amend: don't remove unselected removals from memctx When there are removed files in the working copy and they are not selected to be amended into the parent, the `filectxfn` we create for the `memctx` would still return `None` before this patch. That's clearly incorrect; we should return the `filectx` from the unamended commit. Somehow it seems to not matter much except for the case with copies stored in changesets. Thanks to Kyle Lippincott for doing all the debugging and identifying the fix for this issue. Differential Revision: https://phab.mercurial-scm.org/D12573

tests: demonstrate crash on partial amend with copies in changesets See the fix in the next patch for explanation. Differential Revision: https://phab.mercurial-scm.org/D12572

rust-revlog: add methods for getting parent revs and entries Differential Revision: https://phab.mercurial-scm.org/D12442

rust-changelog: start parsing changeset data This patch makes `ChangelogRevisionData` do some coarse, line-level splitting of the changeset data into manifest node, user, timestamp, files list, and description. There are no (in-tree) users of these functions yet, but I've added tests to prevent regressions. We'll surely add callers at some point. Differential Revision: https://phab.mercurial-scm.org/D12439

rust-changelog: remove special parsing of empty changelog data for null rev For the null revision, `Revlog::get_rev_data()` will return an empty string (of bytes). We currently handle that case in `ChangelogRevisionData::manifest_node()`. However, it's going to be ugly to have special handling for the null revision for each future method on `ChangelogRevisionData`. This patch therefore restructures the code so we instead initialize the struct with valid data for the null revision. Differential Revision: https://phab.mercurial-scm.org/D12438

rust-changelog: don't skip empty lines when iterating over changeset lines The first empty line in the changeset indicates the end of headers and beginning of description. Callers can't know figure out where that position is if empty lines are skipped. Differential Revision: https://phab.mercurial-scm.org/D12426

rust-requirements: allow loading repos with `bookmarksinstore` requirement `rhg` does support bookmarks, so it can load repos with the `bookmarksinstore` requirement just as well as other repos. Differential Revision: https://phab.mercurial-scm.org/D12425

rebase: while rewriting desc hashes, ignore ambiguous prefix "hashes" If a repo is sufficiently large, a six digit number "hash prefix" can somewhat easily reference an ambiguous hash prefix. Differential Revision: https://phab.mercurial-scm.org/D12552

tests: add test demonstrating issue with ambiguous has prefixes during rebase Differential Revision: https://phab.mercurial-scm.org/D12551

test: accept another error message on lack of TLS client certificate Differential Revision: https://phab.mercurial-scm.org/D12492

sslutil: support TLSV1_ALERT_PROTOCOL_VERSION reason code It looks like python 3.10 returns a different reason code on protocol version mismatch. Differential Revision: https://phab.mercurial-scm.org/D12491

test: override default cipher selection when connecting to TLS 1.0/1.1 servers The default set of ciphers on python 3.10 is incompatible with old TLS versions. Differential Revision: https://phab.mercurial-scm.org/D12490

sslutil: be less strict about which ciphers are allowed when using --insecure Python 3.10 restricted which ciphers are enabled by default, leading to no available ciphers for TLS < 1.2. When using the --insecure flag we allow old TLS, so also adjust the cipher list to give connections a chance to work. On the server side, also loosen the cipher selection in tests (when using the devel.serverexactprotocol option). Differential Revision: https://phab.mercurial-scm.org/D12489

sslutil: avoid deprecation warnings from python 3.10's ssl module Use ssl.PROTOCOL_TLS_{CLIENT,SERVER} and SSLContext.{min,max}imum_version when supported (3.7+). And, catch deprecation warnings when the user asks for deprecated TLS versions (1.0 and 1.1). Differential Revision: https://phab.mercurial-scm.org/D12488

zeroconf: fix deprecation warning with python 3.10 threading.condition.notifyAll → threading.condition.notify_all Differential Revision: https://phab.mercurial-scm.org/D12487

test: deal with changed error message on python 3.10 Differential Revision: https://phab.mercurial-scm.org/D12493

mail: don't complain about a multi-word email.method I want to be able to set email.method to "ssh relay /usr/sbin/sendmail" without needing an extra trivial shell script. This works fine since we pass the full command to a shell, except for validateconfig trying to find it in $PATH. Differential Revision: https://phab.mercurial-scm.org/D7542

help: set the large-file-limit to 10MB This is a minor increase (5%) and makes the doc much clearer. Differential Revision: https://phab.mercurial-scm.org/D12484

help: clarify the unit of `ui.large-file-limit` config Its might be a bit confusing, especially since `large-file.min-size` uses MB. Differential Revision: https://phab.mercurial-scm.org/D12483

debuglock: ignore ENOENT error when unlocking This is consistent with the main `lock.release` code. Differential Revision: https://phab.mercurial-scm.org/D12481

run-tests: introduce "forward-slash" version of everything on windows This should be useful for some shell invocation. Differential Revision: https://phab.mercurial-scm.org/D12480

tests-racy-mutation: pass the editor through config instead of env variable On Windows msys seems to do awful mangling of the environment variable content that confuses everything to the death. Going through the config works fine, so we do that instead. Differential Revision: https://phab.mercurial-scm.org/D12479

rust-dirstate: don't return a state for untracked entries This `state` API is a remnant of the former API and is slated for removal at some point. Any caller of this function will expect an entry that is tracked in the larger sense. Differential Revision: https://phab.mercurial-scm.org/D12448

dirstate: remove v1_* methods from Python/C/Rust shared API These methods are used for v1 parsing by their respective implementations, but do not need to be shared between them. Differential Revision: https://phab.mercurial-scm.org/D12447

rust-dirstate-entry: fix typo in panic message Differential Revision: https://phab.mercurial-scm.org/D12446

test-issue660: add dirstate-v2 variant It's basically a dirstate test, so it makes sense to test out the new version. Differential Revision: https://phab.mercurial-scm.org/D12445

test-issue660: test inside a repository, not the test dir This causes an issue with a temporary file showing up in test output when adding a dirstate-v2 variant of this test. Differential Revision: https://phab.mercurial-scm.org/D12444

dirstate: fix some typos in docstrings I was passing by and they've been bothering me. :) Differential Revision: https://phab.mercurial-scm.org/D12443

path: explicitly declare the `pushurl` suboption This will help documentation and discovery. Differential Revision: https://phab.mercurial-scm.org/D12437

setup: fix incomplete implementation of Command `test-install.t` fails without the `get_outputs` method being implemented, which is used when, `self.report` is `True`. When 8d7eaff92f9c introduced this change, they probably ran `test-install.t` without `HGTESTS_ALLOW_NETIO=1`, which does not trigger this codepath. Differential Revision: https://phab.mercurial-scm.org/D12482

path: explicitly declare the `pushrev` suboptions This will help documentation and discovery. Differential Revision: https://phab.mercurial-scm.org/D12436

path: explicitly declare the `multi-urls` suboptions This will help documentation and discovery. Differential Revision: https://phab.mercurial-scm.org/D12435

path: explicitly declare the `bookmarks.mode` suboptions This will help documentation and discovery. Differential Revision: https://phab.mercurial-scm.org/D12434

crecord: avoid duplicating lines when reverting noeol->eol change When reversing a patch that looks like this while using crecord: ``` @@ -301,4 +302,4 @@ zza zzb zzc zzd -zze \ No newline at end of file +zze ``` we would previously reverse the `-zze` line to be an add, encounter the "no newline" line and stop inspecting lines. This caused us to duplicate the line, producing `zzezze` (still without a newline). `break` is the correct action if we know there will be no lines afterwards, as would be the case in an eol -> noeol transition. It is incorrect if there are lines afterward, such as if both sides are missing the newline or if only the lhs is missing the newline. Differential Revision: https://phab.mercurial-scm.org/D12441

crecord: add test demonstrating issue when reverting noeol->eol change Differential Revision: https://phab.mercurial-scm.org/D12440

revert: ask user to confirm before tracking new file when interactive If interactively reverting from a commit with `hg revert -i -r`, we would unconditionally add files from that commit that are not already tracked in the working copy. We have prompts for adding back files removed in the working copy, but that's specific to such files and does not apply to adding files from another revision. Differential Revision: https://phab.mercurial-scm.org/D12416

revert: use a `continue` to reduce indentation I'm about to add more code in the block I'm modifying here. Differential Revision: https://phab.mercurial-scm.org/D12415

stringutil: try to avoid running `splitlines()` only to get first line It's wasteful to call `splitlines()` and only get the first line from it. However, Python doesn't seem to provide a built-in way of doing just one split based on the set of bytes used by `splitlines()`. As a workaround, we do an initial split on just LF and then call `splitlines()` on the result. Thanks to Joerg for this suggestion. I didn't bother to also split on CR, so users with old Mac editors (or repos created by such editors) will not get this performance improvement. Differential Revision: https://phab.mercurial-scm.org/D12413

logcmdutil: use new function for getting first line of string Differential Revision: https://phab.mercurial-scm.org/D12412

filemerge: use new function for getting first line of string Differential Revision: https://phab.mercurial-scm.org/D12411

absorb: use new function for getting first line of string Differential Revision: https://phab.mercurial-scm.org/D12410

extensions: use new function for getting first line of string Differential Revision: https://phab.mercurial-scm.org/D12409

bookmarks: use new function for getting first line of string Differential Revision: https://phab.mercurial-scm.org/D12408

help: use new function for getting first line of string Differential Revision: https://phab.mercurial-scm.org/D12407

histedit: remove an unnecessary default value of `b''` for commit message I don't think `ctx.description()` is ever anything falsy other than `b''`. I think the comment added in 4c4232e51167 (histedit: extract common summary code into method, 2016-05-27) is mistaken (I don't see any code following that pattern before the function was added). Differential Revision: https://phab.mercurial-scm.org/D12406

histedit: use new function for getting first line of a string This fixes a crash you can run into if you enter a commit message that's just a "newline-like" byte, like a form feed byte (`hg ci -m \x0f` in Fish). That bug is the motivation for this series. Differential Revision: https://phab.mercurial-scm.org/D12405

templates: extract function to `stringutil` for getting first line of text It's surprisingly hard to get the first line from a string, so let's have our own function in `stringutil` for it. Differential Revision: https://phab.mercurial-scm.org/D12404

templates: make `firstline` filter not keep '\v', '\f' and similar In b288b4bb8448 (hide some functions behind lambdas, so demandload is useful., 2006-02-28), `x.splitlines(1)[0]` was replaced by `x.splitlines(1)[0].rstrip('\r\n')`, i.e. stripping trailing '\r' and '\n'. Combined with the "truthy" `1` passed to `splitlines()` to get it to keep line endings, that results in e.g. trailing '\v' (Line Tabulation) and '\f' (Form Feed) being preserved. I can't see why one would want that, and I doubt that was the intention; I suspect the author just didn't think to instead remove the `1` argument. Perhaps they thought the 1 being passed there - added by themselves in a7e416bf3c1d (improve templating., 2006-02-27) - was to limit the number of splits to 1 (i.e. thinking about it as `maxsplit=1` rather than `keepends=1`). Differential Revision: https://phab.mercurial-scm.org/D12403

pytype: drop py3.6 support Pytype 2022.01.07 only supports 3.7+. Differential Revision: https://phab.mercurial-scm.org/D12400

perf-util: add a `compare-discovery-case` script This script run the same discovery case using multiple variants of the algorithm and report differences in behavior, especially regarding the numbers of roundtrip. Differential Revision: https://phab.mercurial-scm.org/D12399

discovery: also audit the number of queries done In addition to the number of roundtrip, we now also track the number of queries we perform, this is useful to assert the tradeoff between number of roundtrip and the number of queries. Differential Revision: https://phab.mercurial-scm.org/D12398

search-discovery-case: display more information about the interresting case We display information about the total number of revs and the common/missing numbers. This is useful to spot the interresting case. Differential Revision: https://phab.mercurial-scm.org/D12397

subsetmaker: rework the antichain generation to be usable Before this, antichain computation can run for 10s of hours without completion in sight. We use a more direct approach in the computation to keep the computation in complexity in check. With good result. We can now have a full antichain computation on mozilla-try in about one minute. Which is usable. Differential Revision: https://phab.mercurial-scm.org/D12396

subsetmaker: use SortedSet for the scratch variant This provides a massive speedup on wide repository with many heads. For example on mozilla-try, this move from un-usable slow to fairly instant. Differential Revision: https://phab.mercurial-scm.org/D12395

subsetmaker: stabilize the computation of `scratch` subset `heads` is set, order of the element are not deterministic and we need to stabilize that if we want to get reproducible results. Differential Revision: https://phab.mercurial-scm.org/D12394

revlog: recommit 49fd21f32695 with a fix for issue6528 `filelog.size` currently special cases two forms of metadata encoding: - copy data via the parent order as flag bit - censor data by peaking into the raw delta All other forms of metadata encoding including the empty metadata block are mishandled. In `basefilectx.cmp` the empty metadata block is explicitly checked to compensate for this. Restore 49fd21f32695, but disable it for filelog, so that the original flag bit use contines to work. Document all this mess for now in preparation of a proper rework. Differential Revision: https://phab.mercurial-scm.org/D11203

merge-lists: make it possible to specify pattern to match The `merge-lists` tool doesn't know anything about Python other than its regex that attempts to match import lines. Let's make it possible to pass in a custom regex so it's easy to use the tool for e.g. C/C++ `#include` lines or Rust `use` lines (given the limited). Differential Revision: https://phab.mercurial-scm.org/D12392

contrib: add a partial-merge tool for sorted lists (such as Python imports) This is a pretty naive tool that uses a regular expression for matching lines. It is based on a Google-internal tool that worked in a similar way. For now, the regular expression is hard-coded to attempt to match single-line Python imports. The only commit I've found in the hg core repo where the tool helped was commit 9cd6292abfdf. I think that's because we often use multiple imports per import statement. I think this tool is still a decent first step (especially once the regex is made configurable in the next patch). The merging should ideally use a proper Python parser and do the merge at the AST (or CST?) level, but that's significantly harder, especially if you want to preserve comments and whitespace. It's also less generic. Differential Revision: https://phab.mercurial-scm.org/D12380

branching: merge stable into default

Added signature for changeset 5bd6bcd31dd1

Added tag 6.1.1 for changeset 5bd6bcd31dd1

relnotes: add notes for 6.1.1 This also fixes the header for 6.1 from 6.1rc0

merge: stable into default

rust-hgpath: add `repr(transparent)` to `HgPath` It's been stabilized a long time ago, so let's not rely on an implementation detail now. Differential Revision: https://phab.mercurial-scm.org/D12433

rust-dirstatemap: correctly decrement the copies counter This was caught when writing unit tests for the `DirstateMap`. We were always setting `had_copy_source` to `false` since we erased the value just before. Differential Revision: https://phab.mercurial-scm.org/D12432

rust-dirstatemap: properly decrement counter for tracked descendants I found this bug when writing unit tests after the fact for the `DirstateMap`. We never decremented the tracked descendants counter since we were always resetting the node data before reading it. This also drops the use of `state`, in favor of the new API to get that information. Differential Revision: https://phab.mercurial-scm.org/D12431

rust-dirstate: panic if the DirstateMap counters go below 0 When modifying the API I hit some... interesting errors (trying to allocate 178GB of RAM, for example) because I failed to keep the counters correctly updated. This counter underflow is likely to happen when code is changed around and can have up to eat-your-dirstate level of consequences, which is not nice. The very small runtime cost of checking these counters should really not be an issue and will help us uncover bugs when/if they do appear in the future. Differential Revision: https://phab.mercurial-scm.org/D12430

rust: fix unsound `OwningDirstateMap` As per the previous patch, `OwningDirstateMap` is unsound. Self-referential structs are difficult to implement correctly in Rust since the compiler is free to move structs around as much as it wants to. They are also very rarely needed in practice, so the state-of-the-art on how they should be done within the Rust rules is still a bit new. The crate `ouroboros` is an attempt at providing a safe way (in the Rust sense) of declaring self-referential structs. It is getting a lot attention and was improved very quickly when soundness issues were found in the past: rather than relying on our own (limited) review circle, we might as well use the de-facto common crate to fix this problem. This will give us a much better chance of finding issues should any new ones be discovered as well as the benefit of fewer `unsafe` APIs of our own. I was starting to think about how I would present a safe API to the old struct but soon realized that the callback-based approach was already done in `ouroboros`, along with a lot more care towards refusing incorrect structs. In short: we don't return a mutable reference to the `DirstateMap` anymore, we expect users of its API to pass a `FnOnce` that takes the map as an argument. This allows our `OwningDirstateMap` to control the input and output lifetimes of the code that modifies it to prevent such issues. Changing to `ouroboros` meant changing every API with it, but it is relatively low churn in the end. It correctly identified the example buggy modification of `copy_map_insert` outlined in the previous patch as violating the borrow rules. Differential Revision: https://phab.mercurial-scm.org/D12429

rust: explain why the current `OwningDirstateMap` is unsound See inline comments. Differential Revision: https://phab.mercurial-scm.org/D12428

dispatch: fix silly blackbox entries when hg is interrupted When hg is interrupted, it creates ui.log like this: 1970/01/01 00:00:00 user @0000000000000000000000000000000000000000 (62488)> killed! exited 255 after 1.78 seconds This is due to a scoping problem: two different uses of the name "msg" collide. So rename one of them. Differential Revision: https://phab.mercurial-scm.org/D12427

tests: stop excluding the pycompat module from pyflakes I assume this was skipped because of all of the py2 stuff causing a lot of spew. The "unused" imports are left in place in case any 3rd party stuff is using it. I don't care about most of it, but TortoiseHg uses `io` and `queue`, so minimally I'd like to keep those. Differential Revision: https://phab.mercurial-scm.org/D12423

tests: drop some py2 specific pyflake failures Differential Revision: https://phab.mercurial-scm.org/D12422

util: drop a duplicate import This was already imported several lines above. Differential Revision: https://phab.mercurial-scm.org/D12421

pycompat: drop the pickle import I suspect this is what df56e6bd37f6 meant to eliminate. Differential Revision: https://phab.mercurial-scm.org/D12420

util: restore the util.pickle symbol This was accidently dropped in df56e6bd37f6, which started importing pickle directly. That commit explicitly says it will retain it for compatibility with external stuff though. The unused import in pycompat isn't flagged because that module is skipped. Just importing with a comment seemed cleaner than `import X as Y` and then assigning to a `pickle` variable, just to avoid the pyflakes warning. Differential Revision: https://phab.mercurial-scm.org/D12419

merge: stable into default

precheck: fix false warning about content-divergence creation Before this patch, if we try to `hg prune` (without any successors) an already obsoleted cset which has at least one successor, it would false warn about new content-divergence. As we know, pruning cset without any successors can not create any divergence. Differential Revision: https://phab.mercurial-scm.org/D12002

streamclone: avoid some obscure error in a corner case I don't really know how, but I ran into this error: $ hg clone --stream ssh://user@dummy/empty-repo local-empty-repo streaming all changes abort: unable to apply stream clone: unsupported format: [255] I think you need an empty list of requirements for this to happen, which is weird, but an obscure error like this is not exactly helpful either. Since this is the result of an encoding bug anyway, just fix it. Differential Revision: https://phab.mercurial-scm.org/D12402

tags: fix typo in fast path detection of fnode resolution (issue6673) If I understand it, mctx.readfast() is unreliable here if p1/p2 .hgtags nodes differ, and tags on that branch would be randomly discarded depending on which parent were picked. The test case added by this patch would fail only on zstd-compressed repository. I didn't try hard to stabilize the failure case.

completion: install completers to conventional locations Installs the bash and zsh completers to the convential locations so they will automatically be picked up without user intervention. The zsh completer on Debian is still installed to vendor-completions to match their policy. bash: https://github.com/scop/bash-completion#faq zsh: https://github.com/zsh-users/zsh/blob/57305cf245853b8b30895b41a90142dffab97e38/INSTALL#L254 Debian zsh: https://salsa.debian.org/debian/zsh/-/blob/5086b5356abcef8849dc8a09902b7c55f01db3c0/debian/README.Debian#L73

dirstate-cext: properly invalidate mtime and data in `set_untracked` This was forgotten about in the initial implementation and was revealed while adding the `dirstate-v2` variant of `test-issue660.t`. Neither the existing Python implementation nor the upcoming Rust implementation suffer from this bug since they respectively have `None` and `Option<T>` to represent the lack of information. Differential Revision: https://phab.mercurial-scm.org/D12414

pytype: disable a few errors about Windows specific module attributes These were flagged by pytype 2022.03.21. Differential Revision: https://phab.mercurial-scm.org/D12401

rhg: sort unsupported extensions in error message This caused some flakiness in test output, and is also just better for users. Differential Revision: https://phab.mercurial-scm.org/D12389

hgignore: ignore .testtimes in more location See the inline comment. Differential Revision: https://phab.mercurial-scm.org/D12393

merge: with stable

partial-merge: add support for `.args` config (`$local` etc.) It will be useful to be able to define custom command-line arguments per partial merge tool just like we have for regular merge tools. In particular, I expect the same binary to handle multiple languages, so it will be useful to be able to pass some argument indicating the language, or perhaps simply an argument defining a regex that's used for finding lines to merge as a sorted set. Differential Revision: https://phab.mercurial-scm.org/D12383

filemerge: add support for partial conflict resolution by external tool A common class of merge conflicts is in imports/#includes/etc. It's relatively easy to write a tool that can resolve these conflicts, perhaps by naively just unioning the statements and leaving any cleanup to other tools to do later [1]. Such specialized tools cannot generally resolve all conflicts in a file, of course. Let's therefore call them "partial merge tools". Note that the internal simplemerge algorithm is such a partial merge tool - one that only resolves trivial "conflicts" where one side is unchanged or both sides change in the same way. One can also imagine having smarter language-aware partial tools that merge the AST. It may be useful for such tools to interactively let the user resolve any conflicts it can't resolve itself. However, having the option of implementing it as a partial merge tool means that the developer doesn't *need* to create a UI for it. Instead, the user can resolve any remaining conflicts with their regular merge tool (e.g. `:merge3` or `meld). We don't currently have a way to let the user define such partial merge tools. That's what this patch addresses. It lets the user configure partial merge tools to run. Each tool can be configured to run only on files matching certain patterns (e.g. "*.py"). The tool takes three inputs (local, base, other) and resolves conflicts by updating these in place. For example, let's say the inputs are these: base: ``` import sys def main(): print('Hello') ``` local: ``` import os import sys def main(): print('Hi') ``` other: ``` import re import sys def main(): print('Howdy') ``` A partial merge tool could now resolve the conflicting imports by replacing the import statements in *all* files by the following snippet, while leaving the remainder of the files unchanged. ``` import os import re import sys ``` As a result, simplemerge and any regular merge tool that runs after the partial merge tool(s) will consider the imports to be non-conflicting and will only present the conflict in `main()` to the user. Differential Revision: https://phab.mercurial-scm.org/D12356

pullbundle: fix file name in the help text It is pullbundles.manifest and not pullbundle.manifest. Differential Revision: https://phab.mercurial-scm.org/D12391

unamend: abort if commit was not created by `hg [un]amend` `hg unamend` can currently undo any kind of rewrite, as long as it has an obsmarker. However, that has quite unexpected results if you run it after e.g. `hg rebase` (expecting it to behave like a generic `hg undo` command), because it updates to the predecessor and leaves the old changes in the working copy. I think it's better to allow `hg unamend` only after `hg amend` (and after `hg unamend` because that's documented as being supported). Differential Revision: https://phab.mercurial-scm.org/D12390

branching: merge stable into default

test: use `wait-on-file` in `test-racy-mutations.t` The official utility scale its timeout with the run-tests.py one. So lets use it. Differential Revision: https://phab.mercurial-scm.org/D12382

amend: fix amend with copies in extras If copy information is stored only in the commit extras and not in filelogs, then they get lost on amend if the file wasn't also modified in the working copy. That's because we create `filectx` object from the old commit in those cases, and the `.copysource()` of such objects read only from the filelog. This patch fixes it by always creating a new `memfilectx` in these cases, passing the calculated copy information to it. Differential Revision: https://phab.mercurial-scm.org/D12387

tests: demonstrate that copy info in changeset gets lost on amend When copy information is stored in changesets, it gets lost on amend. We didn't notice that until now because our users at Google have the config set to `compatibility`, which means copy information is stored in both changeset and filelogs. Differential Revision: https://phab.mercurial-scm.org/D12386

ci: use the `v1.0` flavor of the docker images in the CI This new versioning will help us to maintain backward compatibility in the docker image. This will be useful to deal with mismatch between default/stable in version and the re-run CI on older changesets in the future. Once this changeset land on stable, we will have to merge it in default. Then we can start make backward incompatible changes in a new image version. Differential Revision: https://phab.mercurial-scm.org/D12388

rust-status: cap the number of concurrent threads to 16 During benchmarking it was determined that the use of more threads is very advantageous... until we use more than 16. This is most likely due to some resource contention (thrashing, etc.). Until we have time to figure out and fix the underlying cause, let's just cap at 16 threads. Differential Revision: https://phab.mercurial-scm.org/D12384

rhg: add support for ignoring all extensions Some workflows just want what `rhg` does and don't care about any extensions, this makes it easier. Differential Revision: https://phab.mercurial-scm.org/D12385

branching: merge stable into default

revlog: fix index_fast_rank (wip) As far as I can tell, rank is stored as a 32-bit big endian value, I'm not sure how grabbing the first byte can possibly work. I assume there's no test coverage here? cc @pacien Fixes: https://www.mercurial-scm.org/repo/hg/rev/e633e660158f Differential Revision: https://phab.mercurial-scm.org/D12376

tests: fix glob pattern for dynamic timer alignment The number of space characters varies depending on the number of digits of the timer, making some tests fail on slow machines in an unintended way: ```diff --- /build/mercurial-6.1/tests/test-merge-halt.t +++ /build/mercurial-6.1/tests/test-merge-halt.t.err @@ -210,6 +210,6 @@ merge halted after failed merge (see hg resolve) [240] $ hg shelve --list - default (* ago) changes to: foo (glob) + default (11s ago) changes to: foo $ hg unshelve --abort unshelve of 'default' aborted ERROR: test-merge-halt.t output changed ``` Differential Revision: https://phab.mercurial-scm.org/D12381

test: update test-clone-stream.t to pass on bigendian Fixes: a3cf460a6b1b ("stream-clone: also filter the requirement we put in the bundle 2") Differential Revision: https://phab.mercurial-scm.org/D12377

filemerge: when merge tool uses $output, don't leave markers in $local As explained in the previous patch, we incorrectly leave conflict markers in both `$local` and `$output` since D12190. I don't understand why it broke but the fix is simple and clear after all the recent refactoring. Differential Revision: https://phab.mercurial-scm.org/D12379

tests: demonstrate how conflict markers end up $local *and* $output When a merge tool is configured to keep conflict markers, they are supposed to be written to `$local` if `$output` is not mentioned in the tool's `merge-tools.<tool>.args` config, and in `$output` if it is mentioned. However, I broke the latter case in D12190. Differential Revision: https://phab.mercurial-scm.org/D12378

branching: merge stable into default

revlog: fix wrong type of rank_unknown variable We treat "rank" as an int everywhere, but declare rank_unknown as a char. On architectures where char is signed, that works out ok, but when char is unsigned, rank_unknown is 255 instead of -1. Differential Revision: https://phab.mercurial-scm.org/D12374

tests: fix formatting issue in run-tests.py after c194e93d1ebc Differential Revision: https://phab.mercurial-scm.org/D12375

rust-hg-core: use correct type for libc hostname buffer The type of libc::c_char is u8 on aarch64 rather than i8, which causes the use of a specifically-typed constant to fail. Differential Revision: https://phab.mercurial-scm.org/D12373

import-checker: allow symbol imports from typing module As we add typing annotations, we'll want to use a lot of symbols from the `typing` module. Typing `typing` all the time will be annoying. Let's allow symbol imports from this module. While I was here, I changed some comments from "whitelist" to "allow list" as the former is non-inclusive terminology. Differential Revision: https://phab.mercurial-scm.org/D12365

pycompat: remove json.loads polyfill for Python 3.5 We no longer support Python 3.5 so this can be deleted. Differential Revision: https://phab.mercurial-scm.org/D12364

pycompat: remove check for Python >= 3.6 We dropped support for Python 3.5 so this is always true. Differential Revision: https://phab.mercurial-scm.org/D12363

hgdemandimport: delete check for Python 3.5 We dropped support for Python 3.5. So we no longer need to do this. Differential Revision: https://phab.mercurial-scm.org/D12362

hg: always import hgdemandimport The deleted if condition is always true now that we dropped Python 2 and 3.5. Differential Revision: https://phab.mercurial-scm.org/D12361

tests: support another error case when detecting ipv4/ipv6 support I encountered this on Linux in a VM environment with a rather strange networking setup (both on the host and in the VM). Differential Revision: https://phab.mercurial-scm.org/D12371

debugdiscovery: fix a typo in the help Differential Revision: https://phab.mercurial-scm.org/D12372

rust-hg-cypython: upgrade dependencies This upgrades all dependencies to their latest version. This is routinely done to keep-up. Differential Revision: https://phab.mercurial-scm.org/D12359

rust-hg-core: upgrade dependencies This upgrades all dependencies to their latest version, except `clap` and `zstd` whose latest versions do not support our minimum supported Rust version 1.48.0. Same as for `rhg`, it contains security fix for `regex` which does not affect us too much, but doesn't hurt, and the rest of the upgrades are there simply to keep up. Differential Revision: https://phab.mercurial-scm.org/D12358

rhg: upgrade dependencies This upgrades all dependencies to their latest version, except `clap`, which is upgraded to the last 2.x series version, since 3.x does not support our minimum supported Rust version of 1.48.0. This contains a security fix for `regex` which does not affect us too much, but doesn't hurt, and the rest of the upgrades are there simply to keep up. Differential Revision: https://phab.mercurial-scm.org/D12357

chgserver: remove Python 2 file descriptor logic Follows up 0bb28b7736bc "chgserver: remove Python 2 support code." On Python 2, we had to close newfp prior to restoring the original file description since "delete newfp" would otherwise close the file descriptor shared with the long-lived fp: in attachio(): newfp = os.fdopen(fp.fileno(), mode, bufsize) in _restoreio(): newfp.close() # temporarily close newfp.fileno() (= fp.fileno()) os.dup2(fd, fp.fileno()) # reopen fp.fileno() with original fd On the other hand, we shouldn't call newfp.close() on Python 3 since any function calls are proxied to the underlying file object by procutil.LineBufferedWrapper.

dirstate-tree: optimize HashMap lookups with raw_entry_mut This switches to using `HashMap` from the hashbrown crate, in order to use its `raw_entry_mut` method. The standard library’s `HashMap` is also based on this same crate, but `raw_entry_mut` is not yet stable there: https://github.com/rust-lang/rust/issues/56167 Using version 0.9 because 0.10 is yanked and 0.11 requires Rust 1.49 This replaces in `DirstateMap::get_or_insert_node` a call to `HashMap<K, V>::entry` with `K = WithBasename<Cow<'on_disk, HgPath>>`. `entry` takes and consumes an "owned" `key: K` parameter, in case a new entry ends up inserted. This key is converted by `to_cow` from a value that borrows the `'path` lifetime. When this function is called by `Dirstate::new_v1`, `'path` is in fact the same as `'on_disk` so `to_cow` can return an owned key that contains `Cow::Borrowed`. For other callers, `to_cow` needs to create a `Cow::Owned` and thus make a costly heap memory allocation. This is wasteful if this key was already present in the map. Even when inserting a new node this is typically the case for its ancestor nodes (assuming most directories have numerous descendants). Differential Revision: https://phab.mercurial-scm.org/D12317

heptapod-ci: remove useless mentions of Python 3 Now that we don't have Python 2 jobs, we can go with shorter names. Differential Revision: https://phab.mercurial-scm.org/D12354

revlog: remove unused `rank` variable from D12209 Differential Revision: https://phab.mercurial-scm.org/D12355

demandimport: delete demandimportpy2 module We no longer support Python 2. Differential Revision: https://phab.mercurial-scm.org/D12353

py3: use class X: instead of class X(object): The inheritance from object is implied in Python 3. So this should be equivalent. This change was generated via an automated search and replace. So there may have been some accidental changes. Differential Revision: https://phab.mercurial-scm.org/D12352

check-code: remove ban on old-style classes In Python 3, `class foo:` is equivalent to `class foo(object):`. So we can allow the former form now. Differential Revision: https://phab.mercurial-scm.org/D12351

tests: remove Python 2 support code Differential Revision: https://phab.mercurial-scm.org/D12350

tests: remove Python < 3 check We require Python 3 now. Differential Revision: https://phab.mercurial-scm.org/D12349

cborutil: remove Python 2 definition of _elementtointeger() Differential Revision: https://phab.mercurial-scm.org/D12348

py3: stop using util.iterfile() The Python 3 implementation is a no-op. So this is equivalent. We still keep util.iterfile() around for backwards API compatibility to help the Python 3 migration. It can be deleted in a future release. Differential Revision: https://phab.mercurial-scm.org/D12347

util: remove iterfile() variant for buggy EINTR handling The workaround for Python 2 is no longer needed. So we can delete some code. Differential Revision: https://phab.mercurial-scm.org/D12346

statprof: remove superfluous sys.version_info check Always true since we require Python 3 now. Differential Revision: https://phab.mercurial-scm.org/D12345

policy: remove Python 2.7 compatibility code Differential Revision: https://phab.mercurial-scm.org/D12344

lsprof: remove some Python 2.7 compatibility code Differential Revision: https://phab.mercurial-scm.org/D12343

url: remove Python 2.7 support code Differential Revision: https://phab.mercurial-scm.org/D12342

py3: replace pycompat.itervalues(x) with x.values() pycompat.itervalues(x) just calls x.values(). So this is equivalent. The rewrite was perfomed via an automated search and replace. Differential Revision: https://phab.mercurial-scm.org/D12341

py3: use str instead of pycompat.unicode pycompat.unicode is an alias to str. Differential Revision: https://phab.mercurial-scm.org/D12340

ui: use input() directly pycompat.rawinput() is an alias to input() Differential Revision: https://phab.mercurial-scm.org/D12339

py3: use int instead of pycompat.long pycompat.long is aliased to int. So this should have no change in functionality. Differential Revision: https://phab.mercurial-scm.org/D12338

verify: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12336

templatekw: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12335

templatefuncs: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12334

revlog: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12333

merge: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12332

manifest: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12331

localrepo: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12330

keepalive: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12329

help: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12328

dirstate: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12327

debugcommands: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12326

config: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12325

commands: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12324

remotefilelog: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12323

lfs: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12322

largefiles: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12321

infinitepush: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12320

convert: remove pycompat.iteritems() Differential Revision: https://phab.mercurial-scm.org/D12319

global: bulk replace simple pycompat.iteritems(x) with x.items() pycompat.iteritems() just calls .items(). This commit applies a regular expression search and replace to convert simple instances of pycompat.iteritems() with .items(). There are still a handful of calls to pycompat.iteritems() remaining. But these all have more complicated expressions that I wasn't comfortable performing an automated replace on. In addition, some simple replacements were withheld because they broke pytype. These will be handled by their own changesets. Differential Revision: https://phab.mercurial-scm.org/D12318

worker: silence type error when calling pickle pytype is complaining that the argument to `pickle.load()` is not an `IO`. pytype isn't wrong: `_blockingreader` doesn't implement `io.RawIOBase`, only `read()` and `readline()`. But it appears this is enough for pickle. So we silence the false positive. This fixes a regression introduced by D12304 / cc0e059d2af8: worker: remove Python 2 support code. Differential Revision: https://phab.mercurial-scm.org/D12337

stringutil: remove Python 2 support code Differential Revision: https://phab.mercurial-scm.org/D12314

procutil: delete Python 2 support code This entailed deleting a function for Python 2 support and renaming the Python 3 function to match the exported symbol name. Differential Revision: https://phab.mercurial-scm.org/D12313

revlogutils: unconditionally pass version to random seed Differential Revision: https://phab.mercurial-scm.org/D12312

revlogutils: remove Python 2 variant for iter_seed Differential Revision: https://phab.mercurial-scm.org/D12311

charencode: remove Python 2 support code Differential Revision: https://phab.mercurial-scm.org/D12310

hgweb: remove Python 2 support code Differential Revision: https://phab.mercurial-scm.org/D12309

hgweb: remove Python 3 conditional We probably have a better tobytes() implementation somewhere in pycompat. But I don't want to bloat scope of this commit. Differential Revision: https://phab.mercurial-scm.org/D12308

hgweb: simplify uenv assignment We don't need the Python 3 conditional. We can call items() directly since we're on Python 3 now. Differential Revision: https://phab.mercurial-scm.org/D12307

chgserver: remove Python 2 support code The logic here is more complicated than most Python 2/3 support code. But the rewritten logic should be identical. Differential Revision: https://phab.mercurial-scm.org/D12306

chgserver: remove Python 2 branch Differential Revision: https://phab.mercurial-scm.org/D12305

worker: remove Python 2 support code Differential Revision: https://phab.mercurial-scm.org/D12304

wireprotoframing: remove Python 2 support code Differential Revision: https://phab.mercurial-scm.org/D12303

windows: remove write throttling support This mode would only be active on Python 2, which is no longer supported. Differential Revision: https://phab.mercurial-scm.org/D12302

windows: remove conditional for Python 3 Differential Revision: https://phab.mercurial-scm.org/D12301

util: remove superfluous ispy3 test Differential Revision: https://phab.mercurial-scm.org/D12300

urllibcompat: remove Python 2 support code We had to move the `import` statements to appease the import checker. This whole module could probably be deleted as its point in life is to pave over Python 2/3 differences. But that's for a different commit. Differential Revision: https://phab.mercurial-scm.org/D12299

keepalive: remove Python 2 support code Differential Revision: https://phab.mercurial-scm.org/D12298

extensions: remove superfluous pycompat.ispy3 check This is always True now. Differential Revision: https://phab.mercurial-scm.org/D12297

archival: remove check for Python 2 Differential Revision: https://phab.mercurial-scm.org/D12296

encoding: remove Python 2 support code Differential Revision: https://phab.mercurial-scm.org/D12295

dispatch: remove Python 2 function variants Differential Revision: https://phab.mercurial-scm.org/D12294

config: remove conditional asserts We always run on Python 3 now. Differential Revision: https://phab.mercurial-scm.org/D12293

error: unconditionally define __str__ We always run on Python 3 now. Differential Revision: https://phab.mercurial-scm.org/D12292

formatter: remove conditional assert We always run on Python 3 now. Differential Revision: https://phab.mercurial-scm.org/D12291

httppeer: inline simplified _reqdata() The function can be reduced to an attribute lookup on Python 3. So inline it. Differential Revision: https://phab.mercurial-scm.org/D12290

url: remove passing of strict This was needed to support Python 2.7. Differential Revision: https://phab.mercurial-scm.org/D12289

posix: delete Python 2 posixfile() The comment no longer makes sense since the stdlib open() behaves the way we want on Python 3. So it was removed. Differential Revision: https://phab.mercurial-scm.org/D12288

match: delete Python 2 conditional code Differential Revision: https://phab.mercurial-scm.org/D12287

mail: delete conditional code for Python 2 Differential Revision: https://phab.mercurial-scm.org/D12286

archival: remove GzipFileWithTime This was required for Python 2 support, which we no longer need to support. Differential Revision: https://phab.mercurial-scm.org/D12285

check-code: allow importing Python 3 modules Now that we no longer support Python 2, we should be able to import and use the Python 3 only modules in our code. So remove a lint banning this. Differential Revision: https://phab.mercurial-scm.org/D12284

tests: move Python 3.5 check higher in file Per code review comment on the changeset that introduced the Python 3.5+ checks. Differential Revision: https://phab.mercurial-scm.org/D12259

py2: drop some more from __future__ statements These are no longer needed after dropping support for Python 2. Differential Revision: https://phab.mercurial-scm.org/D12258

tests: delete some no-py3 blocks And drop some nearby avoidable py3 checks in close proximity while we are here. Differential Revision: https://phab.mercurial-scm.org/D12257

tests: remove __future__ import from test-debugcommands.t This one is slightly more involved since it affects test output. Differential Revision: https://phab.mercurial-scm.org/D12256

tests: remove from __future__ from inline Python in tests This is no longer required since we require Python 3 and the linter no longer requires these statements. Differential Revision: https://phab.mercurial-scm.org/D12255

py2: remove simple from __future__ statements These were needed for Python 2 support. Now that our linter no longer mandates these, we can start deleting them. Differential Revision: https://phab.mercurial-scm.org/D12254

tests: remove output conditionalized on no-py3 I simply did a search for `^.* \(no-py3 !\)\n` and removed all matched lines. There are still some references to no-py3. But these were the simpler ones to match against. Differential Revision: https://phab.mercurial-scm.org/D12253

py3: use io.BytesIO directly Previously, pycompat.bytesio and pycompat.stringio referred to io.BytesIO. And util.bytesio and util.stringio aliased the pycompat symbols. This commit switches everything to use io.BytesIO directly. util.bytesio and util.stringio still exist to provide backwards compatibility, as they were the preferred symbols. Differential Revision: https://phab.mercurial-scm.org/D12252

check-py3-compat: drop support for Python 2 We no longer support Python 2 so we can drop support for linting code for Python 2 support. This gets rid of the check for `from __future__`, enabling us to delete those imports. Differential Revision: https://phab.mercurial-scm.org/D12251

pycompat: remove large Python 2 block We no longer support Python 2. So we can delete its compatibility code and remove the conditional and dedent the Python 3 code. In order to make the linter happy, we had to inline imports in the stanza at the top of the file. Differential Revision: https://phab.mercurial-scm.org/D12250

py3: use pickle directly pycompat.pickle abstracted over the different pickle modules in Python 2 and 3. Now that we're Python 3 only, it is safe to use the `pickle` module directly. So this commit does that. As part of this we remove the rules from check-code.py that were forbidden direct pickle module use. We retain the `util.pickle` symbol for backwards compatibility, just in case some extensions were using it. Differential Revision: https://phab.mercurial-scm.org/D12249

pycompat: remove first not ispy3 block We now require Python 3. So we can remove the first block supporting Python 2. Differential Revision: https://phab.mercurial-scm.org/D12247

import-checker: assume absolute and use modern import checker Since we require Python 3 now, we can assume we always use absolute imports and the modern import checker should be used. Differential Revision: https://phab.mercurial-scm.org/D12246

tests: unconditionalize some imports Now that we require Python 3 we can simplify these imports. Differential Revision: https://phab.mercurial-scm.org/D12245

tests: remove last references to PYTHON3 This removes the last references to PYTHON3. Differential Revision: https://phab.mercurial-scm.org/D12244

tests: simplify Windows and PYTHON3 conditionals PYTHON3 is always True. So this flow can be reduced. Differential Revision: https://phab.mercurial-scm.org/D12243

tests: delete some not PYTHON3 blocks These can never be used anymore. Differential Revision: https://phab.mercurial-scm.org/D12242

tests: unconditionalize _bytes2sys() As part of requiring Python 3. Differential Revision: https://phab.mercurial-scm.org/D12241

tests: remove Python 3 conditionalizing from variables Differential Revision: https://phab.mercurial-scm.org/D12240

tests: always encode session Differential Revision: https://phab.mercurial-scm.org/D12239

tests: unconditionalize bchr We could probably just do bytes([x]) everywhere. But this eliminates use of PYTHON3. Differential Revision: https://phab.mercurial-scm.org/D12238

tests: unconditionalize _unified_diff Now that we're Python 3 only we can make this logic simpler. Differential Revision: https://phab.mercurial-scm.org/D12237

tests: collapse elif PYTHON3 block PYTHON3 is always True now so this logic should be identical as to before. Differential Revision: https://phab.mercurial-scm.org/D12236

tests: collapse some more trivial if PYTHON3 blocks This doesn't account for all of the references to PYTHON3. But it accounts for the ones that are more trivial and don't entail logical changes. Differential Revision: https://phab.mercurial-scm.org/D12235

tests: require Python 3.5+ in run-tests.py We change the version check logic to hard fail if running on <= 3.5.0. The branch for <3.5 has been deleted. And the >=3.5 branch block has been dedented. Differential Revision: https://phab.mercurial-scm.org/D12234

revlog: use rust rank computation if available Differential Revision: https://phab.mercurial-scm.org/D12212

rust: expose rank computation function to python Differential Revision: https://phab.mercurial-scm.org/D12211

rust: implement vcsgraph::RankedGraph for Index Differential Revision: https://phab.mercurial-scm.org/D12210

revlog: implement fast_rank retrieval in C This will be useful in particular to avoid going through the Python interpreter in native Rust functions. Differential Revision: https://phab.mercurial-scm.org/D12209

revlog: return 0 for the fast_rank of nullrev By convention, the rank of the null revision is 0. This particular revision is never "physically" stored in the changelog, so it is a special case. For consistency, the value `None` is still being returned for revlogs which do not store the fast_rank property for any revision. Differential Revision: https://phab.mercurial-scm.org/D12208

setup: drop support for Python 3.5 We talked about this on the mailing list [1] and there seemed to be agreement that Python 3.5 is effectively dead and no longer worth supporting. So this commit changes our minimum version requirement to 3.6.2. [1] https://www.mercurial-scm.org/pipermail/mercurial-devel/2022-February/147885.html Differential Revision: https://phab.mercurial-scm.org/D12315

packaging: remove requirements constraints to support Python 2 We just deleted support for Python 2 from the packaging code. We no longer need these package constraints in the requirements file to support Python 2. Differential Revision: https://phab.mercurial-scm.org/D12267

packaging: remove py2exe / Python 2.7 support This commit started by deleting references to py2exe (which is only used on Python 2). After pulling the thread, quite a lot of code was orphaned and was deleted. Differential Revision: https://phab.mercurial-scm.org/D12265

automation: delete code related to Python 2.7 support The building of Inno and WiX installers took a python_version argument that allowed us to specify "2" or "3" for the major Python version. Since we no longer support Python 2, we can delete this argument and everything feeding into it. Differential Revision: https://phab.mercurial-scm.org/D12264

automation: drop support for Python 2.7 in Windows environment We stop installing Python 2.7 in the Windows environment. We remove support for building Python 2.7 wheels and installers. There is still some Python 2.7 support cleanup to perform in automation. But this removes the biggest remaining chunk of references to 2.7. Differential Revision: https://phab.mercurial-scm.org/D12263

automation: drop support for Python 2.7 in Linux environment We stop installing Python 2.7 via pyenv. We stop installing the system Python 2 packages. We delete support for running tests on Python 2.7. Differential Revision: https://phab.mercurial-scm.org/D12262

automation: make system3 the default for run-tests-linux We'll soon drop support for Python 2.7. Let's use Python 3 by default. Differential Revision: https://phab.mercurial-scm.org/D12261

automation: run hg with python3 Python 2.7 support will go away soon. Let's use Python 3 as part of the automation. Differential Revision: https://phab.mercurial-scm.org/D12260

rust: enable Python 3 support unconditionally Note: `cpython/python3-sys` is a default feature. Differential Revision: https://phab.mercurial-scm.org/D12316

cext: really remove Python 2 file handling code Disclaimer: This is _WIN32 code and I have no machine to test.

cext: backout e9ca736f5b52 "remove Python 2 file handling code" It's if"n"def.

rust: jettison Python 2 support Differential Revision: https://phab.mercurial-scm.org/D12283

setup: always decode xcode version Not decoding was a Python 2 thing. Differential Revision: https://phab.mercurial-scm.org/D12282

setup: remove Rust support for Python 2 Differential Revision: https://phab.mercurial-scm.org/D12281

cleanup: stop bundling concurrent.futures on Python 2 We no longer support Python 2. Differential Revision: https://phab.mercurial-scm.org/D12280

cleanup: directly use concurrent.futures instead of via pycompat Python 2 is gone. Differential Revision: https://phab.mercurial-scm.org/D12279

imports: allow importing futures from concurrent Differential Revision: https://phab.mercurial-scm.org/D12278

setup: remove pygit2 Python 2 logic Differential Revision: https://phab.mercurial-scm.org/D12277

setup: inline now-constant list This varied when we supported Python 2. Differential Revision: https://phab.mercurial-scm.org/D12276

setup: unconditionally do this python 3 step Differential Revision: https://phab.mercurial-scm.org/D12275

setup: remove Python 2 support code for determining dylib suffix Differential Revision: https://phab.mercurial-scm.org/D12274

setup: inline os.fsdecode now that we're done with Python 2 Differential Revision: https://phab.mercurial-scm.org/D12273

setup: inline encoding constant that is only used once This was variable back when we supported Python 2. Differential Revision: https://phab.mercurial-scm.org/D12272