config: use the right API to access template access Preventing direct access to the underlying dict will help a coming refactoring of `config`. Differential Revision: https://phab.mercurial-scm.org/D9922

config: use the right API to access subrepository section Preventing direct access to the underlying dict will help a coming refactoring of `config`. Differential Revision: https://phab.mercurial-scm.org/D9921

config: test priority involving the command line Differential Revision: https://phab.mercurial-scm.org/D9917

config: test priority involving include Differential Revision: https://phab.mercurial-scm.org/D9916

config: add a test for priority when includes are involved Differential Revision: https://phab.mercurial-scm.org/D9915

errors: use InputError for incorrectly formatted dates Differential Revision: https://phab.mercurial-scm.org/D9913

errors: use StateError more in merge module Differential Revision: https://phab.mercurial-scm.org/D9912

errors: use exit code 40 for when a hook fails A hook can be used for checking inputs, state, configuration, security, etc., so it's unclear which of the existing exit codes to use. Let's instead add one specifically for failed hooks. I picked 40. Differential Revision: https://phab.mercurial-scm.org/D9910

context: add missing manifest invalidation after write in overlayworkingctx This was breaking my merge-diff logic that will be in the next patch. Differential Revision: https://phab.mercurial-scm.org/D9908

merge with stable

copies: fix an incorrect comment in graftcopies() from recent D9802 I don't think the bit about creating a merge in the filelog was correct. Or at least I couldn't find a case where it happened. Differential Revision: https://phab.mercurial-scm.org/D9899

narrow: delete a stale TODO about not sending groups the client already has 2c5835b4246b changed the changegroup generation to not send treemanifests for directories the client had before widening. As that commit mentions, we had already stopped before that commit to send the changelog and filelogs for files the client already had. Differential Revision: https://phab.mercurial-scm.org/D9898

copies: add an devel option to trace all files Filelog based copy tracing only trace copy for file that have been added. This is a trade off between skipping some rare copy case in exchange for avoiding atrocious-to-the-point-of-unusable performance. The changeset centric copy tracing does not need this trade off and naturally trace all copy, include the one involving non-new files. In order to ease the comparison from both algorithm, we add a small devel option to trace copy for all files in the target revisions. Differential Revision: https://phab.mercurial-scm.org/D9796

revlog: prepare pure parser for being overloaded The current class uses module-level variables which don't allow for reusing the current logic for a slightly different revlog version. Differential Revision: https://phab.mercurial-scm.org/D9903

revlog: use a less probable revlog version for backwards compatibility This allows us to define a usable v2. Differential Revision: https://phab.mercurial-scm.org/D9902

clang-format: show required version in skip message This will help people to configure their black Differential Revision: https://phab.mercurial-scm.org/D9890

heptapod-ci: indicate which version of clang-format is used for the run This will help understand possible mismatch Differential Revision: https://phab.mercurial-scm.org/D9889

black: show required version in skip message This will help people to configure their black Differential Revision: https://phab.mercurial-scm.org/D9888

heptapod-ci: indicate which version of black is used for the run This will help understand possible mismatch Differential Revision: https://phab.mercurial-scm.org/D9887

heptapod-ci: add a default value for HG_CI_IMAGE_TAG Differential Revision: https://phab.mercurial-scm.org/D9886

convert: option to set date and time for svn commits Converting to subversion repository is not preserving original commit dates as it may break some subversion functionality if commit dates are not monotonically increasing. This patch adds `convert.svn.dangerous-set-commit-dates` configuration option to change this behaviour and enable commit dates convertion for those who want to take risks. Subversion always uses commit dates with UTC timezone, so only timestamps are used. Test `test-convert-svn-sink.t` uses `svnxml.py` script to dump history of svn repositories. Atm the script is not printing `date` field from svn log. This patch changes this to allow checks on correctness of date and time convertion. Documentation is updated. Additional test case is added to test commit dates convertion. Differential Revision: https://phab.mercurial-scm.org/D9721

copies: simplify the conditional for _filter's case 3 The conditional is much simpler and the test are actually happier. This clarification of the conditional will also be necessary to properly support tracing more renames in a coming changeset. Differential Revision: https://phab.mercurial-scm.org/D9795

copies: clarify which case some conditional are handling This make the function a bit clearer. The middle conditional get no label because we about about to remove it. See next changeset for details. Differential Revision: https://phab.mercurial-scm.org/D9794

copies: explicitly filter out existing file in graftcopies The `graftcopies` function does something very strange (maybe even wrong), it calls `_filter` with a pair of changeset that does not match the one used to compute the copies informations. We are about to do some rework of `_filter` to make it closer to its documented intent and fix a couple of bug. This means some of the logic that only make sense for graft need to go somewhere else. We add the extra filtering with proper documentation to `graftcopies`. Differential Revision: https://phab.mercurial-scm.org/D9802

histedit: rip out mysterious catch-all ignore curses.error handler I have no idea why this was here, and ripping it out doesn't obviously break anything for me (tests all pass, I can poke around chistedit UI a bit without issue), so I'm thinking we should rip it out and see if we get bug reports. Differential Revision: https://phab.mercurial-scm.org/D9855

histedit: notice when the main window underflows height and abort If you try to have a 13-line-tall terminal and use curses histedit, it fails by spinning in an infinite loop due to the catch-all ignore of curses errors on line 1682 of histedit.py. We should also fix that catch-all ignore of curses errors (what other demons lurk here, I wonder?) but we can trivially catch this case and guide the user to a happy path. We've seen this mostly in IDE users that have a tendency to have really tiny embedded terminal windows. Differential Revision: https://phab.mercurial-scm.org/D9854

histedit: don't assign to _ for unused values I don't know what this ignored value is, but we need to not clobber the _() function from gettext, or we get mysterious crashes instead of internationalizing some strings in my upcoming patches. Differential Revision: https://phab.mercurial-scm.org/D9853

hghave: list the module needed for the `vcr` check I'm tired of having to look up modules each time I setup a system, and try to distinguish between similar package names to get the right one. Now that the search API has been disabled, it's even harder. There are other python packages here that should be listed like this, but this is the one that came up missing today, so it's a start. Differential Revision: https://phab.mercurial-scm.org/D9879

debugshell: add a simple command for starting an interactive shell This simplifies interactive exploration of the Mercurial APIs. The ui and repo instances are provided as local variables. Differential Revision: https://phab.mercurial-scm.org/D9866

rust: replace an unsafe use of transmute with a safe use of bytes-cast Differential Revision: https://phab.mercurial-scm.org/D9859

rust: use the bytes-cast crate to parse persistent nodemaps This crate casts pointers to custom structs, with compile-time safety checks, for easy and efficient binary data parsing. See https://crates.io/crates/bytes-cast and https://docs.rs/bytes-cast/0.1.0/bytes_cast/ Differential Revision: https://phab.mercurial-scm.org/D9788

cext: restore the ability to build on Windows with py2 This broke in e92ca942ddca. Differential Revision: https://phab.mercurial-scm.org/D9867

command-namespace: use `::` are the command separator This was discussed on the mailing list / phab and `::` got a couple of positive feedback. I dediced to not introduce automatic folding of `::`, I feel like it make sense for `-`, but I like the expliciteness of `::`. So I am adding alias for backward compatibility. Differential Revision: https://phab.mercurial-scm.org/D9872

perf: test the formatting of a command help Since we are about to addd ':' in command name I want this covered. Differential Revision: https://phab.mercurial-scm.org/D9871

help: escape ':' (as '\:') when generating command names We need this before we can use ":" in command names. Differential Revision: https://phab.mercurial-scm.org/D9870

minirst: respect escaping in definition list key If we start adding `:` in command names… we need minirst to not choke on them. Differential Revision: https://phab.mercurial-scm.org/D9869

run-test: avoid byte issue when replacing output file of python test Otherwise we get error like:: FileNotFoundError: [Errno 2] No such file or directory: "b'…/tests/test-minirst.py'.out" Differential Revision: https://phab.mercurial-scm.org/D9868

store: use `endswith` to detect revlog extension Suggested by Gregory Szorc. Differential Revision: https://phab.mercurial-scm.org/D9865

black: merge config into main pyproject.toml now that we have it This means that naive contributors who just run `black` on a source file will get reasonable behavior as long as they have a recent black. Yay! Differential Revision: https://phab.mercurial-scm.org/D9834

pyproject: add config file This will tell pip et al to call our setup.py for the majority of packaging concerns, but also gives us a place to put standard config stuff like black. Differential Revision: https://phab.mercurial-scm.org/D9833

upgrade: update only requirements if we can Upgrade operations which involves just upgrading requirements earlier used to go through whole revlog cloning business. Now we just upgrade the requirement and skip the cloning part. Differential Revision: https://phab.mercurial-scm.org/D9775

engine: add `if True` to prepare for next patch This will help making next patch easier to read. Differential Revision: https://phab.mercurial-scm.org/D9774

test: unquiet few tests to demonstrate changes in upcoming patches Upcoming patches will skip revlog cloning for share-safe upgrades. Differential Revision: https://phab.mercurial-scm.org/D9773

upgrade: mark sharesafe improvement as only touching requirements Upgrading the repository to use share safe functionality only touches requirements. Differential Revision: https://phab.mercurial-scm.org/D9772

actions: calculate what all parts does the operation touches To make upgrade work less in certain situations, we need to teach it to find out what all parts of repository current operation is touching. This patch starts doing that. In next patch we will be setting values in improvement objects. For now, we assume everything touches everything. Differential Revision: https://phab.mercurial-scm.org/D9771

upgrade: don't create store backup if `--no-backup` is passed If the user explicitly mentioned that they don't need backup, then let's not create it. Differential Revision: https://phab.mercurial-scm.org/D9770

cext: add Python 3.10 support * Replace "Py_TYPE(obj) = type;" with "Py_SET_TYPE(obj, type);" * Add pythoncapi_compat.h header file to get Py_SET_TYPE() on Python 2.7-3.8. Header file added to mercurial/ and contrib/python-zstandard/zstd/common/. In Python 3.10, Py_TYPE(obj) must not longer be used as an l-value. pythoncapi_compat.h comes from: https://github.com/pythoncapi/pythoncapi_compat Differential Revision: https://phab.mercurial-scm.org/D9825

revlog: decouple caching from addrevision callback for addgroup For changesets, it is useful to cache the content as it will almost always be processed afterwards. For manifests on the other hand, the content is often not used directly as there is a fast path for deltas. Explicitly disable the cache in exchangev2's manifest handling for that reason. Differential Revision: https://phab.mercurial-scm.org/D9847

branchmap: update rev-branch-cache incrementally Historically, the revision to branch mapping cache was updated on demand and shared via bundle2 to avoid the cost of rebuilding on first use. Use the new `register_changeset` callback and update rbc directly on every change. Make the transfer of the bundle part redundant, but keep it for the moment to avoid the test churn. Over all, "hg unbundle" for large bundles is less than 1.8% slower for different larger repositories and that seems to a reasonable trade off. Differential Revision: https://phab.mercurial-scm.org/D9781

repository: introduce register_changeset callback The new callback is called whenever a changeset is added to the repository (commit, unbundle or exchange). Since the bulk operations already parse the changeset (readfiles or full changesetrevision), always use the latter to avoid redundant lookups. The first consumer of the new interface needs to look at extra. Differential Revision: https://phab.mercurial-scm.org/D9780

changelog: move branchinfo to changelogrevision The function parses the extra dictionary after looking up the changelogrevision. To avoid duplicated look up, it is better to provide it as property of changelogrevision instead. Keep the function for a release cycle as at least the topic extension depends on it. Differential Revision: https://phab.mercurial-scm.org/D9779

rust: lower compile error on non-linux platforms to a warning As discussed on D9671#146704 this lowers the error to a warning. Differential Revision: https://phab.mercurial-scm.org/D9852

purge: move extension into core mercurial The motivation is simple: it's nicer to avoid gating basic functionality. To reduce the risk of people shooting themselves in the feet, `--confirm` is now the default, unless the extensions is loaded.. For review of the body of the purge command, use this instead of what hg/phabricator will show (the block of code is modified, not just moved): Differential Revision: https://phab.mercurial-scm.org/D9820

purge: also deal with directory with --confirm Getting the exact number is simple (you have to project the result of previous deletion to detect super-directory with only about-to-be-deleted content). So we use a vaguer message. Differential Revision: https://phab.mercurial-scm.org/D9819

purge: add a --confirm option The options provide a prompt to the user before permanent deletion are made. The prompt is currently not aware of directory deletion. I'll fix this in the next changesets. Differential Revision: https://phab.mercurial-scm.org/D9818

filemerge: add a hacktastic version of internal:merge3 for merge diffs This is a version of merge3 that always reports success, so that conflict markers get preserved without us having to implement conflict storage for in-memory merge. Credit to martinvonz for the idea. The only planned consumer of this "merge tool" is my upcoming merge-diffs functionality, though I suspect it could be useful in other ways. Differential Revision: https://phab.mercurial-scm.org/D8515

rebase: add a config knob for forcing in-memory rebasing Use it in a test case where we know the rebase should proceed without incident in-memory, so we can see tracebacks rather than fallbacks. This makes it easier to try and debug when things are broken. Differential Revision: https://phab.mercurial-scm.org/D8623

cleanup: use mergestate.unresolvedcount() instead of bool(list(unresolved())) This avoids some pointless copying. Differential Revision: https://phab.mercurial-scm.org/D8566

reverse-branch-cache: switch to doubling allocating scheme In preperation for updating the reverse-branch-cache incrementally whenever a new changeset comes in, avoid bad performance on resize with Python 3.7 (and likely other 3.x versions). Differential Revision: https://phab.mercurial-scm.org/D9778

discovery: add config options to control sample size This get useful for running analysis of the current algorithm. Differential Revision: https://phab.mercurial-scm.org/D9813

exchangev2: avoid second look-up by node Accessing the revlog by node is slightly more expensive than by revision, so look up the revision first and use it afterwards. Differential Revision: https://phab.mercurial-scm.org/D9831

commit: look-up new revision once Look-up by node is slightly more expensive, so since it is necessary more than once, do it explicitly. Differential Revision: https://phab.mercurial-scm.org/D9830

resolve: also detect new :mergediff conflict markers The conflict markers created by `:mergediff` were not detected as conflicts, which affects both `commands.resolve.mark-check` and `mergetools.<tool>.check`. This patch fixes that. The new regex it uses for finding conflict markers is less restrictive because it `:mergediff` doesn't follow the `<<<<<<<` and `>>>>>>>` lines by a space (and a description). Hopefully lines like that don't give too many false positives. We can add back the space and make `:mergediff` add trailing spaces if it turns out to be a problem. OTOH, there will always be some false positives and we have ways of overriding the checks already. This patch can go onto the default or stable branch, depending on how much we care about an experimental feature. Differential Revision: https://phab.mercurial-scm.org/D9835

simplemerge: delete unused find_unconflicted() The function has been unused ever since it was introduced in 465b9ea02868 (Import 3-way merge code from bzr, 2007-04-16). Differential Revision: https://phab.mercurial-scm.org/D9832