divergence The tests shows the something is still broken related to abort while resolving phase-divergence which will be fixed in upcoming patches.

topic: suggest to clear a topic that becomes empty We add a hint suggesting "hg topic --clear" when a topic becomes empty. This usually happens when all its changesets get public, that is when the user does a 'phase -p' or when they pull from a server. To discriminate relevant cases for which we should display the hint (and typically exclude explicit 'hg topic --clear' operation), we look for the transaction "type" to only consider 'phase' or 'push' transaction. In the latter case, the transaction name is not simply 'push-response' but it also included the remote URL so we need to match on string content.

topic: display a hint pointing at help when a topic becomes non-empty This is an attempt at avoiding "mistakes" by new users of topics when they pull a topic and publish its changesets while being checked out. In this case, the topic will become empty but still be active; however, if this is a "foreign" topic, the user will not likely want to keep it active for the next commit operation. So we add a hint message pointing at general help of "topics" command so that they can eventually do something and learn about "hg topics --clear".

branching: merge with compat fix on default

evolve: add compat for repair.stripbmrevset which is moved to scmutil This patch adds compatibility logic for repair.stripbmrevset() which is moved to scmutil as bookmarkrevs(). This fixes test-prune.t failure.

evolve: update the readme with more information about contribution

topic: dedent items about publishing behavior in docstring Also change item bullets from '#' to '*'.

topic: use preformatted block for config snippets documentation We add :: markers, empty line and properly indent blocks where needed.

tests: add a test for "hg help -e topic" Currently, some configuration snippets do not render well; we'll fix this in forthcoming changesets.

branching: merge with stable

content-div: add test showing `--stop` works during interrupted resolution This patch adds a test showing that the --stop flag of `hg evolve` works well during interrupted content-divergence resolution. The --stop flag now works with all the three different type of instability resolution very well and we should advertise it more.

evolve: create new commit while resolving content-divergence Before this patch, we used to amend one of the content-divergent changes with the changes from the merge. Creating a new commit will certain advantages like letting user enter a commit message, showing user the new merge diff before commiting the content-divergent resolution and give us more control in the implementation part. The reason amend was used initially was just because it was easy. There are some tests update due to change.

content-divergence: store the resolved revs in evolvestate['replacements'] The support for continuing a conflicted content-divergence resolution was added some time ago. That support lacked storing of the resolved rev in the evolvestate['replacements'] which marks the node as resolved and does not try to resolve that node again. Since we didn't store, the resolution, processes the same node again. This patch makes sure we store the information in evolvestate that a certain content-divergent node is resolved to prevent pre-processing it again.

utility: use ui.interactive() instead of checking config value Before this patch, the rev selection prompt which pops up in case of ambiguity to ask user to select a rev used to check the config value of ui.interactive which can is by default set to None. We should use ui.interactive() instead which is more smart in such cases. Thanks to martinvonz for suggesting this.

evolve: make `--abort` strip the temporary revs created during resolution During divergence resolution, `hg evolve` created temporary relocation commits which needs to be stripped off during `hg evolve --abort` to get back to the same state where we were. Previous patch added a new list temprevs to be stored in evolvestate. This patch starts storing the temp relocated node created during phase-divergence resolution in evolvestate and make sure we consider stripping it too while `hg evolve --abort`. The test changes demonstrate the fix. Now we are sucessfully aborting a phase-divergence resolution.

evolve: store a list of temprevs in evolvestate There are temporary revisions which are created as a part of divergence resolution which needs to be stripped off if user runs `hg evolve --abort`. We need to save the nodeids of those temporary revisions in the evolvestate. The temprevs list will help in the above mentioned use cases.

evolve: check for None before adding in set of revisions to strip After previous patch, replacements can contain None. Let's make sure we don't add None to the revisions to strip during `hg evolve --abort` The test changes demonstrate the usefulness of this and the previous patch. Now abort is working and public changesets are not a problem.

evolve: return the new replacement node to be stored in evolvestate While resolving phase-divergence, we can end up obsoleting the phase-divergent commit in favor of the public commit. Before this patch we return the node of public commit to store as a replacement of the phase-divergent commit. The above will not cause any problem till the time we use `hg evolve --abort` on an interrupted evolve which tries to strip the replacement nodes because it thinks that the replacements nodes are the ones which are created during the resolution and are new. Since we will be stripping a public node, `evolve --abort` will error out saying cannot strip public changeset, unable to abort evolve which is bad. We should make sure, replacements should only consist of new nodes formed. If the instablity is resolved by obsoleting in favour of old changeset, we should not store the old changeset in replacements in evolvestate.

evolve: don't make values of replacements dict lists There was inconsistency in code base related to the values stored in replacements dict in evolvestate. At some places we store a list, at some places we just store the nodeid. Let's make sure we store only the nodeid, not a list. If there are cases, when there are multiple replacements or changesets created during resolution, later patch will introduce storing temprevs in evolvestate which should be used.

tests: add a test file `--abort` flag while resolving phase-divergence This patch adds a new test file test-evolve-abort-phasediv.t which will contain tests for `hg evolve --abort` flag while resolving phase-divergence. There is a case in this test which does not work right which will be fixed in upcoming patches.

evolve: pass revisions as strings to commands.*() It's usually fine to pass integer revisions to e.g. commands.update(), but when directaccess is enabled, it attempts to parse it a revset, which fails.

tests: rename test-evolve-abort.t to test-evolve-abort-orphan.t We renamed the test file because it contains all the cases related to orphan resolution. Now we care about divergence resolution too and hence we must have separate specific test files for each type.

branching: merge with branch for stable

compat: update obshashrange test output Mercurial core bfe8ef6e370e seems to have changed the order of some messages. Simply accept the change in the test file. CORE-TEST-OUTPUT-UPDATE: bfe8ef6e370e

test: update to new outputs Mercurial core ea63a2004d09 update the hidden changeset warning during update after the effective update. CORE-TEST-OUTPUT-UPDATE: ea63a2004d09

compat: fix obslog with Mercurial 4.6+ Mercurial core 3fe1c9263024 removed the diffopts attribute on changesetprinter. Backport consuming diffopts in the constructor for old versions of changesetprinter and update the non-graph code to extract the patch option and pass it directly.

debian: add line to clean-docs target to delete additional docs file

amend: use context manager for locks (issue5887)

evolve: point to '--abort' in verbose status message f6979d64b9fb (evolve: point to '--abort' in error message, 2018-04-22) changed "hg update -C ." to "hg evolve --abort" in error messages. This commit makes the same change in the verbose status message (the one that warns about unfinished evolve). There's still one mention "hg update -C" in our codebase. That's after "hg evolve --content-divergent" fails, but I don't think "hg evolve --abort" works in that case (yet?).

evolvecmd: add comments and new lines in content-divergence handling logic I have visited this logic three times in last 10 days and I have to re-read the whole code to make sure I don't miss any part. Adding comments will help understanding the code in better and fast way.

evolve: show cset hash in "merging conetnt-divergent changeset" While I was here I also mae de it repo.ui.status() instead of repo.ui.note() to align the behavior with rest of the messages.

evolve: show cset hash in "update to local conflict" message

evolve: use hg.updaterepo() instead of hg.update() to hide stats

evolve: fix the conflicts error message during content-divergence resolution Before last patch, we didn't had an automated way to continue an interrupted content-divergence resolution and we used to show a very long error message on how to continue manually. This patch replaces that long error message with nice small message which tells how to continue and how to abort. We still show `hg update -C .` to abort as we need to test and make `--abort` work in case of content-divergence.

evolve: add logic to continue interrupted content-divergence resolution We continue the interrupted content-divergence using the logic to complete content-divergence which we refactored in it's own function few patches ago. We had to delete evolvestate file before using cmdutil.amend() because otherwise we can't use the function. This leads to a hack which deletes the evolvestate file, we may want to not to use the high level API function here. Also we had to pass evolvestate also to _completecontentdivergent function because of this. This patch also, adds a test which demonstrates that the logic works fine.

evolve: store the evolvestate file when encountering conflicts This patch makes `hg evolve` store the evolvestate file when it encounters merge conflicts while resolving content-divergence.

evolve: take out logic to complete phase-divergence in separate fn This patch takes out the logic to complete the content-divergence resolution to it's own separate function so that we can reuse it in case of interrupted resolution.

evolve: store divergent nodes in evolvestate when resolving content-divergence We need to store the divergent nodes in evolvestate so that we can continue an interrupted phase divergence resolution.

branching: merge stable into default

tests: add test showing `hg evolve --stop` works fine with phase-divergence This patch adds a new test file which contains test demonstrating that `--stop` flag works well while resolving phase-divergence using `hg evolve`.

tests: rename the test file to add 'orphan' in the name test-evolve-stop.t contains test for `hg evolve --stop` for orphan resolution only. We also need to test the flag while resolving divergence. It will be good if we have a test file for each type of instability.

tests: test that `hg obslog wdir()` aborts with a useful message

obshistory: make obslog work when a commit doesn't have any description This logic is taken from logcmdutil.changesetprinter class. It checks that commit description is not empty before extracting the first line (before this patch it would try to index an empty list). Stripping description also makes obslog behave more like regular log when there are trailing/leading whitespaces.

obshistory: use a more appropriate type for obslog entries cand is actually a regular changeset and not 'M' (graphmod.MISSINGPARENT). And 'M' is only used for graph edges, but cand is drawn as a graph node. Visually there's no effect, but the code makes more sense this way.

packaging: mark as development version This help avoiding confusion when installing from source.

branching: merge with latest release

packaking: mark as development version This avoid confusion when installing from source.

test-compat: merge mercurial-4.4 into mercurial-4.3

test-compat: merge mercurial-4.5 into mercurial-4.4

test-compat: merge stable into mercurial-4.5

Added tag 8.0.0 for changeset 116cdd8c102a

packaging: update version number

packagin: update `tested-with` statement

changelog: mention 4.6 compatibility

Changelog: pre-version gardening

testcompat: update output for Mercurial 4.3 Lets of minor update + one major update to tests/test-evolve-phase-divergence.t since all hashed chance (different instability name).

test-compat: merge mercurial-4.4 into mercurial-4.3

test-compat: merge mercurial-4.5 into mercurial-4.4

testcompat: revert output change from new capabilities This new capabilities is 4.6 only.

testcompat: revert output change from 54a469cd5b26 These are 4.6 only changes.

testcompat: revert output change from 1de6ba0ac6d7 These are 4.6 only changes.

testcompat: merge stable into mercurial-4.5

compat: handle different `ui.edit` signature with Mercurial 4.3

wirepeer: try modern module first This avoid some issue with remaining .pyc

compat: handle lack of allpredecessors for Mercurial 4.3

branching: merge further fix into stable

wrapping: fix a silly error were we checking on module and wrapping the other ...

brancing: merge next release into default We are about to release hg-evolve 8.0.0

test-compat: open mercurial-4.5 branch

evolve: point to '--abort' in error message The way to escape and evolve run has been updated and we need to tell our user about it.

compat: drop compatibility layer for pager All versions we support now support `ui.pager`.

compat: drop compatibility layer for duplicatecopies We no longer support version before Mercurial 4.3.

compat: drop compatibility layer for bookmark.applychanges We now only support Mercurial version down to 4.3.

compat: from compatibility layer for allprecursors We only support down to 4.3 now.

compat: drop compatibility layer for successorssets We support down to Mercurial 4.3 now.

compat: drop compatibility wrapper for "relevantmarkers" We no longer support version without the logic we have been adding.

compat: drop compat layer for cleanupnodes We no longer support version lower than 4.3.

compat: from compatibility layer for createmarkers We support down to 4.3 now. It supports operation.

compat: drop obsutile layer Mercurial 4.2 is no longer supported.

compat: drop vfsmod compat layer Mercurial 4.2 is no longer supported.

evolve: use in code filtered error for 4.6 Things have been upstream into core for a while, let us use them instead of the evolve version.

documentation: refer to 4.3 in the help text Minimal version supported is now 4.3. Drop reference to 4.2 in the help.

topic: remove compatibility for older version in discovery wrapping Support for 4.1 and 4.2 has been dropped.

obscache: drop compat layer to access public changeset We can now access precomputed sets. (compat to 4.1 and 4.2 dropped)

obscache: drop compat layer for obsmarkers reading We cna now use a version with the argument we needs.

template: drop compatibility layer Support has been dropped for 4.1 and 4.2

timer: drop compat layer for time Support for 4.1 and 4.2 has been dropped.

compat: drop support for Mercurial 4.2 Since Mercurial 4.6 is about to be released, we drop support for 4.2 too. (code still need to be cleaned up).

compat: close test-compat branch for mercurial-4.2 Support for this version have just been dropped.

evolve: add support for hg >= b4d85bc1 Mercurial commit b4d85bc1 deleted the wireproto module and its contents are now in various other modules, so we can not even do unchecked imports of it now.

discovery: handle deletion of 'wireproto' module Mercurial commit b4d85bc122bd. deleted the wireproto module and its contents are now in various other modules, so we can not even do unchecked imports of it now.

evolve: handle wireproto module deletion for registering new commands Module have been deleted in b4d85bc122bd. We can no longer import it blindly.

minitopic: handle wireproto module change from b4d85bc122bd This module have been scattered in other place, so we need to detect and handle this for 4.6+

topic: handle wireproto module change from b4d85bc122bd This module have been scattered in other place, so we need to detect and handle this for 4.6+

obsexchange: remove unused import of wireprotoserver This seems to be unused since 300d7d019b9.

compat: rename resolvepartialhexnodeid to resolvehexnodeidprefix The name changed in 5f8f013e7d52

compat: access datestr in a way compatible with verison prior to 4.6

compat: fallback to compatible partial lookup prior to 4.6

import: adapt to fd1dd79cff20 With the latest state of the code, the patch is passed to "import try one" as parsed.

import: grab the expected node value on the fly We no longer parse the patch beforehand this will reduce inconsistency with newer API.

import: use a less direct approach to store the expected node This will help later refactoring.

compat: adapt `amend --patch` to the new `patch.extract` API

compat: deal with change in the wireproto module

branching: merge with stable

topic: allow '.' in topic names This is also commonly used.

relocate: look up partial nodeids in unfiltered repo This accidentally changed to using the filtered repo in commit 0f0f3135.

compat: avoid using partial node ID in lookup This is getting deprecated in 4.6.

test: apply output change from protocaps CORE-TEST-OUTPUT-UPDATE: afcfdf53e4b5

test: update output to match upstream The exact wording for filtered error has been changed. CORE-TEST-OUTPUT-UPDATE: ecd3f6909184

evolve: don't pass stringified int to repo.__getitem__ We were doing repo['-1'], which is now deprecated. Switch to repo['tip'], which is much clearer anyway (I initially thought this was nullrev and sent an incorrect patch until Pierre-Yves corrected me).

patch: offer the user a chance to fix the patch if it fails to apply The experience is still not great, but better than purely dropping the changes.

amend: add a new flag `--patch` to `hg amend` This patch adds a new flag `--patch` to `hg amend` which pops up an editor with the patch of working directory parent which you can change, and when you exit the editor the patch with changes is applied to current working directory with old changeset being obsoleted in favour of new one created by the applied patch. If supplied filenames, only those filenames are present in the popped editor and rest files stay the same way in the commit as they were. The extension of the file which opens up in editor is '.diff', we cannot have it as '.patch' as there will be develwarns related to that. We need to change to patch core and undo some change to achieve this. The implementation does not use any core API rather it has picked chunks from API which are required. One main reason to not use core import API is that we have to change wdir parent before using patch.patch() which I will like to avoid to make sure we handle merge cases too. While writing this patch I have spend lot of time try to use internal API's to work for this but none of them served the purpose well. If I have time in future and work on similar problem again, I am going to write better high-level API's which uses patchstore to achieve this. A new test file test-amend-patch.t which contains a lot of testing of the feature.

stablerangecache: include units in ui.progress() calls

stablerange: include units in ui.progress() calls

tests: add test demonstarting issue5833 is fixed Previous patch which fixed issue5832 also fixed issue5833 as the root cause for both was same. This patch adds test of issue5833 as well. CHANEGLOG entry is also added for the fixed issues.

utility: convert node from MultipleSuccessorsError to rev numbers (issue5832) utility.builddependencies() deals with rev numbers but MultipleSuccessorsError returns nodeid which breaks the function and also breaks instability resolution. Previou patch demonstrate the breakage. This patch makes sure we convert those nodeids to rev numbers before processing ahead.

tests: add a test demonstarting issue5832 traceback The traceback will be fixed in upcoming patches.

CHANGELOG: add an entry about `hg evolve --abort` Previous patches added a new flag `--abort` for `hg evolve` which aborts an interrupted evolve. Currently this only works for orphan resolution cases which is very common and most people use `evolve` for that. So let's add an entry about that.

evolve: restore bookmarks correctly after `hg evolve --abort` This patch adds logic to store bookmark movements in evolvestate class and restore bookmarks to their original place if `hg evolve --abort` is called.

topic-ext: restrict the format of topic names This patch restrict the format of topic names to make sure they consist only of alphanumeric, '_' and '-' characters.

serverminitopic: wrap context.changectx.branch instead of context.changectx Recently some changes in core resulted in this wrapping of class leading to an infinite recursion. Let's wrap the particular function instead of wrapping the whole command. This fixes test-minitopic.t which was breaking.

templatekw: remove obsfatedata templatekw Individuals fields are available in core as single template functions.

evolve: do not exit 1 when there are no troubled changesets (issue5823) There are possibly other cases we also do not wish to exit non-zero for in this function, but I did not analyze them closely and am just looking at resolving issue5823.

changelog: add an entry about the memory fixes

safeguard: use self instead of repo in noautopublishrepo.checkpush() Referring to repo here was somehow preventing it from being garbage-collected (important in hgweb, where currently every request gets a new repo).

stablerangecache: use self instead of repo in stablerangerepo.stablerange() Referring to repo here was somehow preventing it from being garbage-collected (important in hgweb, where currently every request gets a new repo).

obscache: use self instead of repo in obscacherepo.updatecaches() Referring to repo here was somehow preventing it from being garbage-collected (important in hgweb, where currently every request gets a new repo).

templatekw: fix templatedata for 46

style: fix a couple of pyflakes complains

evolve: add a `--abort` flag for `hg evolve` command This patch adds a new flag `--abort` to `hg evolve` command which will abort the interrupted evolve going on undoes the changes created by the evolve command till now. The changes are bookmark movements, creation of evolved commits, obsolete the old commits in favor of evolved commits. In case when user changed things while the interrupted evolve, like did some actions which created a new commit on top of evolved commits, or changed phase of evolved commits to public, the evolve fails to abort as we should not strip out the new commit formed by user neither we can strip the public changeset. The abort fails and tell user to use `hg evolve --stop` instead. Right now bookmark movement is broken, and will be fixed in upcoming patches. Tests are added for the new flag.

evolve: don't allow updating or commiting when there is interrupted evolve We should make things simple in the starting and restrict user from doing things which can lead to problems. This is same as the rebase behavior.

evolve: internationalize the entry of --list in hg evolve help

evolve: add evolvestate as an argument to _finalizerelocate() This will help us to store bookmark changes to evolvestate to revert them in case when `hg evolve --abort` is called.

evolve: add evolvestate as an argument to evolvecmd.relocate() This will help us in passing it to _finalizerelocate() and store the bookmark changes which happend.

tests: add test about handling phase-divergent merges We don't handle phase-divergent merges and we should. Upcoming patches will add the ability to handle them.

evolve: show a message when evolution leads to no changes to commit In case of orphan resolution, evolution may result in no changes to commit. This patch adds a use message in such cases. This is same as the rebase behavior.

evolve: pass the instability category to relocate() and _finalizerelocate() This patch adds category argument to relocate() and _finalizerelocate() functions to take the type of instability which we are processing so that we perform actions accordingly.

tests: add more tests for solving phase-divergence using `hg evolve` This patch adds more tests for resolving phase-divergence caused due to fold or split using `hg evolve` command. The resolution in case of split is not very good and will be fixed in upcoming patches

evolve: handle interrupted phase-divergence resolution gracefully This patch adds logic to handle the interrupted phase-divergence resolution by `hg evolve` command gracefully. We first complete the relocation of the changeset and then do the divergence resolution. Previous functions moved required code separate fucntions so that we can use them here.

evolve: move logic to resolve phase-divergence in a new function This patch moves logic which does the revert and create a commit on top of public changeset because we need to repeat that logic while handling continuation of `hg evolve` when there are conflicts while rebasing the phase-divergent changeset.

evolve: store the precursor of phase-divergent in evolvestate This patch makes `hg evolve` store the precursor of phase-divergent changeset in evolvestate so that we don't need to calculate that again.

evolve: move logic to complete interrupted relocation to new fn This moves logic which completed the interrupted relocation to it's own separate function as we will need that function in continuing phase-divergence and content-divergence too.

evolve: move logic to complete interrupted orphan resolution to new fn This patch moves logic which completes an interrupted orphan resolution to it's own new function. We will defining such functions for completing phase-divergence and content-divergence too.

evolve: write data to evolvestate in case on conflicts in phase-divergence This patch makes phase-divergence resolution write data to evolvestate instead of graftstate. The `evolve --continue` handling of phase-divergence is broken which will be done in upcoming patches.

evolve: explicitly turn revs to list before storing into evolvestate The cbor library which we are using cannot handle encoding and decoding Python sets, and revs was set there. Let's convert revs explicitly to list to make sure we don't try writing sets.

tests: add tests of content+parent change creating phase-divergence This patch adds tests which shows handling of phase divergence cases by `hg evolve` command which involved content+parent change. The conflicts case is not handled and will be fixed in upcoming patches.

tests: show first line of commit message in test-evolve-phase-divergence.t This patch adds 'firstline' filter to description in the log template.

evolve: make sure bookmarks are moved after phase-divergence resolution This patch adds logic to move bookmarks from the rebased node to the phase-divergence resolved node.

evolve: use the new rebased changeset to resolve phase-divergence When there is a parent change in phase-divergence, `hg evolve` first rebases the diveregent changeset on parent of public changeset and then continue the resolution of phase-divergence *but* we used to continue with the old divergent changeset rather than the new rebased one. This patch makes sure we use the new rebased changeset to proceed with the resolution. Tests changes shows that it fixes things which were broken earlier.

evolvecmd: add a detailed comment and some newlines This patch adds a detailed comment on a chunk and some newlines to make things look better and more understandable for upcoming patches.

tests: add one more test case about handling of phase-divergence by evolve This patch adds a test case when phasedivergence is caused by rebase command and there is parent change involved. As the tests show, we are not handling things correctly here. Upcoming patches will fix that. Fixing things in later patches showed there was a wrong case in one another test which will be fixed by later patches, so let's add description on why that test is wrong in this patch.

tests: add more tests for handling of phase divergence by `hg evolve` This patch adds more tests related to handling of phase-divergence by `hg evolve` command.

tests: temporary delete the merge test of handling phase divergence This patch temporary deletes the merge test of handling phase divergence by `hg evolve` command to that other important tests can be plugged in before this. This will make adding more tests easy. As I say, this is temporary to ease out adding more tests, the test will be added back.

tests: various comment on test-evolve-phase-divergence

tests: make test-evolve-phase-divergence more descriptive This patch adds description to what each part is doing in test-evolve-phase-divergence.t and some commands which should be used to analyze the divergence before resolving that.

compat: use updateresult API if available The status return is no longer a tuple, we access attribute instead.

compat: provide an explicit default value to the 'odiff' config There is a devel warning about this now.

compat: keep passing arguments as keyword argument Otherwise, This break with new changes in upstream Mercurial.

compat: search for stringmatcher in the new location More of util has been split into sub modules.

rebase: update the rebase wrapping to works with hg 4.6 Mercurial core commit bd0086bd3af4 removed _makeextrafn hence breaking our wrapping. Update the code to wrap rebaseruntime directly in a way that's compatible with hg 4.1+.

compat: call current obsolete.createmarkers(), not one from load time Other extensions may wrap obsolete.createmarkers(). If they do, we should call their wrapped version. However, since f61a23a84dac (compat: add wrapper for obsolete.createmarkers() that accepts "operation" arg, 2018-03-16), we started storing the version of the function as it was when our compat module was loaded. This broke an extension we use internally at Google.

tests: rename test-evolve-bumped to test-evolve-phase-divergence.t While I was here, I added a heading to test mentioning what the test if for.

evolve: replace bumped in function name to phasedivergence bumped was the old name of phasedivergence. Lets replace with its new name so that we can forget the old one.

evolve: drop duplicate import of rewriteutil._bookmarksupdater

obslog: more 4.1 code compatibility drop

template: drop some 4.1 compatibility code Simplify all the things.

checkheads: drop upstreamed code Mercurial 4.2 has all the code that used to live in the evolve extensions so we can drop it now.

obslog: cleanup patch handling after 4.1 compat drop We can remove some of your compatibility code.

packaging: further mark 4.1 support as dropped

branching: merge with stable

tests: extend the globing to the user name Silly me.

branching: close mercurial-4.1 test branch

evolve: kill `--obsolete` and `--old-obsolete` flags for graft Evolve extension used to add the couple of flags to `hg graft` command for reasons I don't know but they were deprecated. For the functionality of `hg graft --old-obsolete`, we have `hg grab` now which can do that. One more reason to kill this is that we have logic to parse the graftstate, whereas in core, we are going to change how state files are read and write. Tweaked the tests a bit to use `hg grab` instead of these flags to make rest of the test output consistent. Replaced an entry of `hg graft -O` in doc by `hg grab`.

compat: drop support for Mercurial 4.1 This version is over 1 year old and we are dropping support for it.

packaging: mark versiond as devel versions This will help prevent confusion when installing a non-tagged version.

packaging: merge stable back into default Version have been released.

packaging: mark version as devel version This will help prevent confusion when installing non-tagged version.

Added tag 7.3.0 for changeset e4ac2e2c2086

test-compat: merge mercurial-4.2 into mercurial-4.1

test-compat: merge mercurial-4.3 into mercurial-4.2

test-compat: merge mercurial-4.4 into mercurial-4.3

test-compat: merge stable into mercurial-4.4

branching: merge future 7.3.0 into stable Release belong to stable

tests: update test output for the stable branch

debian: add version 7.3.0 to changelog

packaging: update version numbers

changelog: prepare release 7.3.0

tests: updated tested with version

__init__: show some internationalization love

evolvecmd: show some internationalization love Some outputs are not internationalized because I doubt that they are gonna stay here for long.

utility: make sure prompt text is internationalized This patch internationlize the prompt text by passing the into mercurial.i18n._

test: do not use revision number This get confused with older version.

compat: use mergetoolopts from commands.py Mercurial <= 4.2 do not have cmdutils.py yet.

branching: merge stable into default Some bugfix that we could use here (also getting ready for a release).

evolve: add a `--stop` flag for `hg evolve` command This patch adds a `--stop` flag for `hg evolve` command. The stop flag lets you stop the evolution from where you are and let you move to other things if you want. This fuctionality is great as user can decide to resolve conflicts later on, stop the command execution, do something else and get back once they has time or care about resolving them. As implementation details, it clears the evolvestate, updates to the parent of working directory clearing the uncommitted changes which also deletes the mergestate. A new test file test-evolve-stop.t is added with tests of the new flag added.

obsexchange: make compatible with some unreleased 4.5 versions We don't generally care to be compatible with unreleased versions of hg, but this is just a one-line change to be compatible with a range of commits before 158d4ecc03c8 (wireprotoserver: move all wire protocol handling logic out of hgweb, 2018-03-08), so hopefull it can be worth it.

compat: make override of createmarkers work on wrapped function If createmarkers() has been wrapped using functools.partial(), the __code__ attribute won't exist. Fix by unwrapping partial functions. I tried to find another attribute to use as proxy for the "__code__.co_argcount < 6", but there were no new functions or similar added in that commit.

previous/next: add -m short form for --merge This matches core's option to `hg update`. This patch also covers `hg next`, but test-check-commit.t only lets me mention one of them.

evolve: handle stabilization of merge changeset with both parents obsoleted This patch adds logic to stabilize an orphan merge changeset having both parents obsoleted. The logic tries to first stabilize the merge changeset on successor of second parent and then stabilize the new changeset formed on the successor of first parent. We are stabilizing on second parent first and then to second parent to preserve the first parent's successor as first parent of the merge changeset. Conflicts can occcur and we can loose processing information, therefore we store a variable in statefile `orphanmerge` which represents whether we are processing a merge changeset with both parents obsoleted. Thanks to Pierre-Yves David for suggesting this way and helping to understand the correctness of this. More rigourous test cases for this will be added in next patch.

state: add __setitem__() for cmdstate class Adding __setitem__() will help us in setting the value of variables stored in state files using `state[key] = value` in our code which is very convenient.

evolve: preserve the order of merge parents in `hg evolve --continue` This patch makes sure we preserves the order of merge parents in case of `hg evolve --continue`.

topics: include "operation" metadata in obsmarkers

grab: include "operation" metadata in obsmarkers

touch: include "operation" metadata in obsmarkers

split: include "operation" metadata in obsmarkers

prune: include "operation" metadata in obsmarkers

metaedit: include "operation" metadata in obsmarkers

fold: include "operation" metadata in obsmarkers

uncommit: include "operation" metadata in obsmarkers

evolve: include "operation" metadata in obsmarkers

amend: include "operation" metadata in obsmarkers This is for the legacy handling support for "hg commit --amend" case. I picked "amend" because it feels more natural than "commit". Besides, and more importantly, that's consistent what core uses.

compat: add wrapper for obsolete.createmarkers() that accepts "operation" arg I want to start passing the operation argument, but hg < 4.3 does not support it, so we need a compat wrapper. For older hg, the operation is simply ignored.

tests: do not glob out fixed dates Perhaps these dates historically have not been fixed, but they seem to be now. I got tired of not being able to use "run-tests.py -i".

evolve: de-duplicate some function calls _cleanup() was called in continueevolve() and evolve() function too. Let's call it at a single place for both the cases. Also call progresscb() outside of conditionals to call it both times.

evolve: move deletion of evolvestate out of continueevolve() function We need to extract some data at evolve() function from evolvestate.

evolve: load evolvestate before passing to continueevolve() This patch moves some code from continueevolve() to evolve() function so that we have a loaded evolvestate present in the evolve() function for post processing.

evolve: move the handling for non-continue part in else conditional This will help us do factor out which operations we want to do for continue, which for non-continue and which operations we want to do for both of them.

evolve: move the logic of continuation handling of `hg evolve` in own fn The function which defines the evolve command is pretty big with all the logic messed up there. It's better to have individual functions for individual uses. The movement of code will help in adding more logic to continue part in clean way. Moreover for implementing `hg continue`, we need all continuation handling in their own functions, so it will help there too.

CHANGELOG: add an entry about next and prev being interactive

prev: show changeset hash in --dry-run instead of rev numbers rev numbers can change, so it's better to show changeset hash.

prev: prompt user to choose parent in case of multiple parents This patch adds functionality to `hg prev` to prompt user to choose one parent to update when multiple parents exist. Surprisingly there were no tests for this case before this patch, I added one with this patch.

next: cleanup logic around return by returning early In the cmdnext function, there are some conditionals which returns early and some which sets a result value to be returned later. Let's return from all the conditionals early. There are if, elif and else involved, so it's sure we will end up in one of the conditional, each conditional will return, so it's guranteed that we are returning the correct value and returning everytime.

next: prompt user to choose child in ambiguity in `hg next --evolve` This patch makes `hg next --evolve` prompt user to choose a child to evolve and update when multiple of them exists. This is improvement over previous behavior where we used to error out saying ambiguous next unstable changeset and told user to manually use `hg evolve -r rev`. While writing this patch, I am personally very happy to have this feature as I have encountered this stage many times and I wanted `next --evolve` to take care of that. It uses the revselectionprompt thing added in earlier patches.

next: move logic to evolve the children to it's own function This patch moves the logic which evolves the children and updates to it in `hg next --evolve` to it's own function as we want to use that again.

next: prompt user to choose a children in case of ambiguity This patch adds functionality to prompt user and ask the children to which next should move to incase of multiple childrens. This is a very nice UI because otherwise user has to lookup changeset hash, and then manually do `hg update` to that. In this case we show hash and description so user can choose which one to update easily. It is using the interactive revision choosing functionality added in a previous patch.

next: show changeset hash in `--dry-run` instead of rev numbers Rev numbers can change, hence we should show rev hash instead of rev numbers in dry-run.

evolve: make sure we consider all cases in if-else The part of code which is touched decides on evolution of merge changeset. We need proper handling of each case there, so before adding logic, let's add conditional for each case so that upcoming patches are easy to understand.

tests: add more cases about stabilization of an orphaned merge commit This patch adds three more cases related to stabilization of orphaned merge commits.

evolve: fix missing negation Spotted by Pulkit Goyal.

evolve: clarify message for ambiguous destination from split Message is a bit long, but should be less scary.

evolve: prompt user to select destination for stabilization in case of split This patch adds functionality to prompt user to select a revision when there exists multiple stabilization destination because of split on multiple branches. Test are also added for the feature. I am not sure whether this is the best message which we can show, if people have suggestions I will like to improve the messages shown in prompt.

tests: remove some tests which are moved to dedicated files The tests in test-unstable.t files are now moved to their dedicated file with all the cases to that in that file. Let's delete them from here.

next: factor out the logic to update to changeset in another function This patch moves the logic to update to children changeset during `hg next` in a separate function because we want to re-use the logic when we will prompt user to choose a changeset in case of multiple childrens.

utility: add a function to prompt user to choose a revision This patch adds a utility function to interactively prompt user to choose a revision to proceed in cases when there are multiple revisions which are candidates. The function will be used in `hg evolve`, `hg next`, `hg prev` etc. If user feedback will be good, I plan to move it to core and use it at more possible places.

tests: add a test about evolve stabilizing orphan caused by splits There are some similar tests lying around in the test suite but they are not well arranged. I wanted the test suite to be a goto reference for all the behaviors till the time we get all of them documented.

test: fix output after rev-branch-cache has been merged Mercurial core 2090044a288d added a new revbranchcache part which changed some exchange-related outputs. Glob them to ease merging with old Mercurial versions branches.

cleanup: remove duplicate import of "obsolete"

previous: fix behavior on obsolete cset when topic is involved (issue5708) This patch fixes the behavior of `hg previous` when done on an obsolete changeset having a successor of different topic. The underlying code is bit messed up, I wanted to split the topic handling to topic extension but I have postponed it to next bug.

previous: add test case when it's broken on obsolete cset This patch adds couple of tests showing `hg previous` is broken on obsolete changeset when the successor has different topic. Upcoming patch will fix this.

prev: fix the breakage of `hg previous` from obsolete cset This patch fix the breakage of `hg prev` from obsolete changeset when topic is involved. We get the topicidx of it's succesor and proceed.

prev: add a test case which shows traceback when doing prev on obsolete cset This patch adds a test which shows how `hg prev` is borken when called from a obsolete changeset and there is an active topic. The next patch will fix that.

tests: add more tests related to stabilization of orphaned merge commit This patch more tests related to stabilization of orphaned merge commit. Stabilization of a merge changeset with both the parents is not supported right now and prints a message on how to do that. Maybe we should do that internally as a part of `hg evolve` but I need to think about this.

evolve: fix stabilization of orphan merge changesets in case of conflicts As demonstrated by tests added in previous patch, if one of the parent of a merge changeset has a new version and merge changeset gets, running `hg evolve` to stabilize that if results in conflicts will lead to lose of other parent of that merge changeset when we resolve conflicts and do `hg evolve --continue`. This patch fixes the behavior, by checking whether we are continuing evolving a merge changeset and if yes, setting the dirstate parents according to that.

tests: add a test file of resolving orphaness of merge changesets This patch adds a new test file named test-evolve-orphan-merge.t which contains test cases when `hg evolve` try to stabilize a merge changeset which was orphan. This file contains 20-30% of all the cases as we have got something to fix before we add tests for rest of the cases. Going by this approach will make sure we don't live unfixed cases in the test file and the test files does not become a pile of hacks which are used to get in state when we fix the bugs.

grab: gracefully handle the case when we try to grab parent of wdir Before this patch, we go all the way down to obsmarker creation logic and gets an error saying that changeset cannot obsolete itself. We should be more smart here and error out early saying we cannot grab that changeset.

grab: move the initialization of pctx variable outside of if-else The pctx variable is used outside of the if-else but it was initialized under the conditional statements. Added test for the case when the initialization inside the if-else broke `hg grab --continue`.

compat: drop hack around http request handling for 4.6 The approach we used in 4.5 works again :-)

obsdiscovery: add compatibility layer to register wireproto command Mercurial core 0b18604db95e changed the way permissions are registered.

compat: pass all hgweb dispatch argument to the orig function This make use compatible with whatever changes on this side.

grab: add short version for continue and abort

obsdiscovery: include units in ui.progress() calls (issue5773)

tests: commit tests changes missing in the previous changeset (guess who forgot to amend)...

obslog: add the operation to the Obslog output The operation is helpful when debugging a case where we don't know how we ended up in this case while effect-flag are useful to understand what semantic changes happens to a changeset. Add the operation when displaying an obs-marker in the following form: `rewritten as X using amend` The other proposal is interesting but seems harder to internationalize and might requires update each time we introduce new operation: `Amended as X` Let's use the first proposal right now.

serverminitopic: attempt to fix assertion for repoview in branchmap This seems fine since we topic never impact public only caches.

compat: register a version of 'obsfatedata' template that works with 4.6 Since showlist is deprecated, let us use compatlist.

compat: register a version of 'troubles' template that works with 4.6 API Since showlist is deprecated, let us use compatlist.

exthelper: add support for "requires" argument for template keyword This will help use to support the new template API.

compat: add compat layer for date related functions

branching: merge with stable

compat: more straight forward alias for old template names The template registration have significantly changed in core:7b74afec6772. We take simpler approach to define the alias in a way that works for all versions.

serverminitopic: fix wrong object being passed to phase

legacy: respect explicit config for server.bundle1 Unconditionally disabling bundle1 is a problem for larger hosting website. We adopt a less aggressive approach and only disable it the default would be used.

branching: merge with stable

compat: only wrap wireprotoserver function if they exist Older version does not need the wrapping so everything should be fine.

pure: make sure what we expect to be list are list In the pure version, some of these object would be generator.

evolve: update code for not implicitly converting ctx to revision Mercurial core 38f480502043 removed the capacity to implicitly convert contexts to revisions using `int(ctx)`. Update Evolve code to explicitly call rev instead.

legacy-push: fix the code forbidding legacy obsmarkers pulling changeset e69e65b2b4a9 in code remove a key try/except. We reintroduce it for compatibility.

updatecaches: handle any new parameters to the method

evolvecmd: fix broken `evolve --continue` ran after `next --evolve` Recent series broke `evolve --continue`in cases when run after `hg next --evolve`. The previous patch adds test demonstarting the broken behaviour, this patch fixes that by storing the command which leads us to evolve --continue in the evolvestate.

tests: add test showing `evolve --continue` after `next --evolve` is broken Recent series fixing `hg evolve --continue` break the continue functionality after a conflicting `hg next --evolve`. This patch adds a test to demonstrate that. It will be fixed in the upcoming patch.

color: add some colors to evolve command itself

wireproto: fix *_capabilities for change of API types Mercurial commit 2f7290555c96 introduced new types for *_capabilities, update the code to use them.

obsfate: fix changeset description diff computing Mercurial core 8b6dd3922f70 changed the mdiff.unidiff API. Add a new compat method for all supported Mercurial versions.

evolve: add comptability to read old evolvestate files When a user runs `hg evolve` with old evolve where we used to write evolvestate the old ways i.e. without using cbor, and faces conflicts, the evolve is interrupted. If the user updates evolve before completing the evolve, that can result in traceback as the new evolve cannot read the old evolvestate file. The previous patch added function to read old evolvestate file and this patch uses that when we encounter the above mentioned case.

state: bring back the function to read old evolvestate files Before introduction of this nice cmdstate file wrapper, we had individual methods to write data to evolvestate file. Let's bring back the function which used to read data from the old evolvestae file so that we can use it in next patch for BC purposes.

evolve: add entry in CHANGELOG about fixed behavior of `hg evolve --continue`

evolve: store the skippedrevs in evolvestate While doing `hg evolve`, there are some revs which are skipped as they can't be evolved. Before this patch, we did not store this information in evolvestate and neither they were evolved, so `hg evolve --continue` will again try to stabilize those revisions. As much time as we run `hg evolve --continue`, this will happen which is not a good behavior. This patch adds a skippedrevs list to evolvestate where we store such revisions so that we don't try to stabilize them again.

evolve: show the updated working directory after `hg evolve --continue` `hg evolve` shows a nice message about the new working directory when the evolve ends but if the evolve is interrupted by conflicts and you continue the evolve resolving the conflicts, that nice message is lost. Let's have that in case of `--continue` too.

evolve: make `hg evolve --continue` respect `--confirm` Previous patch added capability to `hg evolve --continue` to continue the rest of the stack, not just the current revision but that was missing respecting the `--confirm` flag which user passed earlier. This patch makes sure we respect the `--confirm` flag.

evolve: make `--continue` continue evolving all remaining revisions This patch makes `hg evolve --continue` to continue all remaining revisions instead of just the current conflicted one. There are some still improvements required to this behavior after which the CHANGELOG entry will be added.

tests: add more tests to test-evolve-continue.t This patch adds more tests for `hg evolve --continue`. The tests demonstrates how the continuation handling of hg evolve is broken and the later patches fixes that handling.

evolve: handle case when newnode can be None while continuing evolve There can be case when `hg evolve --continue` is run, when working directory is clean and there is nothing to be included in the new commit. Let's make sure we handle such cases correctly and obsolete the node which is evolved with parent of wdir as successor. This is same as what rebase does.

tests: add tests for `hg evolve --continue` This patch adds a new test file containing tests for `hg evolve --continue`. The test file still does not contains all the tests related to `hg evolve --continue` as one of the case is broken which needs to be fixed first. The case is when resolving a conflicting evolve results in empty wdir. Next patch will fix the traceback and then a later patch will add more tests.

merge with stable

CHANGELOG: add an entry about resolve showing how to continue evolve

test: explicitly run HGEDITOR shell command with `sh` Otherwise Windows pops up the "What program do you want to use to open this?" dialog.

test: conditionalize blackbox output for Windows

branching: merge with stable

compat: add wrapper for logcmdutil functions These classes and functions were bulk-renamed at hg c8e2d6ed1f9e.

evolve: store the revs to be processed in the evolvestate This will help us in getting a full functional `hg evolve` where `hg evolve --continue` does not only continue one changeset.

evolve: store the replacements in the evolvestate This will greatly help when we will be implementing `hg evolve --abort`.

evolve: store the target category in evolvestate

evolvecmd: use the evolvestate object passed as argument in _solveunstable Earlier if get a merge conflict, we used to initialize a new evolvestate object and used to save it. Let's start using the state object which we get in as argument as that will contain more data about the state.

evolvecmd: pass evolvestate object to _solve* functions This patchs adds an argument evolvestate to _solve* functions. This will help us in having one evolvestate object which we can pass around various functions and keep storing what's going on.

evolve: create a state object in all case Let's create a state object even when we are not in the continue mode. This will help us passing the state object to rest of the functions.

branching: merge with stable

grab: preserve phase of the grabbed changeset This patch fixes the wrong behavior of loosing secret phase of the grabbed changeset.

tests: add a test showing grabbing of secret changeset turn it into draft This patch adds a test showing wrong behavior of `hg grab` command which looses the secret phase while grabbing a secret changeset.

grab: add grab info to cmdutil.afterresolvedstates This will help us in showing how to continue the grab when user resolves all the conflicts.

evolve: add evolve info to cmdutil.afterresolvedstates Adding evolve info to cmdutil.afterresolvedstate will improve the ui when we run into merge conflicts while evolving. Once we resolve all the conflicts using `hg resolve`, this will show a message saying "continue: hg evolve --continue". Other extensions like rebase and shelve also use it and it's very helpful.

doc: add link to roadmap Add a link to the Evolution roadmap in the main page of the documentation.

evolve: move the evolve command to evolvecmd.py This patch moves finishes the code movement by moving the evolve command to evolvecmd.py.

evolve: move _orderrevs() function evolvecmd.py _orderrevs() is a function which is used to order the revs in which the instability needs to be solved if we have multiple revisions to resolve.

evolve: move builddependencies() and _singlesuccessor() to utility module The builddependencies() and _singlesuccessor() function are very useful functions and can be used at a lot of places. They are also used in topic extension also, so let's move them to utility. However moving them to utility won't help anything in the topic extension.

utility: move MultipleSuccessorsError from __init__.py The MultipleSuccessorsError is required in multiple modules in upcoming patches, so let's move it to utility first.

evolvecmd: move more core from __init__.py to evolvecmd.py We are done with moving helper code, we will now be moving the evolve command in the new module.

evolvecmd: move code from __init__.py to evolvecmd.py

evolvecmd: move more code from __init__.py to evolvecmd.py We are now very close to have most of the evolve related code in evolvecmd.py.

evolvecmd: move more chunks of code from __init__.py to evolvecmd.py

evolvecmd: move more functions from __init__.py to evolvecmd.py If things are looking ugly, hold on.

evolvecmd: introduce a new module to handle `hg evolve` related code __init__.py is to cluttered with wrapping and all, it's better to move code related to evolve command to new a module which will help us in great refactoring and introducing nice and better flow of data through functions. There is a temporary import cycle is introduced which will be fixed in the series.

evolve: remove not required calls to phases.retractboundary() To make sure phases are preserved during evolve, we started using the phases.new-commit config option. Now we don't need to call phases.retractboundary(). Let's remove the redundant code.

evolve: store the replacements occurred during `hg evolve` at top level This patch starts storing the replacements made during the `hg evolve` command. The current dictionary is not used anywhere but will be used in future as we want to introduce abort functionality, so we want to delay writing of obsmarkers, storing these exchanges in statefile etc.

evolve: make the _solve* functions return a tuple This patch makes _solve* functions return a tuple which will contain information about whether we were able to solve the instability and if yes, then what was the newnode. For more details about tuple, look the fn docs added. There was some refactoring required in next command because of change in return values. This also leds to omitting the wdir update message in case of dry-run which is good.

evolvestate: rename the file to state.py and class name to cmdstate The evolvestate.py and the evolvestate class in it were introduced to help `hg evolve` grow continue and abort options. But the class and it's API can be easily used to store state information for other state files like we stored the information for grab state in the previous commit. Let's rename the the file and class.

branching: merge with stable

update: show "use 'hg evolve' to..." hint iff evolve command enabled Before this patch, one had to enable *all* commands (by including "allnewcommands" in the experimental.evolution config) for the hint to be shown.

obslog: drop period from summary line in accordance with convention

grab: add a command to grab a commit and update to it This patch adds anew command `hg grab` which will pick a commit and put it on the top of the working directory parent and updates to it. Earlier grab was an alias for `hg rebase -r <rev> -d .` which is now deleted to make sure `hg grab` resolves to the grab command and that alias. Continuing interrupted grab functionality is also there using the evolvestate class. Tests are also added for the new introduced command.

amend: query the wdir parent after taking lock (issue5266) If we query wdir parent without taking a lock, that can lead to bugs because the wdir parent can change is another process has changed the wdir parent. One such example of this was issue 5266. When a user is running amend and that amend is waiting for commit message, the user runs another amend which waits for lock. The second amend is waiting for lock, but has already read the description from the parent of working directory to use. Once the first amend completes the wdir parent changes but we still have the description from an old wdir parent. This patch fixes the bug by querying the description after taking lock. Attempts were made to add test for this but the results were unstable because they depend on time in which lock is released.

fold: make sure we move bookmarks after folding (issue5772) The rewriteutil.rewrite() API currently only handles moving bookmark from a old node to new node. In case on fold, we have multiple old nodes, from where a bookmark move is possible. The rewriteutil.rewrite() API is used by another commands also and it's not good to plug in some logic there. This patch handles bookmark movement from all the old nodes to the new node.

tests: add a test showing bookmark not moving in-case of `hg fold` This patch adds a test showing that bookmark is not moved when doing `hg fold`. This test represents issue 5772.

branching: merge stable back into default

test: update test output ot reply caps size change CORE-TEST-OUTPUT-UPDATE: 2e703ed1c713

branching: merge default test output into stable Mercurial 4.5 has been merged, we update the expected test output.

branching: open Mercurial 4.4 compatibility branch

branching: merge release 7.2.1 into default

packaging: mark source as development version This help avoiding consusion

Added tag 7.2.1 for changeset 6c772398eb4e

test-compat: merge mercurial-4.2 into mercurial-4.1

test-compat: merge mercurial-4.3 into mercurial-4.2

test-compat: merge stable into mercurial-4.3

packaging: prepare version 7.2.1

branching: merge with tutorial output update

test-compat: merge mercurial-4.2 into mercurial-4.1

test-compat: merge mercurial-4.3 into mercurial-4.2

test-compat: merge stable into mercurial-4.3

tutorial: docgraph output seems to have been dedented Update the outputs following latest version of docgrpah and its dependencies.

test: update output for more verbose blackbox CORE-TEST-OUTPUT-UPDATE: 853bf7d90804

test: update with new 'update' hint about obsfate CORE-TEST-OUTPUT-UPDATE: 35a0f6f31eef

evolve: update output for wider reporting of troubled changesets CORE-TEST-OUTPUT-UPDATE: 1a09dad8b85a

test: update output for new `merge --abort` message Core have updated its command to cancel a failed merge, update the test output to match it. CORE-TEST-OUTPUT-UPDATE: 41ef02ba329b

branching: merge with stable

changelog: mention 4.5 compatibility

trouble-reports: yield to in code reporting of new troubles Mercurial 4.5 has the feature in core, let core do the reporting in this case.

caches: fix automatic warming trigger The caches were supposed to get warmed when we receive a push. However the transaction created when we -issue- a push use a similar name and was confusing the detection. We now properly detect the case.

changelog: mention the packaging bugfix

packaging: properly install the evolve.thirdparty module It is properly added to the Manifest.in but that is not enough. We need to find another tool capable of catching things.

branching: merge release tags into default

packaging: mark source as development version This help avoiding consusion

Added tag 7.2.0 for changeset bf6b859807ba

test-compat: merge mercurial-4.2 into mercurial-4.1

test-compat: merge mercurial-4.3 into mercurial-4.2

test-compat: merge stable into mercurial-4.3

branching: merge stable back into default The test change has to be reverted.

branching: merge further compat change into stable

branching: merge further compat change into stable

branching: merge more compatibility fix into stable

packaging: merge 7.2.0 into stable

tests: update test output to match 4.4 The stable branch is sitll 4.4.

branching: backed out test output changes from core-1905aac253f0 This backs out evolve-ae6fddf39933

branching: backed out test output changes from core-02fdb8c018aa This backs out evolve-1905aac253f0.

compat: fix argument check for marker reading The previous check was broken for 4.2

cachevfs: use a compatibility later for all access

diff: drop default argument value This argument use the default value for an argument not available in earlier version. So we just stop passing the argument.

test: use hash in test-evolve-topic.t This helps testing compat with older version.

packaging: prepare version 7.2.0

changelog: prepare for coming release

stablerange: add missing licence and copyright header

obshistory: pass the csets description in getmarkerdescriptionpatch Previous patches removed any use of ctx in the function and we now just need the cset description in the function. Let's pass that only. This will save couple of calls of repo[node].

obshistory: remove the unused function The function is not used anymore.

obslog: add header to the changeset description diff This patch adds `diff -r <hash> -r <hash> <filename>` header to the output of the changeset description diff which is shown using the --patch flag of obslog command. This fixes the label/color issue for description diff also.

obslog: colorize the description diff shown in obslog -p This patch colorizes the description diff shown in `hg obslog -p`. The labels for the header are different from that of the patch as the main header which says `diff -r <somehash> -r <somehash> <filename>`. Next patch will hack to add that.

obslog: colorize the patch shown using `hg obslog -p` This patch adds color support to the patch shown in obslog command.

topic: suggest using topic when user creates a new head on branch This patch adds a message to use topics when user creates a new head on a branch.

changelog: update changelog with some of the benefit from the new format

evolve: use better words in conflict markers of `hg evolve` Though we call merge.graft internally, we still have the functionality to pass the words we want to show up in conflict markers. So let's use better words here.

tests: add a test demonstrating we still use graft in conflict markers This patch adds a test which shows that we still use the word graft in the conflict markers. Next patch will start using evolve instead of graft.

evolve: remove the old functions for writing evolvestate Previous patch made all the users of these functions to use the evolvestate class and now they are not required. Let's remove them.

evolve: use the new evolvestate class in place of old methods Earlier we had three methods, which used to handle all the storage around the evolvestate but in previous patch we have introduced an evolvestate class wrapper which can serve for all our needs. This patch replaces the existing usage of the evolvestate with the new class introduced. Upcoming patches will start using the class at more places.

evolvestate: add a class to wrap the state of `hg evolve` command This patch adds a new file which contains a class which will act as a wrapper for the data to be stored or used during the `hg evolve` command. This patch just introduces the file and does not use it at the moment. The upcoming patches will start using it.

evolve: start vendoring third party package cbor This patch adds a third party package named cbor. cbor is a data serialization format and this library provides API to convert Python Objects into cbor formatted data. To know more about cbor, hop onto cbor.io The files are copied from it's official bitbucket repository which is at https://bitbucket.org/bodhisnarkva/cbor/src/113f42203c941470c98f7174065ba9b384cff727/ cbor.py can be found at https://bitbucket.org/bodhisnarkva/cbor/src/113f42203c941470c98f7174065ba9b384cff727/py/cbor/cbor.py # no-check-commit as we are importing a third party package Modified test-check-flake8.t to not lint on thirdparty packages.

uncommit: add a new flag `--revert` to discard wdir changes after uncommit This patch adds a new flag to `hg uncommit` using which one can discard the wdir changes which are left after the operation. Both the changes, the ones which were before the uncommit and the ones which were added to wdir as a result of uncommit as cleared out.

evolve: use phases.new-commit config to retain phase information (issue5720) Since the previous patches replaced the use of graftstate with evolvestate and we don't graft anymore on evolve --continue, we have now the capability to fix underlying bugs. This patch makes sure we use the phases.new-commit config to make a new commit of the same phase as that of the predecessor. Tests added in the previous patch shows the bug being fixed.

tests: add a test demonstarting lose of phase in `hg evolve` This patch adds a new test file which demonstrates the lose of phase information when hg evolve command runs into conflict. This is algo bug5720. The upcoming patch will fix it. A new test file is added as existing test files serve their own purpose and are already cluttered.

evolve: add a utility fn to check validity of opts passed to `hg evolve`

evolve: don't use graftstate for continuing `hg evolve` Before this patch, if user runs `hg evolve`, we read the evolve state, write a graftstate from that information and run graft command. This patch replaces that logic with new logic which does not depends on the graft state. The new logic is very filtered part of logic from graft command. We did not require a lot of opts checking and filtering revs logic from the graft command as our use case is fixed. This is the start of series/work which will make `hg evolve` better and have a good statefile for itself.

topics: add a new templatekeyword `topic` This patch adds a new templatekeyword topic which can be used to list the topic of the changeset when using commands which supports templating.

packaging: update MANIFEST.in to include new tutorial files I forget to update the MANIFEST.in file when important the new tutorial files. This commit fix that.

test-compat: merge mercurial-4.2 into mercurial-4.1

test-compat: merge mercurial-4.3 into mercurial-4.2

test-compat: merge stable into mercurial-4.3

test: add a test to reproduce issue 5720

test: add missing cycle test Also fix precursors to break on obs markers cycles.

Small typos fixing in the topic tutorial

doc: import the training support Import the training support which was stored in a private-repository before.

branching: merge with stable

evolve: use new trouble names in commit messages There is case when we add trouble to the commit message while using `hg evolve`. This match makes sure, we use the new trouble names. The hash changed because it depends on commit messages.

evolve: use compat.TROUBLES to show troubles in user interface compat.TROUBLES is a dict which contains the trouble name for each trouble according to the hg version user is using. All of the replacements are related to `hg evolve` command.

topics: list only topic names in quiet mode This will be helpful in bash completion.

topics: take logic to parse username to a separate function In next patch we will be adding support to `hg stack` to show users, and this logic will be required there too. So let's take it out in a separate function.

changelog: fix a typo in the changelog There is probably many others but that one is fixed now.

caches: switch to 'auto' warming by default This options seems good let us make it the default.

caches: add a 'auto' option for obshashrange cache warming This option will only warm the cache when used as a server.

caches: extract some config reading in 'shouldwarmcache'

caches: pass the transaction to the "shouldwarncache" logic This will allow to have smarter mode about cache warming (eg: only warm them for server transaction.

caches: record 'desc' attribute on transaction This is useful to know if a transaction if server side or not.

caches: protect against missing connection If the connection cannot be established, we should not try to use it.

evolve: don't show working directory obsolete message if we were on it This patch tweaks showing wc obsolete message functionality to not show the message if the operation led us to the same changeset we were on before and it didn't obsoleted it. This make failed updates, update on current changeset and a pull with no change omit showing the message. This has some cons like if you are on rev 1 which is obsolete and you do `hg up <revset>` where revset is some revset which resolves to the rev 1, we won't show the warning as shown earlier.

prev: jump to parent's successor if parent is obsolete and topic is involved `hg stack` shows a linear chain of commits even when they are not linear chain and some changesets are obsoleted and have successors as separate head. This is very nice to have a post evolution view of the stack. However when `hg prev` is run on an unstable changeset whose parent is obsolete and does not topic but parents successor has the topic, it shows no parent on the topic. This patch makes `hg prev` update to parent's successor and make things follow the stack order.

tests: add a test showing wrong behaviour doing `hg prev` on orphan cset If there is an unstable changeset in the topic stack, whose parent is not a part of stack but has a successor which is part of stack, hg stack does shows them in linear order, but `hg prev` on that orphan changset says no parent on that topic. However it should update to the successor of the changeset. It will be fixed in the next patch.

test: update output to 02fdb8c018aa Now that we have directaccess in core, we get a new warning message when updating to a hidden changeset. Update the tests. CORE-TEST-OUTPUT-UPDATE: 02fdb8c018aa

test: update output to 9b3f95d9783d The unstable graph nodes are now shown with the character "*", update the test outputs. CORE-TEST-OUTPUT-UPDATE: 9b3f95d9783d

stablesort: use parent filtering in a place we forgot to Freak merge striked again.

obshashrange: add a progressbar to upgrade This section is the slowest, have some progress idea would be useful.

stablerange: add an sql index on subranges id too This should speed up query that find super ranges.

obshashrange: less brutal reset when receiving markers on existing node We now target the affected ranges only. The implementation is still slow but that should be better than what we had before.

obshashrange: update the cache updating logic Use the more modern option when available. (Same as the other caches).

stablerange: cleanup the update logic

stablesort: cleanup the update logic

firstmergecache: cleanup the update logic

depthcache: cleanup the update logic

changelog: add an entry about the obshashrange changes

stablerange: be more cautious when deleting the sql connection This triggered some crash in further development. This also match what is done elsewhere with this attribute.

stablerange: clarify sql transaction logic This should help concurrent access.

stablerange: use mergepoint based algorithm for the official stable range Changing the official stable range will impact all users of the infrastructure. We update version number of the cache file and discovery methods to clarify this. We do no keep compatibility with the old method over the wire. The new algorithm is much better and keeping compat is more work than we have time for right now.

stablerange: explicitly create a stablerange entry for branchpoint method We are about to change the default.

stablerange: have a mergepoint based sql backed class Everything is so generic that getting a class is 3 line of codes.

stablerange: introduce a sql backed version of the generic on disk cache This is getting use close to actually using the new sorting and algorithm for real.

stablerange: add a base class for on disk stored stablerange class

stablesort: use on disk cache for test

caches: factorise the cache warming check The old conditional was duplicated and hard to read. We factorise everything into the `utility` module.

stablerange: use repo-carried stablesortcache That one is common to all and disk persisted

stablesort: expose the cache through the repository Let have a unique instance, keep it warm and accessible.

stablesort: implement an ondisk cache Persisting the cache on disk will be important for large repositories.

stablesort: realign a misaligned continue We should make a better use of the jump cache now.

stablesort: remove some dead code This closure is no longer in use.

stablesort: abstract all cache access This prepare an on disk version of the cache

stablerange: use first merge cache to skip over linear section If we are to skip over a large section of the range, skipping everything until the first merge seems like a good idea.

stablerange: add a new 'firstmerge' cache This cache store the first merge ancestors of a changesets. This is useful to avoid iteration over the changelog when building stablerange.

stablerange: drop unused `until` utility We no longer needs this function.

stablerange: use cached size data instead of walking the graph Less iteration is better.

stablesort: record previous segment size in the jump That was, we can skip over a full jump without walking it. We'll make use of this in the next changeset.

stablesort: move jump recording inside the exclusive function This will allow use to record to cache size of segment in a later changeset. If the exclusive set is empty, we need to record it outside that function.

stablerange: compute jump size after jump retrieval only This prepares the caching of jump sizes.

stablesort: warm jump cache more efficiently We no longer for a full depth iteration for all request: - we exit early for non-merge, - for merge, we only iterate over the exclusive area.

stablesort: use a regular dict for jumps This will help knowing what revision we have already queried or not.

stablerange: use the jump information for faster iteration Instead of doing a full iteration over the exclusive set, we compare the jump information instead. This perform the same kind of logic than before but will allow for much faster iteration once all the caches are in place.

stablesort: expose the jumps sequence to other code The stable range needs it to build exclusive subrange efficiently.

stablesort: use 'depth' in mergepoint tie breaker The parents with the most depth will is considered lower. It has a couple of advantages. 1) the more shallow parent probably have less exclusive revision, 2) it makes Oedipus merge behave like close to the linear case,

stablesort: rework jump gathering The old code was buggy in some case, the new code give the same result as the 'headstart' version.

stablesort: stop recording jump type There are no mature user of this "jump type" data and the current implementation is known to be buggy. So we drop the logic for now. This will make on disk storage simpler. The data will be reintroduced later

stablesort: pass a jump recording function instead of a list This seems cleaner to abstract the underlying storage to the lower level. This also makes use more flexible about jump detection.

stablesort: minor indent fix

stablesort: clarify subcall to the exclusive side Before doing deeper rework of this logic, we tackle the simplest parts.

docgraph: update test output with new output

stablesort: record, cache and reuse jump Iterating below a merge means two things: 1) iterate over the part exclusive to the higher parents, 2) iterate from the lower parents. While iterating on the exclusive part, there will be case were we just go the next natural parent, and case were we'll have to "jump" to another revision. If we record all point this "jump" happens and their target, we can easily reproduce the iteration in the future. With that information we can iterate over the exclusive part of the merge without having to compute it entirely. In addition we store the reason of the jump. This will help the stable range processing later.

stablesort: fix head start computation The recursion was wrong and provided wrong result. As a side effect, this seems to hint that `stablesort_mergepoint_bounded` is not giving the right result for complex set.

stablesort: avoid attempting to sort a tuple This seems silly.

stablerange: use the filterparents utility

stablesort: use the filtered parents utility

depthcache: use parents filter in depth cache

parents: add a utility to filter parents There is various strange case of merge out there that needs to be handled. We add an utility function to handle them all.

stablerange: abstract the bit able to store cache into sql The part about actually storing the range is independent of the algorithm, so we should extract this before using it for the next iteration.