py3: use native string as fallback value for __file__ for consistency This is not a bugfix (pycommpat.fsencode(b'') is a no-op on py3), but the b'' value was inconsistent and confusing. Differential Revision: https://phab.mercurial-scm.org/D7412

scmutil: convert status data object from a tuple to an attrs (API) We've been pushing towards the property names for a while, and the subclassing of the tuple confuses pytype. Rather than bend over backwards to try and annotate the tuple subclass, let's just use attrs here. Differential Revision: https://phab.mercurial-scm.org/D7406

perf: bool() elements of dirstate.status return instead of len() I'm about to make scmutil.status no longer have a len(), so we need to do something else to "use" the results in this perf method. Differential Revision: https://phab.mercurial-scm.org/D7405

tags: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7404

subrepo: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7403

revset: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7402

patch: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7401

logcmdutil: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7400

context: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7399

commands: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7398

cmdutil: convert terse status result back to an scmutil.status This is also part of detupling scmutil.status. Differential Revision: https://phab.mercurial-scm.org/D7397

cmdutil: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7396

uncommit: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7395

transplant: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7394

split: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7393

mq: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7392

hgk: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7391

extdiff: use field names instead of field numbers on scmutil.status As part of my pytype adventures I want to make scmutil.status no longer a subclass of tuple. This is part of that process. Differential Revision: https://phab.mercurial-scm.org/D7390

encoding: fix bad type annotation This function returns utf-8 in a bytes, not a unicode. Differential Revision: https://phab.mercurial-scm.org/D7379

branchmap: annotate constructor type for branchcache This type signature is...big. But it's correct as far as I can tell, and it detected a bug. Differential Revision: https://phab.mercurial-scm.org/D7389

debugcommands: don't shadow the error module Caught by pytype. Differential Revision: https://phab.mercurial-scm.org/D7378

branchmap: correctly set()-ify list argument Caught with pytype. I'm more than a little curious how this never caused problems. Differential Revision: https://phab.mercurial-scm.org/D7289

utils: move finddirs() to pathutil This is a follow-up to c21aca51b392 (utils: move the `dirs` definition in pathutil (API), 2019-11-06). finddirs() is closely related to dirs and used by it. Differential Revision: https://phab.mercurial-scm.org/D7388

help: create packages for the help text These files need to be loaded as resources with PyOxidizer, instead of using filesystem representations. AFAICT, the resource loading mechanisms only work for the named package given to it, and can't reach into a subdirectory. While here, the `help` directory is renamed to `helptext`. Without this, trying to load external help text crashed in mercurial/help.py when importing `.i18n`, saying there's no `mercurial.help.i18n` module. Differential Revision: https://phab.mercurial-scm.org/D7376

tests: test "hg log" with --line-range and --copies This now works (does not crash), due to previous changeset. Since --line-range implies --follow, --copies option is redundant.

logcmdutil: let getlinerangerevs() return "revs" as a smartset This makes it consistent in "hg log" command where 'revs' can come from logcmdutil.getrevs() as a smartset or from getlinerangerevs(), previously as a list. This will help type hinting as noticed in https://phab.mercurial-scm.org/D7377.

patchbomb: fix wrong argument type when calling mail generator.flatten()

mail: move strtolocal call in _addressencode()

mail: use a native string for "subtype" value This is somehow similar to previous changeset and avoids one str conversion.

mail: let all charset values be native strings Charset values will typically be used to build email.header.Header instances, which takes str (though it tolerates bytes) or passed to decode()/encode() methods of string values (which want str). It seems that using native str involves less conversions than before and this also helps type hinting (as illustrates removal of pytype disabling instructions).

mail: add type hints for pytype We essentially annotate functions in which handling of bytes/str is not obvious in order to hopefully clear things out. See also changeset 2ade00f3b03b introducing typing hints in Mercurial. Most types are straightforward but a few is wrong, and we need to either disable pytype on respective instructions or use wrong annotations. These will be fixed in next changesets. Notice the type Union[bytes, str] of "s" parameter of headencode(), this reflects how email.header.Header.append() behaves.

mail: fix a bad return type in _encode() This particular instruction returned only a string and omitted the charset value.

packaging: stage files and dynamically generate WiX installer Like we did for Inno, we want to make the WiX installer "dumb" and simply consume source files from a directory tree rather than have to define every single file in installer files. This will greatly decrease the amount of effort required to maintain the WiX installer since we don't have to think that much about keeping files in sync. This commit changes the WiX packager to populate a staging directory as part of packaging. After it does so, it scans that directory and dynamically generates WiX XML defining the content within. The IDs and GUIDs being generated are deterministic. So, upgrades should work as expected in Windows Installer land. (WiX has a "heat" tool that can generate XML by walking the filesystem but it doesn't have this deterministic property, sadly.) As part of this change, GUIDs are now effectively reset. So the next upgrade should be a complete wipe and replace. This could potentially cause issues. But in my local testing, I was able to upgrade an existing 5.1.2 install without issue. Compared to the previous commit, the installed files differ in the following: * A ReleaseNotes.txt file is now included * A hgrc.d/editor.rc file is now generated (mercurial.rc has been updated to reflect this logical change to the content source) * All files are marked as read-only. Previously, only a subset of files were. This should help prevent unwanted tampering. Although we may want to consider use cases like modifying template files... This change also means that Inno and WiX are now using very similar code for managing the install layout. This means that on disk both packages are nearly identical. The differences in install layout are as follows: * Inno has a Copying.txt vs a COPYING.rtf for WiX. (The WiX installer wants to use RTF.) * Inno has a Mercurial.url file that is an internet shortcut to www.mercurial-scm.org. (This could potentially be removed.) * Inno includes msvc[mpr]90.dll files and WiX does not. (WiX installs the MSVC runtime via merge modules.) * Inno includes unins000.{dat,exe} files. (WiX's state is managed by Windows Installer, which places things elsewhere.) Because file lists are dynamically generated now, the test ensuring things remain in sync has been deleted. Good riddance. While this is a huge step towards unifying the Windows installers, there's still some improvements that can be made. But I think it is worth celebrating the milestone of getting both Inno and WiX to essentially share core packaging code and workflows. That should make it much easier to change the installers going forward. This will aid support of Python 3. Differential Revision: https://phab.mercurial-scm.org/D7173

tests: add optional output when wheel is missing If the `wheel` package isn't installed, pip will emit a warning on stderr complaining about that. Alternatively, we could redirect stderr to stdout and swallow the warning. But I like tests having visibility of warnings, as it helps keep us honest regarding edge cases. Differential Revision: https://phab.mercurial-scm.org/D7371

py3: replace "%r" by"'%s'% for py3-compatible (and clearer) quoting in chg Differential Revision: https://phab.mercurial-scm.org/D7375

dirstate: respect request to not list unknown/ignored/clean files (API) Unknown files that are explicitly mentioned by the matcher are returned even if the caller said unknown=False (and it seems the same is done for ignored files). That seems pretty surprising. Let's make the interface less surprising by respecting the caller's request. Differential Revision: https://phab.mercurial-scm.org/D7150

py3: pass a bytes value for "msg" to nouideprecwarn() That function formats "msg" with the "version" value. On Python 3, this leads to "TypeError: can only concatenate str (not "bytes") to str". Also eliminate spurious strings concatenation in single-line declarations. Differential Revision: https://phab.mercurial-scm.org/D7373

remotefilelog: handle **kwargs correctly when overriding changelog.add() Differential Revision: https://phab.mercurial-scm.org/D7372

packaging: ship all help .txt files on WiX These are technically not needed. But it is easier to ship all files than to cherry-pick. A `make install` will copy these files, so the new behavior is consistent with that. This also makes WiX consistent with Inno, which is my main reason for doing this. If we don't want to ship the files (which is a valid argument), I think we can do that in a follow up. Differential Revision: https://phab.mercurial-scm.org/D7166

packaging: install contrib/mq.el on WiX This makes things consistent with the Inno installer. Differential Revision: https://phab.mercurial-scm.org/D7167

fuzz: clean out most of fuzzutil It's now a header-only setup that just selects absl:: or std:: versions of things as needed, and a logging helper. There's some room for future cleanups here: we could move to just requiring a C++17 compiler and get rid of the absl stuff. Also, the mpatch parser has a fair amount of parsing the input string into char* blocks that we can and probably should fix that up to use FuzzedDataProvider as well. Differential Revision: https://phab.mercurial-scm.org/D7370

fuzz: migrate xdiff fuzzer to use FuzzedDataProvider Differential Revision: https://phab.mercurial-scm.org/D7369

fuzz: migrate bdiff fuzzer to use FuzzedDataProvider Differential Revision: https://phab.mercurial-scm.org/D7368

dirstate: re-blacken file This got overlooked in landing a recent patch, but it's easy to fix up. Differential Revision: https://phab.mercurial-scm.org/D7367

rust-matchers: remove default implementations for `Matcher` trait We don't expect a whole lot of matchers to be defined, and this makes it more obvious what a matcher does by reading its `impl Matcher for FooMatcher`. This patch has the added benefit of fixing the `AlwaysMatcher`, its `matches` function differs from the former default. Differential Revision: https://phab.mercurial-scm.org/D7255

packaging: use lower case for RC files The file case doesn't matter. I'm not sure why it is capitalized. This is a source of inconsistency between Inno and WiX. So let's standardize on the lower case names. The MSI should remove the old files before installing the new one, so there shouldn't be a problem with a case collision. And, users should not have modified these files, so it should be safe to rename them. Differential Revision: https://phab.mercurial-scm.org/D7172

packaging: write out editor.rc in Python Let's dynamically generate the file from Python instead of relying on Inno functionality, as that will give more visibility into what's going on. I spotted this as part of converging the install layout of Inno and WiX: the former doesn't install an editor.rc and instead defines the editor via its mercurial.rc file. We will change the behavior of WiX later once it is consuming the staged install directory for packaging. Differential Revision: https://phab.mercurial-scm.org/D7171

packaging: stop installing Contributors.txt This file isn't up to date. If we are going to call out contributors, we should auto generate a file from `hg log` or something. I don't think we need to include it in the Windows install. This change makes Inno more consistent with WiX, which doesn't ship this file. Differential Revision: https://phab.mercurial-scm.org/D7170

packaging: remove hg-ssh.8.html from Inno installer We don't ship hg-ssh because it requires a python.exe to run, which we don't ship. So it doesn't make sense to ship the HTML documentation for this tool. This change makes the Inno install layout more consistent with WiX, which doesn't ship this file. Functionality for removing files has been made generic, in anticipation of future expansion. Differential Revision: https://phab.mercurial-scm.org/D7169

packaging: install .rc files to hgrc.d Currently, Inno installs these .rc files to a default.d directory and WiX to a hgrc.d directory. Mercurial automatically loads files in an hgrc.d directory next to the executable and in an "internal" default.d directory. This "internal" directory is util.datadir, which on Windows installs made with py2exe is the directory containing hg.exe. So on Windows both hgrc.d and default.d directories relative to hg.exe are loaded. For the install layout, I think it makes sense for hgrc.d to be used, as that path is documented as non-internal. So this commit changes the Inno install layout to use hgrc.d instead of default.d. .. bc:: Changed default config file location in Windows .exe installers Mercurial's installer installs some default .rc config files. Previously, these were installed in a default.d directory. These are now installed in an hgrc.d directory. Custom default.d/*.rc files should still be loaded. But it is recommended to move these to hgrc.d/*.rc. This change only affects the .exe installers (not the .msi installers). Differential Revision: https://phab.mercurial-scm.org/D7168

rust: introduce SIZE_FROM_OTHER_PARENT constant This will clarify what `-2` meant in the codebase. This change also merged imports in affected files for cleanup. Differential Revision: https://phab.mercurial-scm.org/D7301

rust-status: refactor dispatch case for normal files This should make the code easier to read and more idiomatic. Differential Revision: https://phab.mercurial-scm.org/D7300

rust-status: return a ParallelIterator instead of a Vec from stat_dmap_entries This allows the caller function to choose when and how the iteration and/or collection happens. This change also cleans up the now unused `filter_map`. Differential Revision: https://phab.mercurial-scm.org/D7299

rust-status: improve status performance This change does more things in the parallel loop, refactors the file-level logic into two functions for added clarity. This bit of Rust code takes 55ms to execute on a repo where the stat'ing part of Valentin's fast path takes 40ms. While the code differs a bit and it's hard to get an exact measurement of how much of a performance impact it has, I can be fairly certain that this implementation is *at worse* twice as slow. Differential Revision: https://phab.mercurial-scm.org/D7254

bookmarks: accept explicit -r 'wdir()' when adding new bookmarks (issue6218) Even though the bookmark semantics can't be fully encoded to the virtual working changeset idea, the active bookmark can be considered a bookmark of the working revision. Before, 'tgt' was None, and changes=[(bm, None)] means deleting a bookmark named 'bm'.

bookmarks: use changectx instead of remembering hex of hidden revision It should be better to not depend on the ctx variable which was assigned conditionally.

bookmarks: resolve target revision out of the bookmarks loop The target revision doesn't depend on the bookmark to be added.

bookmarks: fix handling of multiple bookmarks with one to be deactivated Before, "hg bookmark --inactive Z Y" would ignore "Y" if "Z" were currently active. I'm pretty sure it is a bug.

py3: avoid iterating over a literal bytes in highlight In Python 3, iterating over a bytes literal yields integers. Since we use the value in `text.replace()`, this fails on Python 3 with the following trackback: Traceback (most recent call last): File "/usr/lib/python3/dist-packages/mercurial/hgweb/hgwebdir_mod.py", line 378, in run_wsgi for r in self._runwsgi(req, res): File "/usr/lib/python3/dist-packages/mercurial/hgweb/hgweb_mod.py", line 326, in run_wsgi for r in self._runwsgi(req, res, repo): File "/usr/lib/python3/dist-packages/mercurial/hgweb/hgweb_mod.py", line 449, in _runwsgi return getattr(webcommands, cmd)(rctx) File "/usr/lib/python3/dist-packages/mercurial/hgweb/webcommands.py", line 211, in file return _filerevision(web, webutil.filectx(web.repo, web.req)) File "/usr/lib/python3/dist-packages/hgext/highlight/__init__.py", line 72, in filerevision_highlight pygmentize(web, b'fileline', fctx, web.tmpl) File "/usr/lib/python3/dist-packages/hgext/highlight/__init__.py", line 58, in pygmentize field, fctx, style, tmpl, guessfilenameonly=filenameonly File "/usr/lib/python3/dist-packages/hgext/highlight/highlight.py", line 62, in pygmentize text = text.replace(c, b'') TypeError: a bytes-like object is required, not 'int'

import-checker: allow 'from typing import ...' Suppresses the following error in test-check-module-imports.t: mercurial/encoding.py:24: relative import of stdlib module

match: drop support for passing '.' for root dir to visit*() methods We said we'd drop support for it after 5.1, so it's time to clean it up now. Differential Revision: https://phab.mercurial-scm.org/D7249

tests: use time.time() for relative start and stop times os.times() does not work on Windows. This was resulting in the test start, stop, and duration times being reported as 0. This commit swaps in time.time() for wall clock measurements. This isn't ideal, as time.time() is not monotonic. But Python 2.7 does not have a monotonic timer that works on Windows. So it is the best we have which is trivially usable. And test times aren't terribly important, so variances due to clock skew are arguably acceptable. Differential Revision: https://phab.mercurial-scm.org/D7126

tests: rename stopped and started variables to reflect times In preparation for introducing more variables that will have similar names. Differential Revision: https://phab.mercurial-scm.org/D7125

revset: simplify checkstatus() by using any() Differential Revision: https://phab.mercurial-scm.org/D7179

rebase: check for unfinished ops even when inmemory (issue6214) When using rebase.experimental.inmemory, we should be able to work well with a dirty working directory, but we can not reliably work if we're in the middle of another operation (such as another rebase), as we'll potentially stomp on some state that the other operation needs. Differential Revision: https://phab.mercurial-scm.org/D7298

status: move initialization closer together The default initialization happened long before it needed to happen. Differential Revision: https://phab.mercurial-scm.org/D7145

largefiles: avoid dynamically subclassing context instances E.g. copies.pathcopies() compares context objects for equality and basectx.__eq__ compares the type. But since largefiles was dynamically creating classes, they would all be unequal. That led pathcopies(), after some changes I made, to not short-circuit to get copy info only from the dirstate. This patch fixes that short-circuiting (and other places where context are being compared for equality). Differential Revision: https://phab.mercurial-scm.org/D7143

relnotes: copy "next" to "5.2" and clear "next" This is the same thing as we did for 5.1 in cba59b338976 (relnotes: copy "next" to "5.1" and clear "next", 2019-08-01). Differential Revision: https://phab.mercurial-scm.org/D7231

py3: avoid `b'%s' % type(...)` in a ProgrammingError Differential Revision: https://phab.mercurial-scm.org/D7363

largefiles: delete unused repo.status_nolfiles() Differential Revision: https://phab.mercurial-scm.org/D7142

largefiles: use context manager for wlock in repo.status() override Differential Revision: https://phab.mercurial-scm.org/D7141

largefiles: allow "lfstatus" context manager to set value to False Differential Revision: https://phab.mercurial-scm.org/D7140

largefiles: move lfstatus context manager to lfutil Differential Revision: https://phab.mercurial-scm.org/D7139

revlog: delete references to deleted nullid sentinel value We stopped keeping the nullid in the index in 781b2720d2ac (index: don't include nullid in len(), 2018-07-20). I forgot to update some comments about it. Differential Revision: https://phab.mercurial-scm.org/D7366

revlog: remove the, now unused, `revlog._nodepos` attribute Spotted by Yuya Nishihara.

py3: fix sorting of obsolete markers in obsutil (issue6217) This is similar to 01e8eefd9434 and others. We move the sortedmarkers() function from exchange module to obsutil.

nodemap: use bytes for the error message Spotted by Yuya Nishihara.

py3: use native strings when forming email headers in patchbomb Per previous changesets, encoded header's values are native str. We complete the change in patchbomb extension to have literal header values native str as well. Then we can also change headers' keys to be str. In _msgid(), we still need to use encoding.strfromlocal() because usage of os.environ is not allowed by check-code. This finally removes the "if pycompat.ispy3:" TODO.

mail: convert addr to str early in addrlistencode()

mail: let addressencode() / addrlistencode() return native strings Avoids conversion to "str" on py3.

mail: let headencode() return a native string This is to avoid conversion to/from str on py3.

revlog: deprecate direct `nodemap` access Now that all in-core user have been updated to the new API, we can deprecated direct access to nodemap. Differential Revision: https://phab.mercurial-scm.org/D7362

revlog: drop special hack in the `nodemap` property This code does not seems necessary anymore. Differential Revision: https://phab.mercurial-scm.org/D7361

perf: make `perfrevlogindex` use the new `index.rev` api if available Differential Revision: https://phab.mercurial-scm.org/D7360

perf: make `perfnodemap` use the new `index.get_rev` api if available Differential Revision: https://phab.mercurial-scm.org/D7358

index: use `index.get_rev` in `remotefilelog` Differential Revision: https://phab.mercurial-scm.org/D7357

index: use `index.get_rev` in `histedit.processreplacement` Differential Revision: https://phab.mercurial-scm.org/D7354

index: use `index.get_rev` in `histedit.adjustreplacementsfrommarkers` Differential Revision: https://phab.mercurial-scm.org/D7353

index: use `index.get_rev` in `rebase._computeobsoletenotrebased` Differential Revision: https://phab.mercurial-scm.org/D7350

index: use `index.get_rev` in `rebase.successorrevs` The code is slightly updated in the process for simplicity. Differential Revision: https://phab.mercurial-scm.org/D7349

index: use `index.get_rev` in `unionrepo.unionrevlog` Differential Revision: https://phab.mercurial-scm.org/D7338

index: use `index.rev` in `bundlerepo.bundlerevlog` Differential Revision: https://phab.mercurial-scm.org/D7340

index: use `index.get_rev` in `exchange._pullapplyphases` Differential Revision: https://phab.mercurial-scm.org/D7347

index: use `index.rev` in `exchange.fallbackheads` Differential Revision: https://phab.mercurial-scm.org/D7345

index: use `index.get_rev` in `revset._mapbynodefunc` We slightly update the code in the process. Differential Revision: https://phab.mercurial-scm.org/D7343

index: use `index.get_rev` in `discovery._postprocessobsolete` Differential Revision: https://phab.mercurial-scm.org/D7341

index: use `index.get_rev` in `repoview.pinnedrevs` We simplify the code in the process Differential Revision: https://phab.mercurial-scm.org/D7337

index: use `index.get_rev` in `obsolete._computephasedivergentset` Differential Revision: https://phab.mercurial-scm.org/D7336

index: use `index.get_rev` in `obsutil.getobsoleted` Differential Revision: https://phab.mercurial-scm.org/D7335

index: use `index.get_rev` in `phases.newheads` Differential Revision: https://phab.mercurial-scm.org/D7332

index: use `index.get_rev` in `localrepo.known` Differential Revision: https://phab.mercurial-scm.org/D7327

index: add a `get_rev` method (API) The new `index.getrev(node)` is to be preferred over using: `index.nodemap.get(node)`. This get us closer to be able to remove the `nodemap` attribute of the index. Differential Revision: https://phab.mercurial-scm.org/D7326

index: use `index.rev` in `revlog.rev` Differential Revision: https://phab.mercurial-scm.org/D7325

index: add a `rev` method (API) The new `index.rev(node)` is to be preferred over using `node index.nodemap[node]`. This get us closer to be able to remove the `nodemap` attribute of the index. Differential Revision: https://phab.mercurial-scm.org/D7324

perf: make `perfphasesremote` use the new `index.has_node` api (If available) Differential Revision: https://phab.mercurial-scm.org/D7359

index: use `index.has_node` in `infinitypush` Differential Revision: https://phab.mercurial-scm.org/D7356

index: use `index.has_node` in `phabricator.getoldnodedrevmap` Differential Revision: https://phab.mercurial-scm.org/D7355

index: use `index.has_node` in `histedit.cleanupnode` Differential Revision: https://phab.mercurial-scm.org/D7352

index: use `index.has_node` in `histedit._finishhistedit Differential Revision: https://phab.mercurial-scm.org/D7351

index: use `index.has_node` in `exchangev2._pullchangesetdiscovery` Differential Revision: https://phab.mercurial-scm.org/D7348

index: use `index.has_node` in `exchange._pulldiscoverychangegroup` Differential Revision: https://phab.mercurial-scm.org/D7346

index: use `index.has_node` in `bookmarks.bmstore` Differential Revision: https://phab.mercurial-scm.org/D7344

index: use `index.has_node` in `tags.findglobaltags` Differential Revision: https://phab.mercurial-scm.org/D7342

index: use `index.has_node` in `bundlerepo.bundlerevlog` Differential Revision: https://phab.mercurial-scm.org/D7339

index: use `index.has_node` in `obsutil.foreground` Differential Revision: https://phab.mercurial-scm.org/D7334

index: use `index.has_node` in `obsutil.exclusivemarkers` Differential Revision: https://phab.mercurial-scm.org/D7333

index: use `index.has_node` in `phases.filterunknown` Differential Revision: https://phab.mercurial-scm.org/D7331

index: use `index.has_node` in `analyzeremotephases` Differential Revision: https://phab.mercurial-scm.org/D7330

index: use `index.has_node` in `repo._rollback` Differential Revision: https://phab.mercurial-scm.org/D7329

index: use `index.has_node` in `revlog.addgroup` Differential Revision: https://phab.mercurial-scm.org/D7328

index: use `index.has_node` in `revlog.addrevision` Differential Revision: https://phab.mercurial-scm.org/D7323

index: add a `has_node` method (API) The new `index.has_node(node)` is to be preferred over: `node in index.nodemap`. This get us closer to be able to remove the `nodemap` attribute of the index. Differential Revision: https://phab.mercurial-scm.org/D7322

revlog: deal with nodemap deletion within the index Since the nodemap data now live in the index, it should be the index responsibility to ensure the data are up to date. The C version of the index is already dealing with such deletion. This work is part of a refactoring to unify the revlog index and the nodemap. This unification prepare the use of a persistent nodemap. Differential Revision: https://phab.mercurial-scm.org/D7321

revlog: clean up the node of all revision stripped in the C code For some obscure reason, the loop cleaning up node was skipping the first element… I cannot see a reason for it. The overall code is running fine nevertheless because the node are also explicitly deleted from python. We want to delete this explicit deletion, so we need to fix that code first. This work is part of a refactoring to unify the revlog index and the nodemap. This unification prepare the use of a persistent nodemap. Differential Revision: https://phab.mercurial-scm.org/D7320

revlog: move nodemap update within the index code Since the nodemap data now live in the index, it should be the index responsibility to ensure the data are up to date. This work is part of a refactoring to unify the revlog index and the nodemap. This unification prepare the use of a persistent nodemap. Differential Revision: https://phab.mercurial-scm.org/D7319

revlog: deprecate the _nodecache attribute (API) This attribute have been the same object for a long time. Code should directly access the nodemap in case of need. (Or the final API when this will be done). This work is part of a refactoring to unify the revlog index and the nodemap. This unification prepare the use of a persistent nodemap. Differential Revision: https://phab.mercurial-scm.org/D7318

revlog: access the nodemap through the index The `revlog.rev` method now access the nodemap through `self.index.nodemap`, instead of using the `_nodecache` attribute that we are trying to remove. This work is part of a refactoring to unify the revlog index and the nodemap. This unification prepare the use of a persistent nodemap. Differential Revision: https://phab.mercurial-scm.org/D7317

revlog: return the nodemap as the nodecache Ultimately, the nodecache is the nodemap. To stop having this disctinction, we start using the nodemap where the nodecache was used. This work is part of a refactoring to unify the revlog index and the nodemap. This unification prepare the use of a persistent nodemap. Differential Revision: https://phab.mercurial-scm.org/D7316

revlog: clarify which version use the older API in perf This work is part of a refactoring to unify the revlog index and the nodemap. This unification prepare the use of a persistent nodemap. Differential Revision: https://phab.mercurial-scm.org/D7315

revlog: no longer return the nodemap after parsing Now that both pure and cext version have a `nodemap` attribute, we do not need to return the `nodemap` object after parsing. This work is part of a refactoring to unify the revlog index and the nodemap. This unification prepare the use of a persistent nodemap. Differential Revision: https://phab.mercurial-scm.org/D7314

revlog: move the nodemap into the index object (for pure) This make the pure code closer to the C extension one. The ultimate goal is to merge the two into a single object and offer a unified API. This changeset focus on gathering the data on the same object. For now the code for `revlogoldindex` and `BaseIndexObject` index object are quite similar. However, there will be larger divergence later on, so I don't think is worth doing a base case. This work is part of a refactoring to unify the revlog index and the nodemap. This unification prepare the use of a persistent nodemap. Differential Revision: https://phab.mercurial-scm.org/D7313

revlog: introduce an explicit NodeMap class for pure code This class make the "pure" nodemap raise the same exception than the C-extension one. This is a step toward unifying nodemap and index, the class is not meant to survive on the long run. This work is part of a refactoring to unify the revlog index and the nodemap. This unification prepare the use of a persistent nodemap. There is a new `isinstance` call, it will be cleaned up in coming changesets. Differential Revision: https://phab.mercurial-scm.org/D7312

utils: move the `dirs` definition in pathutil (API) Before this change, the `dirs` class was accessible through the `mercurial.util` module. That module is expected to stay free of scm specific content. The `pathutil` destination has been selection by Martin von Zweigbergk. This work is part of a refactoring to unify the revlog index and the nodemap. This unification prepare the use of a persistent nodemap. Differential Revision: https://phab.mercurial-scm.org/D7311

packaging: stop installing i18n files The WiX installer has been shipping the content of the i18n/ directory since it was introduced in 2010 in 1e022c88a0a5. And the installer was subsequently refactored to only ship the .po files and hggettext. The .po files and the hggettext script are only used at build time to produce .mo files, which Mercurial does use at run-time. It doesn't make sense to install these files on Windows. So this commit stops doing that. This change further converges the file layout of the Inno and WiX installers. Differential Revision: https://phab.mercurial-scm.org/D7165

packaging: install hgk as hgk.tcl Giving it a file extension will give Windows a better chance at being able to execute it via standard means. This also makes WiX consistent with the Inno installer. Differential Revision: https://phab.mercurial-scm.org/D7164

packaging: add logo-droplets.svg The WiX installer adds this file. Let's add it to the Inno installer for consistency. Differential Revision: https://phab.mercurial-scm.org/D7163

packaging: install documentation to doc/ directory The source directory is "doc" and "doc" is used by the WiX installer. I think it makes sense to converge on the canonical path name. While we're here, we also lower the case because we can. (I think lower case names make more sense plus it is more consistent.) This only affects the Inno installer at the moment because it is the only installer using the staging code. Differential Revision: https://phab.mercurial-scm.org/D7162

packaging: always pass VERSION into Inno invocation The code in the Inno file was a holdover from before we had Python driving execution. With Python in the driver's seat, we can now have it resolve the version string and pass it into Inno, making the code easier to understand for people who aren't packaging gurus. Differential Revision: https://phab.mercurial-scm.org/D7161

packaging: remove hg.exe.local file <exe>.local files are used by Windows to set up DLL redirection. But these files are ignored if you embed an application manifest in your binary, which we do. So the existence of this file serves no purpose. So we remove it. Differential Revision: https://phab.mercurial-scm.org/D7160

packaging: stage installed files for Inno Previously, the Inno installer maintained its own mapping of source files to install location. (We have to maintain a similar mapping in the WiX installer.) Managing the explicit file layout for Windows packages is cumbersome and redundant. Every time you want to change the layout you need to change N locations. We frequently forget to do this and we only find out when people install Mercurial from our packages at release time. This commit starts the process of consolidating and simplifying the logic for managing the install layout on Windows. We introduce a list of install layout rules. These are simply source filenames (which can contain wildcards) and destination paths. The Inno packaging code has been updated to assemble all files into a staging directory that mirrors the final install layout. The list of files to add to the installer is derived by walking this staging directory and dynamically emitting the proper entries for the Inno Setup script. I diffed the file layout before and after this commit and there is no difference. Another benefit of this change is that it facilitates easier testing of the Windows install layout. Before, in order to test the final install layout, you needed to build an installer and run it. Now, you can stage files into the final layout and test from there, without running the installer. This should cut down on overhead when changing Windows code. Differential Revision: https://phab.mercurial-scm.org/D7159

packaging: process Inno Setup files with Jinja2 I want to make the Inno Setup files dynamically generated. This will enable us to do things like automatically derive the set of files to be packaged instead of having to manually keep lists of files in sync across installers. As the first step towards this, we process the Inno Setup files with Jinja2. As part of this conversion, we had to escape syntax in mercurial.iss that conflicts with Jinja2. I also took the opportunity to include modpath.iss via Jinja2 instead of using Inno's preprocessor. This keeps the Python code a bit simpler. Differential Revision: https://phab.mercurial-scm.org/D7158

packaging: install and run Inno files in a build directory Upcoming commits will refactor the Windows installers so their content is dynamically derived. In preparation for this, we copy the Inno files into a new build directory and run them from there. This required changing some relative paths to work from a new directory level but is otherwise a pretty straightforward change. Differential Revision: https://phab.mercurial-scm.org/D7157

packaging: consolidate CLI functionality into packaging.py Consolidating functionality for invoking code in the hgpackaging package through a single CLI entry point will make things simpler when we add more complexity to that package. For example, it will allow us to run things out of a virtualenv with third party packages. This commit consolidates functionality from the Inno and WiX build.py scripts into a new packaging.py script. That script simply creates a virtualenv and runs the CLI functionality in it. The new virtualenv is populated with jinja2 because I felt it easier to incorporate requirements file processing in this commit and we will soon use jinja2 in an upcoming commit. The unified CLI functionality will also make it easier to script other packaging workflows going forward. e.g. RPM, Debian, and macOS packaging. Differential Revision: https://phab.mercurial-scm.org/D7156

packaging: clean up excess packages for Inno This partially reverts e97ffa5511ba. A bunch of package dependencies were getting picked up because I ran pip-compile on Linux and the keyring package has some Linux-only dependencies. I have run pip-compile again - this time from Windows - to undo the addition of these dependencies. I also normalized whitespace to LF. Differential Revision: https://phab.mercurial-scm.org/D7155

unshelve: correct help string; unshelve does not accept file list The non-option arguments to `hg unshelve` are interpreted as the name of the shelve to unshelve, not the list of files to unshelve. While that functionality would probably be nice to have, that's well beyond the scope of this documentation fix. Differential Revision: https://phab.mercurial-scm.org/D7365

repoview: use class literal for creating filteredchangelog The type name is constant, so we don't need to create it dynamically using type(). As suggested by Yuya. Differential Revision: https://phab.mercurial-scm.org/D7364

vagrant: update vagrant image to buster from jessie I honestly have no idea if this works since I don't use vagrant, but looking around online suggests it should work fine. Differential Revision: https://phab.mercurial-scm.org/D7309

packaging: remove references to debian jessie Jessie is on life support at this point, so let's update the list of releases in the Makefiles. Differential Revision: https://phab.mercurial-scm.org/D7308

sqlitestore: remove superfluous r-prefixes on strings I did this file separately because it's a _ton_ of hits and it had some manual reformatting fallout as a result of black squeezing some expressions onto a single line. Differential Revision: https://phab.mercurial-scm.org/D7307

cleanup: remove pointless r-prefixes on single-quoted strings This is the promised second step on single-quoted strings. These had existed because our source transformer didn't turn r'' into b'', so we had tagged some strings as r-strings to get "native" strings on both Pythons. Now that the transformer is gone, we can dispense with this nonsense. Methodology: I ran hg locate 'set:added() or modified() or clean()' | egrep '.*\.py$' | xargs egrep --color=never -n -- \[\^b\]\[\^a-z\]r\'\[\^\'\\\\\]\*\'\[\^\'\ in an emacs grep-mode buffer, and then used a keyboard macro to iterate over the results and remove the r prefix as needed. # skip-blame removing unneeded r prefixes left over from Python 3 migration. Differential Revision: https://phab.mercurial-scm.org/D7306

rust-threads: force Rayon to respect the worker count in config As per the inline comment, this is a workaround because Rust code does not yet know how to read config files. Differential Revision: https://phab.mercurial-scm.org/D7310

phabricator: use context manager form of progress in uploadchunks Follow-up to 453079605242 / D7046. Differential Revision: https://phab.mercurial-scm.org/D7134

cleanup: remove pointless r-prefixes on double-quoted strings This is only double-quoted strings. I'll do single-quoted strings as a second step. These had existed because our source transformer didn't turn r"" into b"", so we had tagged some strings as r-strings to get "native" strings on both Pythons. Now that the transformer is gone, we can dispense with this nonsense. Methodology: I ran hg locate 'set:added() or modified() or clean()' | egrep '.*\.py$' | xargs egrep --color=never -n -- \[\^a-z\]r\"\[\^\"\\\\\]\*\"\[\^\"\] in an emacs grep-mode buffer, and then used a keyboard macro to iterate over the results and remove the r prefix as needed. # skip-blame removing unneeded r prefixes left over from Python 3 migration. Differential Revision: https://phab.mercurial-scm.org/D7305

formatting: remove another data-ogre from the config example The fix extension will pass the contents on stdin and read the output from stdout. By invoking `rustfmt` as `rustfmt {rootpath}`, we tell rustfmt to format the given path and update it on disk. That also makes rustftm not produce the formatted output to stdout. So the fix extension ends up writing out empty files. Differential Revision: https://phab.mercurial-scm.org/D7304

fix: replace str() by b'%d' for formatting integer Differential Revision: https://phab.mercurial-scm.org/D7303

zeroconf: improve the extension's documentation Differential Revision: https://phab.mercurial-scm.org/D7302

branchmap: pytype is confused about bytestr Differential Revision: https://phab.mercurial-scm.org/D7290

mail: suppress a pytype error that's just experimentally wrong It's probably more correct to pass a str here, but I'm not going to worry about it for now. Differential Revision: https://phab.mercurial-scm.org/D7283

dagparser: suppress some pytype errors around pycompat.bytestring I can't justify why we're getting these errors, but nothing I do fixes these handful of calls, so let's just move on with suppressions. Differential Revision: https://phab.mercurial-scm.org/D7277

encoding: add comment-based type hints for pytype Differential Revision: https://phab.mercurial-scm.org/D7275

merge with stable

dirs: reject consecutive slashes in paths We shouldn't ever see those, and the fuzzer go really excited that if it gives us a 65k string with 55k slashes in it we use a lot of RAM. This is a better fix than what I tried in D7105. It was suggested by Yuya, and I verified it does in fact cause the fuzzer to not OOM. This is a revision of D7234, but with the missing set of an error added. I added a unit test of the dirs behavior because I needed to reason more carefully about the failure modes around consecutive slashes. Differential Revision: https://phab.mercurial-scm.org/D7252

tests: quote $PYTHON in test-byteify-strings.t This is needed because the default install location is "Program Files" on Windows. Differential Revision: https://phab.mercurial-scm.org/D7251

repoview: define filteredchangelog as a top-level (non-local) class As suggested by Greg. This makes it easier for extensions to override the filtering. Differential Revision: https://phab.mercurial-scm.org/D7256

cmdutil: suppress bogus pytype errors pytype seems to handle our subclass of tuple (scmutil.status) especially poorly. We should probably file a bug. Differential Revision: https://phab.mercurial-scm.org/D7294

changegroup: suppress pytype error that's wrong Differential Revision: https://phab.mercurial-scm.org/D7293

changegroup: avoid shadowing a set with an int This confuses the heck out of pytype, and I basically agree with it that this shadowing is confusing. There's a chance this causes some memory to be freed later, but I think it's probably not worth worrying about for now. Differential Revision: https://phab.mercurial-scm.org/D7292

bundle2: more pytype suppressions around bytestring We clearly need to work around this somehow. Differential Revision: https://phab.mercurial-scm.org/D7291

archival: suppress some incorrect pytype failures Again, I suspect some missing coverage in type stubs for the stdlib. I'll revisit these later. Differential Revision: https://phab.mercurial-scm.org/D7288

simplemerge: disable a pytype error where it's just confused Differential Revision: https://phab.mercurial-scm.org/D7287

match: suppress error about subscripting an exception I get the sense I should file a bug about subscripting Exception subclasses. Differential Revision: https://phab.mercurial-scm.org/D7286

vfs: more attribute suppressions Differential Revision: https://phab.mercurial-scm.org/D7285

mdiff: mark diffopts as having dynamic attributes This class looks like we could move it to being an attrs and life would be better, but let's do that later. Differential Revision: https://phab.mercurial-scm.org/D7284

vfs: suppress some pytype errors around us using a private attribute Looking at threading._MainThread seems like we're probably a little unsupported, but since this code appears to work on both Python 2 and 3 I'm not going to sweat this for now. Differential Revision: https://phab.mercurial-scm.org/D7282

wireprototypes: disable pytype where it's just confused By inspection validnames is always a Set[bytes] here, but for some reason pytype is convinced it's a bytes. Let's disable the error for now. Differential Revision: https://phab.mercurial-scm.org/D7281

scmposix: another suppression on IOError subscripting As before, this will clear up when we move to Python 3-only. Differential Revision: https://phab.mercurial-scm.org/D7280

pvec: add an explicit type hint to help pytype Differential Revision: https://phab.mercurial-scm.org/D7279

procutil: suppress pytype warnings around windows-only attributes Differential Revision: https://phab.mercurial-scm.org/D7278

posix: add a pytype suppression This one seems a little sketchier than the others, but this will clean up when we're Python 3-only. Differential Revision: https://phab.mercurial-scm.org/D7276

color: suppress pytype warning on a windows-only module Differential Revision: https://phab.mercurial-scm.org/D7274

compression: tell pytype to not sweat a missing `zstd` module Differential Revision: https://phab.mercurial-scm.org/D7273

pytype: don't warn us about ignored-on-py3 metaclasses We can remove this when we're Python 3-only, but for now it's just too awkward to deal with and it's harmless. Differential Revision: https://phab.mercurial-scm.org/D7272

linelog: be more careful about types before looking for _target attribute Without this, pytype (correctly) complains that fakejge might not have a _cmprev attribute. We're operating on some constraints not expressible in the type system, so we detect the invalid case and raise a nicer exception now. Sadly, we also need an `assert` to give PyType a clue that we know what we're doing, but I can absolutely live with that. Differential Revision: https://phab.mercurial-scm.org/D7271

linelog: suppress annoying pytype warning about an ignored metaclass This isn't actively hurting us right now, and it's not trivial to set up this metaclass correctly on both Python 2 and Python 3 without dragging in all of `six`, but we still get lots of typechecking goodness with this error suppressed. Differential Revision: https://phab.mercurial-scm.org/D7270

hgignore: ignore generated pytype stubs Differential Revision: https://phab.mercurial-scm.org/D7269

hghave: add a check for pytype, Google's Python type checker I'm not unbiased, but I strongly prefer pytype to mypy: it does more inference, so we don't need as many annontations. In theory we can use both, but since I'm doing some tinkering with pytype I figured I'd just try setting it up. Differential Revision: https://phab.mercurial-scm.org/D7268

encoding: define local identify functions with explicit type comments This removes some obfuscation as far as pytype is concerned and corrects many bogus type errors. Differential Revision: https://phab.mercurial-scm.org/D7267

vfs: add NotImplementedError version of join Again, this helps out pytype. Differential Revision: https://phab.mercurial-scm.org/D7266

vfs: add a NotImplementedError implementation of __call__ This helps pytype considerably on this file. Differential Revision: https://phab.mercurial-scm.org/D7265

vfs: fix erroneous bytes constants Detected with pytype. Differential Revision: https://phab.mercurial-scm.org/D7264

pvec: migrate to modern integer division Detected with pytype. Differential Revision: https://phab.mercurial-scm.org/D7263

templateutil: fix a missing ABCMeta assignment Caught by pytype. Differential Revision: https://phab.mercurial-scm.org/D7262

pvec: fix overlooked chr() call Caught by pytype. Differential Revision: https://phab.mercurial-scm.org/D7261

vfs: another bytes-str confusion on thread name Differential Revision: https://phab.mercurial-scm.org/D7260

statprof: correctly always pass a str as the thread name Caught by pytype. Differential Revision: https://phab.mercurial-scm.org/D7259

py3: fix handling of ctrl keys in crecord (issue6213) The "keypressed" value in handlekeypressed() is a key name obtained by curses's getkey(); this can be a multibyte string for special keys like CTRL keys. Calling curses.unctrl() with such a value fails on Python 3 with a TypeError as described in issue6213. (On Python 2, this does not crash, but I'm not sure the result is correct, though it does no matter here.) So instead of calling unctrl(), we compare "keypressed" with the expected "^L" obtained by curses.ascii.ctrl("L").

py3: keep "keypressed" a native str in crecord This will help in the next changeset by avoiding a decode step. Also, the actual bytes conversion seems superfluous since values coming from curses's getkey() will be a native string. As a consequence, we open the "testcommands" file (used in test-interactive-curses.t) in text mode.

py3: compare response of crecord's confirmationwindow with str confirmationwindow() returns a native string, as a result of calling chr() on getch(). On Python 3, response.lower().startswith(b"y") leads to a TypeError. This fixes a crash when typing "r" in the curses interface of interactive commit.

py3: compare http server's command with a native string The "command" attribute is an str, so comparing with a bytes would not work on Python 3. This might solve issues in test-lfs-serve-access.t that happens sometimes (especially in CI): --- /hgwork/src/tests/test-lfs-serve-access.t +++ /hgwork/src/tests/test-lfs-serve-access.t.err @@ -163,11 +163,13 @@ $ cat $TESTTMP/access.log $TESTTMP/errors.log $LOCALIP - - [$LOGDATE$] "POST /missing/objects/batch HTTP/1.1" 404 - (glob) + $LOCALIP - - [05/Nov/2019 16:32:34] "{"objects": [{"oid": "f03217a32529a28a42d03b1244fe09b6e0f9fd06d7b966d4d50567be2abe6c0e", "size": 20}], "operation": "download"}" HTTPStatus.BAD_REQUEST - $LOCALIP - - [$LOGDATE$] "GET /subdir/mount/point?cmd=capabilities HTTP/1.1" 200 - (glob) $LOCALIP - - [$LOGDATE$] "GET /subdir/mount/point?cmd=batch HTTP/1.1" 200 - x-hgarg-1:cmds=heads+%3Bknown+nodes%3D x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob) $LOCALIP - - [$LOGDATE$] "GET /subdir/mount/point?cmd=getbundle HTTP/1.1" 200 - x-hgarg-1:bookmarks=1&bundlecaps=HG20%2Cbundle2%3DHG20%250Abookmarks%250Achangegroup%253D01%252C02%252C03%250Adigests%253Dmd5%252Csha1%252Csha512%250Aerror%253Dabort%252Cunsupportedcontent%252Cpushraced%252Cpushkey%250Ahgtagsfnodes%250Alistkeys%250Aphases%253Dheads%250Apushkey%250Aremote-changegroup%253Dhttp%252Chttps%250Arev-branch-cache%250Astream%253Dv2&cg=1&common=0000000000000000000000000000000000000000&heads=525251863cad618e55d483555f3d00a2ca99597e&listkeys=bookmarks&phases=1 x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob) $LOCALIP - - [$LOGDATE$] "POST /subdir/mount/point/.git/info/lfs/objects/batch HTTP/1.1" 200 - (glob) $LOCALIP - - [$LOGDATE$] "GET /subdir/mount/point/.hg/lfs/objects/f03217a32529a28a42d03b1244fe09b6e0f9fd06d7b966d4d50567be2abe6c0e HTTP/1.1" 200 - (glob) + $LOCALIP - - [05/Nov/2019 16:32:34] code 400, message Bad request version ('"download"}') Blobs that already exist in the usercache are linked into the repo store, even though the client doesn't send the blob. @@ -195,6 +197,7 @@ server2/.hg/store/lfs/objects/f0/3217a32529a28a42d03b1244fe09b6e0f9fd06d7b966d4d50567be2abe6c0e $ "$PYTHON" $RUNTESTDIR/killdaemons.py $DAEMON_PIDS $ cat $TESTTMP/errors.log + $LOCALIP - - [05/Nov/2019 16:32:34] code 400, message Bad request version ('"download"}') $ cat >> $TESTTMP/lfsstoreerror.py <<EOF > import errno (from https://ci.hg.gregoryszorc.com/job-info/hg-committed-ca3dca416f8d5863ca6f5a4a6a6bb835dcd5feeb-debian10-cpython-3.7-0)

rust-status: remove dead code The `walk_explicit` function is only called when using a prefix matcher, which the Rust code does not yet support. This function will return in a future patch, probably with a different signature for performance reasons. With it, the `files` argument and its interface code can be removed for now. Differential Revision: https://phab.mercurial-scm.org/D7253

hghave: disallow symlinks on Windows Symlinks on Windows require either a special priviledge, or enabling Developer Mode. It's probably the latter that is enabled on the new CI machine. But since Mercurial itself is saying no to symlinks on Windows, the tests for symlinks shouldn't be attempted. This should fix a lot of the noise in the py3 tests. Differential Revision: https://phab.mercurial-scm.org/D7233

repoview: avoid wrapping changelog if there's nothing to filter This simplifies the code a bit by moving the optimizaton for no filtered revisions to one place. I assume it also makes working with repos without obsmarkers a little faster, but it doesn't seem significant. Differential Revision: https://phab.mercurial-scm.org/D7248

repoview: move changelog.flags() override to filteredchangelog Differential Revision: https://phab.mercurial-scm.org/D7247

repoview: move changelog.parentrevs() override to filteredchangelog Differential Revision: https://phab.mercurial-scm.org/D7246

repoview: move changelog.linkrev() override to filteredchangelog Differential Revision: https://phab.mercurial-scm.org/D7245

repoview: move changelog.node() override to filteredchangelog Differential Revision: https://phab.mercurial-scm.org/D7244

repoview: move changelog.rev() override to filteredchangelog Differential Revision: https://phab.mercurial-scm.org/D7243

repoview: move changelog.strip() override to filteredchangelog Differential Revision: https://phab.mercurial-scm.org/D7242

repoview: move changelog.headrevs() override to filteredchangelog Differential Revision: https://phab.mercurial-scm.org/D7241

repoview: move changelog.revs() override to filteredchangelog Differential Revision: https://phab.mercurial-scm.org/D7240

repoview: move changelog.__iter__() override to filteredchangelog Differential Revision: https://phab.mercurial-scm.org/D7239

repoview: move changelog.__contains__() override to filteredchangelog Differential Revision: https://phab.mercurial-scm.org/D7238

repoview: move changelog.tiprev() override to filteredchangelog Differential Revision: https://phab.mercurial-scm.org/D7237

repoview: wrap changelog class when filtering The class doesn't yet do anything. I'll move the filter-aware overrides from the changelog class over one by one to this class in coming patches. That will leave the changelog class simpler and will centralize more of the filtering logic to repoview. I could not measure any performance difference. Differential Revision: https://phab.mercurial-scm.org/D7236

repoview: extract a function for wrapping changelog I would like to clean up the changelog class by moving out knowledge of filtering. The filtering will instead be done in a class that wraps the changelog. This patch prepares for that. Differential Revision: https://phab.mercurial-scm.org/D7235

revlog: move tiprev() from changelog up to revlog This makes the changelog override simpler and it seems more consistent. Differential Revision: https://phab.mercurial-scm.org/D7250

tests: fix a few `(no-windows !)` conditionals that are really `(symlink !)` The CI for py3 is assuming that symlinks are possible (they are when running as Administrator or when Developer Mode is enabled), and these popped up. Differential Revision: https://phab.mercurial-scm.org/D7232

rust-matchers: add `Matcher` trait and implement `AlwaysMatcher` In our quest of a faster Mercurial, we have arrived at the point where we need to implement the matchers in Rust. This RFC mainly for the `Matcher` trait to see if the changes proposed feel fine to people with more experience on the matter. While the `AlwaysMatcher` implementation is here as a trivial example, it should be the first step towards matchers use in Rust as it is currently the only supported one. Notable changes: - `exact` is renamed to `exact_match` - enums for `visit*` methods with `Recursive` instead of `'all'`, etc. - a new `roots`, separate from `file_set` - no `bad`, `explicitdir` or `traversedir` functions as they can be passed to the high functions instead of the matchers Thanks to Martin for suggesting the last two (most important) changes and for reaching out to help a few weeks ago. Differential Revision: https://phab.mercurial-scm.org/D7178

merge with stable

perf: add a way to benchmark `dirstate.status` Getting more details about time spend in this specific internal bit is meaningful.

largefiles: delete obsolete and unused repo.push() The function was removed from localrepo in 4d52e6eb98ea (locarepo: remove the `push` method (API), 2014-09-25). Differential Revision: https://phab.mercurial-scm.org/D7174

censor: document that some commands simply ignore censored data I can't see a benefit for hg grep to ever error out early when it encounters censored data.

grep: warn on censored revisions instead of erroring out We need most of the grep logic to go through in case we encounter a censored revision, so we just return a None body for a censored node, and we stop just short of trying to record matches with the contents of that censored body. The other parts such as recording that the censored file has been considered at this revision needs to go into the proper dicts. I have also gotten weary of all the abbreviations, so while I did a small refactor to move the file-data-getting operation into a common function, I also expanded the abbreviations of the relevant variables within this little function. Hopefully some day this helps someone figure out what all the abbreviations mean. Although the censoring docs currently state that some commands error out or are ignored depending on the `censor.policy` config, I cannot see a benefit for grep to ever stop dead in its tracks when a censored revision is encountered. I will also amend the docs to indicate that some commands, such as grep, unconditionally ignore censored revisions.

py3: enable legacy fs encoding to fix filename compatibility on Windows This patch is untested. I just followed the instruction: https://docs.python.org/3/whatsnew/3.6.html#pep-529-change-windows-filesystem-encoding-to-utf-8

rust-cpython: run cargo fmt

rust-cpython: remove useless PyResult<> from leak_immutable() The caller should know if the shared data is mutably borrowed or not.

rust-cpython: remove useless PyRefMut wrapper

rust-cpython: drop manual management of mutably_borrowed RefCell::borrow() should guarantee there's no mutable borrow. As a follow up, maybe PySharedState can be a pure data structure + function. Most ref-sharing business has already been moved to PySharedRef* and PyLeaked*.

rust-cpython: leverage RefCell::borrow() to guarantee there's no mutable ref Since the underlying value can't be mutably borrowed by PyLeaked, we don't have to manage yet another mutably-borrowed state. We can just rely on the RefCell implementation. Maybe we can add try_leak_immutable(), but this patch doesn't in order to keep the patch series not too long.

rust-cpython: remove useless Option<$leaked> from py_shared_iterator We no longer need to carefully drop the iterator when it's consumed. Mutation is allowed even if the iterator exists. There's a minor behavior change: next(iter) may return/raise something other than StopIteration if it's called after the iterator has been fully consumed, and if the Rust object isn't a FusedIterator.

rust-cpython: allow mutation unless leaked reference is borrowed In other words, mutation is allowed while a Python iterator holding PyLeaked exists. The iterator will be invalidated instead. We still need a borrow_count to prevent mutation while leaked data is dereferenced in Rust world, but most leak_count business is superseded by the generation counter. decrease_leak_count(py, true) will be removed soon.

rust-cpython: add generation counter to leaked reference This counter increments on borrow_mut() to invalidate existing leaked references. This is modeled after the iterator invalidation in Python. The other checks will be adjusted by the subsequent patches.

rust-cpython: add stub wrapper that'll prevent leaked data from being mutated In order to allow mutation of PySharedRefCell value while PyLeaked reference exists, we need yet another "borrow" scope where mutation is prohibited. try_borrow<'a> and try_borrow_mut<'a> defines the "borrow" scope <'a>. The subsequent patches will implement leak counter based on this scope. PyLeakedRef<T> and PyLeakedRefMut<T> could be unified to PyLeakedRef<&T> and PyLeakedRef<&mut T> respectively, but I didn't do that since it seemed a bit weird that deref_mut() would return a mutable reference to an immutable reference.

rust-cpython: rename PyLeakedRef to PyLeaked This series will make PyLeaked* behave more like a Python iterator, which means mutation of the owner object will be allowed and the leaked reference (i.e. the iterator) will be invalidated instead. I'll add PyLeakedRef/PyLeakedRefMut structs which will represent a "borrowed" state, and prevent the underlying value from being mutably borrowed while the leaked reference is in use: let shared = self.inner_shared(py); let leaked = shared.leak_immutable(); { let leaked_ref: PyLeakedRef<_> = leaked.borrow(py); shared.borrow_mut(); // panics since the underlying value is borrowed } shared.borrow_mut(); // allowed The relation between PyLeaked* structs is quite similar to RefCell/Ref/RefMut, but the implementation can't be reused because the borrowing state will have to be shared across objects having no lifetime relation. PyLeaked isn't named as PyLeakedCell since it isn't actually a cell in that leaked.borrow_mut() will require &mut self.

py3: don't use bytes with vars() or __dict__ Inspired by D7227. These were all the remaining instances I could find. Differential Revision: https://phab.mercurial-scm.org/D7230

Added signature for changeset ca3dca416f8d

Added tag 5.2 for changeset ca3dca416f8d

py3: add inline comment about encoding issue of str(Abort())

py3: do not reimplement Abort.__str__() on Python 2 It isn't necessary on Python 2, and the default implementation should be better than our BaseException_str() clone.

tests: write out file using bytes I/O The encoding of sys.stdout varies between Python versions. So using a one-liner to write a file from a Unicode string is not deterministic. This commit writes out the file using bytes I/O to ensure we have exactly the bytes we want in the file. This change fixes a test failure in Python 3.5/3.6. Differential Revision: https://phab.mercurial-scm.org/D7226