shelve: attempt to make merge labels more helpful Differential Revision: https://phab.mercurial-scm.org/D12024

histedit: attempt to make merge labels more helpful Differential Revision: https://phab.mercurial-scm.org/D12023

update: set custom conflict label for base commit Differential Revision: https://phab.mercurial-scm.org/D12022

merge: set custom conflict label for base commit Differential Revision: https://phab.mercurial-scm.org/D12021

rebase: set custom conflict label for base commit Differential Revision: https://phab.mercurial-scm.org/D12020

simplemerge: take over formatting of label from `filemerge` The padding we do of conflict labels depends on which conflict marker style is used. For two-way conflict markers (the default), the length of the base label shouldn't matter. It does before this patch, however. This patch moves the formatting from `filemerge` to `simplemerge`. The latter knows which conflict marker style to use, so it can easily decide about the padding. This change will allow us to use more descriptive "base" labels without causing illogical padding in 2-way markers. I'll do that next. One wrinkle is that we pass the same labels to external merge tools. I decided to change that in this patch to be simpler: no padding, and no ellipsis to fit within 80 columns. My reasoning is that the typical external, 3-or-4-panel merge tool doesn't show the labels on top of each others, so the padding doesn't make sense there. The ellipsis is probably not necessary because the external tools probably have their own way of dealing with long labels. Also, we limit them to "80 - 8" to fit the "<<<<<<< " before, which is almost definitely not what an external tool would put there. Differential Revision: https://phab.mercurial-scm.org/D12019

filemerge: pass `simplemerge.MergeInput` to tool functions Differential Revision: https://phab.mercurial-scm.org/D12018

filemerge: work with `simplemerge.MergeInput` in `filemerge()` We currently pass around pairs of file context objects and labels between functions in the `filemerge` module. I plan to pass around `simplemerge.MergeInput` instead. This patch prepares for that by using the type internally in `filemerge.filemerge()`. Differential Revision: https://phab.mercurial-scm.org/D12017

filemerge: always define a "base" label It simplifies the code and improves the conflict markers to always have the "base" label defined. Differential Revision: https://phab.mercurial-scm.org/D12016

filemerge: set default labels a little earlier By setting the default labels a little earlier, we can rely on them always being set, as far as I can tell. It may actually even be fine to rely on that even if we don't set them earlier, but it makes more sense to me to do it. Differential Revision: https://phab.mercurial-scm.org/D12015

filemerge: remove unused arguments from `_merge()` Differential Revision: https://phab.mercurial-scm.org/D12014

simplemerge: take arguments as annotated context objects The labels we put in conflict markers are formatted so the part before the ':' (typically says things like "local") is padded so the ':' is aligned among the labels. That means that if you specify a long label for "base" but the conflict marker style is "merge" (i.e. 2-way), the other two will have unwanted padding. We often don't specify a label for the base, so we don't notice the problem (and it may very well be that it didn't exist before my D11972). I think the best fix is to pass the labels along with the context objects, so the low-level code that switches on the marker style to use (i.e. `simplemerge`) can do the formatting. This patch starts doing that by passing a fully-formatted label to `simplemerge`. A coming patch will move the formatting to `simplemerge`. Differential Revision: https://phab.mercurial-scm.org/D12013

filemerge: make `labels` a required arg for `_premerge()` We always pass labels, and it's always a non-empty list, so we can simplify a bit. Differential Revision: https://phab.mercurial-scm.org/D12012

filemerge: remove unnecessary check for empty string Looking at the statement just before, the string is clearly never empty (or otherwise falsy). Differential Revision: https://phab.mercurial-scm.org/D12011

tests: also test spaces in conflict marker labels Our fake merge tool printed the arguments using `$1` etc. without quotes, which means that the shell collapsed repeated spaces within the argument. That resulted in tests that did not pick up changes to the number of spaces. Differential Revision: https://phab.mercurial-scm.org/D12010

blackbox: change year in logs to ISO 8601 format AFAIK, year/month/day is not a format than anyone uses. It seems more sensible to me to use ISO 8601, as that's standard and unambiguously year-month-day. Compatibility-wise, I think it's acceptable to change the default format. It's most for human consumption. It's plausible that a few tools parse this format, but it has already changed in the past (for instance to add the current revision). Differential Revision: https://phab.mercurial-scm.org/D12006

blackbox: add milliseconds to blackbox logs by default The current second granularity is often not specific enough to determine whether an hg command is happening before or after some other event. Given that starting a process takes on the order of 1ms (well, for native processes. It's quite a bit more for python processes), microseconds seems like unnecessary noise. This also lines up behavior with the rust version, where we already switched to millisecond precision. Differential Revision: https://phab.mercurial-scm.org/D12005

blackbox: correct date format doc Differential Revision: https://phab.mercurial-scm.org/D12004

branching: merge with stable

rust: Fix outdated comments in OwningDirstateMap OwningDirstateMap used to own a PyBytes, but was generalized to be more generic when it was moved from hg-cpython to hg-core. This fixes some comments that were still referencing PyBytes. Differential Revision: https://phab.mercurial-scm.org/D12003

merge: consider the file merged when using :merge-{local,other} Returning `None` from `simplemerge()` means that the caller interprets it as "no merge was necessary (because two sides matched)". See 6217040b2780 and issue2680 for some background. However, `simplemerge()` shouldn't even get called in such scenarios, and returning `None` means that the file is not considered merged, even though the contents actually were. See the affected test cases. Differential Revision: https://phab.mercurial-scm.org/D11999

simplemerge: delete unused exception class `CantReprocessAndShowBase` The only user was removed in f18830651811 (simplemerge: burn "minimal" feature to the ground, 2014-08-05). Differential Revision: https://phab.mercurial-scm.org/D11998

simplemerge: change _minimize() to minimize a single conflict `_minimize()` is weird in that it changes `a_lines` and `b_lines` but leaves `base_lines` unchanged, which means they'll be inconsistent. That was fine because the caller never used `base_lines`. With the recent refactorings of this code, we can now make this function cleaner by having it minimize a single conflict and not care about `base_lines`. This also makes the code simpler and makes the code for each `render_*()` function more similar. Differential Revision: https://phab.mercurial-scm.org/D11981

simplemerge: make minimize() a free function IMO, `Merge3Text` should be about merging text without knowing about conflict markers. "Conflict minimization" is about making conflict markers smaller, so it should be moved out. This patch does that. I'll refactor it next. Differential Revision: https://phab.mercurial-scm.org/D11980

simplemerge: simplify and rename `render_markers()` `render_markers()` now always renders minimized 2-way markers, so let's simplify and rename it accordingly. Differential Revision: https://phab.mercurial-scm.org/D11979

simplemerge: split out function for rendering :merge3 conflict markers The results in some duplicate, but the code is very straight-forward and I think it's worth it to have each conflict marker style in a separate function. Differential Revision: https://phab.mercurial-scm.org/D11978

simplemerge: move default labels to simplemerge extension I want to remove knowledge of paths from the `simplemerge` core module. As part of that, let's move the default labels from the core module to the extension. Differential Revision: https://phab.mercurial-scm.org/D11977

simplemerge: use same newline detection for :mergediff as for :merge[3] The `:merge` and `:merge3` markers used the newline style detected from the base input. The `:mergediff` style always used LF as newline character. This patch teaches `:mergediff` to use the same detection as the others did. Differential Revision: https://phab.mercurial-scm.org/D11976

simplemerge: clarify names of functions that render conflict markers Differential Revision: https://phab.mercurial-scm.org/D11975

simplemerge: make merge_lines() a free function IMO, the rendering of a conflict as text with merge markers should be separate from the code for resolving conflicts. The latter is what `Merge3Text` is mostly about already. Differential Revision: https://phab.mercurial-scm.org/D11974

simplemerge: convert `merge_lines()` away from generator We always consume all the lines and put them in a list anyway. By making the function not a generator, we can later make it return an additional value (to indicate if there were conflicts). Differential Revision: https://phab.mercurial-scm.org/D11973

simplemerge: use 3-way markers if mode=='merge3', ignoring number of labels Before this patch, we use a combination of `mode` and `labels` to decide which marker style to use. That's mostly my own fault because I was the one who added the `mergediff` as a "mode" instead of adding a separate argument for marker style. Still, let's make these arguments more orthogonal by ignoring the number of labels when deciding which marker style to use. This refactoring makes it a little clearer in the simplemerge *extension* that passing more than 2 `--label` arguments changes the marker style. Differential Revision: https://phab.mercurial-scm.org/D11972

rust: upgrade `rand*` crates `test-check-cargo-lock.t` is failing for me and I was hoping this would help. It doesn't, but we might as well take the upgrade now that I've done the (small amount of) work for it. Differential Revision: https://phab.mercurial-scm.org/D12000

logcmdutil: raise `StateError` when file to follow doesn't exist Differential Revision: https://phab.mercurial-scm.org/D11969

logcmdutil: raise `InputError` on bad CLI arguments Apparently there were no tests for any of these errors. Differential Revision: https://phab.mercurial-scm.org/D11968

tests: support pygments 2.11 (issue6628) Differential Revision: https://phab.mercurial-scm.org/D11970

exchange: add fast path for subrepo check on push Try to check if .hgsub and .hgsubstate exist at all before looking for them in every changeset to be pushed. The latter can be quite expensive for large repositories and the existance check is almost free. Differential Revision: https://phab.mercurial-scm.org/D11956

simplemerge: rewrite `merge_lines()` using `merge_groups()` Simple is better. Differential Revision: https://phab.mercurial-scm.org/D11967

simplemerge: make merge_groups() yield only 2-tuples `merge_groups()` currently yields 2-tuples or 4-tuples, making the callers check the first element to decide how to interpret the rest. Let's make it yield only 2-tuples, thereby simplifying life a little for the callers. Differential Revision: https://phab.mercurial-scm.org/D11966

rhg: desambiguate status without decompressing filelog if possible When status is unsure based on `stat()` and the dirstate if a file is clean or modified, we need to compare it against the filelog. This comparison can skip looking at contents if the lengths differ. This changeset optimize this further to deduce what we can about the length if the filelog without decompressing it or resolving deltas. Differential Revision: https://phab.mercurial-scm.org/D11965

filelog: add a comment explaining a fast path in filerevisioncopied() Differential Revision: https://phab.mercurial-scm.org/D11964

rhg: Store p1, p2, and hash in RevlogEntry This avoids a duplicate index lookup Differential Revision: https://phab.mercurial-scm.org/D11963

rhg: RevlogEntry::uncompressed_len is signed The corresponding Python code appears to explicitly check for non-negative values. Differential Revision: https://phab.mercurial-scm.org/D11962

rhg: Expose FilelogEntry that wraps RevlogEntry This can be later extended to access metadata such as `uncompressed_len` without necessarily resolving deltas. Differential Revision: https://phab.mercurial-scm.org/D11961

rhg: Add RevlogEntry::data that does delta resolution This requires keeping a `&Revlog` reference inside the `RevlogEntry` struct. This struct already had the appropriate lifetime parameter. Differential Revision: https://phab.mercurial-scm.org/D11960

rhg: Rename some revlog-related types and methods Use "data chunck" and "data" for a revlog entry’s data before and after resolving deltas (if any), repsectively. The former `FilelogEntry` actually only contains data, rename it to `FilelogRevisionData` accordingly. This leaves room to later have a `FilelogEntry` type that wraps `RevlogEntry`. Differential Revision: https://phab.mercurial-scm.org/D11959

git: add opener attribute to gitstore Differential Revision: https://phab.mercurial-scm.org/D11958

branching: merge stable into default

dirstate: remove deprecated API Differential Revision: https://phab.mercurial-scm.org/D11953

discovery: remove deprecated API Differential Revision: https://phab.mercurial-scm.org/D11952

revlog: remove deprecated APIs Differential Revision: https://phab.mercurial-scm.org/D11951

urls: remove deprecated APIs Differential Revision: https://phab.mercurial-scm.org/D11950

remotefilelog: remove deprecated API Differential Revision: https://phab.mercurial-scm.org/D11949

exchangev2: remove it As discussed on the mailing list, this is incomplete and unused with little hope of revival. Differential Revision: https://phab.mercurial-scm.org/D11954

rust: fix build errors on darwin I'm not all _that_ versed in Rust, but I think the root cause is that some constants are u16 rather than u32 on Darwin. I checked that the code still compiles on the latest Ubuntu. Differential Revision: https://phab.mercurial-scm.org/D11955

merge: with stable

help: properly document the default pattern The default isn't to glob. It's relpath, which does no globbing at all. Thus most commands and options like -X/-I and hg files require `glob:` in front of them to actually do any globbing. I think long ago the default was to glob, but that hasn't been true for a long time.

hg-core: dedup LazyAncestors Iterator impl … Which has been moved to the vcsgraph crate. AncestorsIterator cannot yet be removed because it is still being used in hg-core. Differential Revision: https://phab.mercurial-scm.org/D11948

hg-cpython: use ancestor iterator impls from vcsgraph Differential Revision: https://phab.mercurial-scm.org/D11947

hg-cpython: implement vcsgraph::Graph for our Index Differential Revision: https://phab.mercurial-scm.org/D11946

rust: add vcsgraph crate as dependency Preparing the integration of the vcsgraph library, which will provide generic algorithms for computations on version control graphs. Differential Revision: https://phab.mercurial-scm.org/D11945

simplemerge: don't calculate conflict labels when resolving automatically Differential Revision: https://phab.mercurial-scm.org/D11903

simplemerge: remove now-unused `localorother` argument from `merge_lines()` Differential Revision: https://phab.mercurial-scm.org/D11902

simplemerge: add a specialized function for "union", "local", "other" Differential Revision: https://phab.mercurial-scm.org/D11901

rhg: Fall back to Python if verbose status is requested by config Differential Revision: https://phab.mercurial-scm.org/D11943

rhg: Accept different "invalid ignore pattern" error message formatting At the moment rhg compiles all patterns into a single big regular expression, so it’s not practical to find out which file the invalid bit of syntax came from. Differential Revision: https://phab.mercurial-scm.org/D11942

rhg: Properly format warnings related to ignore patterns Differential Revision: https://phab.mercurial-scm.org/D11941

rhg: Sub-repositories are not supported Differential Revision: https://phab.mercurial-scm.org/D11940

simplemerge: make `localorother` a "mode" instead of a separate thing `simplemerge()` takes a `mode` argument, which can be "union", "merge" or "mergediff", and a `localorother` argument, which can be `None`, "local", or "other". The two options are not at all orthogonal -- most combinations don't make sense. Also, at least "union", "local", and "other" are very closely related. Therefore, it makes sense to combine them into one. It probably makes sense to split the `mode` argument into `resolve` and `marker_style`, where the former can be `None`, "union", "local", or "other", and the latter can be "merge", "merge3", "mergediff", or "minimize". This is a good step in that direction whether or not we end up doing that. Differential Revision: https://phab.mercurial-scm.org/D11887

simplemerge: avoid a call to `pycompat.strkwargs()` Differential Revision: https://phab.mercurial-scm.org/D11886

simplemerge: stop merging file flags As 384df4db6520 (merge: merge file flags together with file content, 2013-01-09) explains, we shouldn't do a 3-way merge of the symlink. However, since 84614212ae39 (flags: actually merge flags in simplemerge, 2020-05-16), we do that in `simplemerge.simplemerge()`. What's more, the merging of the executable flag there isn't actually necessary; it was made a no-op by the very next commit, i.e. 4234c9af515d (flags: read flag from dirstate/disk for workingcopyctx (issue5743), 2020-05-16). I found the overall flag-merging code (not the bit in `simplemerge.py`) very hard to follow, but I think I now finally understand how it works. `mergestate.resolve()` calculates the merged file flags and sets them on the local side of the merge (confusingly by calling `_restore_backup()`). Then it calls `filemerge.filemerge()`, which in turn calls `simplemerge.simplemerge()` (if premerge is enabled). That means that the flags on the local side `fcs.flags()` are already correct when the flag-merging code in `simplemerge.simplemerge()` runs. Interestingly, that code still works when the local side already has the merged value, it just doesn't change the value. Here's a truth table to explain why: ``` BLOMCAR 0000000 0011111 0101011 0111111 1000000 1010000 1100000 1111101 ``` B: Base L: Local O: Other M: Merged flags from `mergestate.resolve()`, i.e. what's called "local" when we get to `simplemerge.simplemerge()` C: `commonflags` in `simplemerge.simplemerge()`, i.e. `M & O` A: `addedflags` in `simplemerge.simplemerge()`, i.e. `(M ^ O) - B` R: Re-merged flags `simplemerge.simplemerge()`, i.e. `C | A` As you can see, the re-merged flags are always unchanged compared to the initial merged flags (R equals M). Therefore, this patch effectively backs out 84614212ae39. (I might later refactor this code to have the flags explicitly passed in.) `simplemerge.simplemerge()` is also called from `contrib/simplemerge.py`, but that code never passes any flags. Differential Revision: https://phab.mercurial-scm.org/D11879

filemerge: stop returning always-`True` value Now that we've removed some more leftovers from "merge driver", it's clear that the first element of `filemerge()`'s return value is always `True`. Differential Revision: https://phab.mercurial-scm.org/D11885

filemerge: remove unused `orig` argument from tool functions Differential Revision: https://phab.mercurial-scm.org/D11878

filemerge: stop passing around 3 unused `None` values in `files` argument Since 67cfffbfb6a0 (filemerge: eliminate most uses of tempfiles, 2017-08-31), we have used only the last entry in the `files` tuple that gets passed around to various functions. That commit said "Emphasize that they're unused so we can more easily remove them later.". It's now "later". Differential Revision: https://phab.mercurial-scm.org/D11877

filemerge: rename backup variables from `back` to `backup` `backup` seems like an obviously clearer name for something containing something related to a backup. Differential Revision: https://phab.mercurial-scm.org/D11876

setup.py: fix some documentation typos Differential Revision: https://phab.mercurial-scm.org/D11944

dirstate-v2: Apply SECOND_AMBIGUOUS to directory mtimes too This would only be relevant in contrived scenarios such as a dirstate file being written with a libc that supports sub-second precision in mtimes, then transfered (at the filesystem level, not `hg clone`) to another system where libc *doesn’t* have sub-second precision and truncates the stored mtime of a directory to integer seconds. Differential Revision: https://phab.mercurial-scm.org/D11939

rank: actually persist revision's rank in changelog-v2 The changelog v2 format is now persisting whatever "rank" value is recorded. However keep in mind that for we do not record any value. Differential Revision: https://phab.mercurial-scm.org/D11937

rank: add a "rank" value to the revlog-entry tuple The rank of a revision is the size of sub-graph it defines as a head. In other words, the rank of X is the size of `ancestors(X)` (X included). This is a property that can help various algorithm and we intend to store it in changelog-v2. We start with adding this new information to the "entry tuple", with a default value. We will start to compute and persist the rank later. Differential Revision: https://phab.mercurial-scm.org/D11936

changelog-v2: fix an assertion error to display the right data We were checking on thing and displaying the other. Differential Revision: https://phab.mercurial-scm.org/D11935

changelog-v2: use helper constant in the code to pack/unpack entries This makes the code simpler to read and modify. This is probably less efficient, but we are talking about the Python implementation here, so performance is concerne secondary to readability. Differential Revision: https://phab.mercurial-scm.org/D11934

revlog: add some information about the revision we cannot find Parameter to Exception are good, use them. Differential Revision: https://phab.mercurial-scm.org/D11933

rhg: Use binary search in manifest lookup … instead of linear scan, when looking for a single entry based on its path. Manifest entries are sorted by path, but are variable-size so we can’t use the standard library’s `[T]::binary_search`. We can still jump to a byte index and then look around for entry boundaries. Differential Revision: https://phab.mercurial-scm.org/D11932

rust: Upgrade to format-bytes 0.3 This removes use of the proc-macro-hack crate, which is possible now that we don’t support Rust 1.41 to 1.44 anymore. This in turn fixes spurious errors reported by rust-analyser: https://github.com/rust-analyzer/rust-analyzer/issues/9606#issuecomment-919240134 Differential Revision: https://phab.mercurial-scm.org/D11938

pytype: stop excluding ssutil.py Pytype simply can't distinguish the type for this one entry's value from the other values: File "/mnt/c/Users/Matt/hg/mercurial/sslutil.py", line 142, in _hostsettings: No attribute 'append' on None [attribute-error] In Optional[Union[Any, List[nothing], bool, bytes]] File "/mnt/c/Users/Matt/hg/mercurial/sslutil.py", line 142, in _hostsettings: No attribute 'append' on bool [attribute-error] In Optional[Union[Any, List[nothing], bool, bytes]] File "/mnt/c/Users/Matt/hg/mercurial/sslutil.py", line 142, in _hostsettings: No attribute 'append' on bytes [attribute-error] In Optional[Union[Any, List[nothing], bool, bytes]] File "/mnt/c/Users/Matt/hg/mercurial/sslutil.py", line 147, in _hostsettings: No attribute 'append' on None [attribute-error] In Optional[Union[Any, List[Tuple[Any, Any]], bool, bytes]] File "/mnt/c/Users/Matt/hg/mercurial/sslutil.py", line 147, in _hostsettings: No attribute 'append' on bool [attribute-error] In Optional[Union[Any, List[Tuple[Any, Any]], bool, bytes]] File "/mnt/c/Users/Matt/hg/mercurial/sslutil.py", line 147, in _hostsettings: No attribute 'append' on bytes [attribute-error] In Optional[Union[Any, List[Tuple[Any, Any]], bool, bytes]] Differential Revision: https://phab.mercurial-scm.org/D11931

pytype: stop excluding wireprotoserver.py The `config` entry is a 2 part tuple, which has `__iter__()`: File "/mnt/c/Users/Matt/hg/mercurial/wireprotoserver.py", line 253, in _availableapis: No attribute '__iter__' on Callable[[Any, Any], Any] [attribute-error] In Union[Callable[[Any, Any, Any, Any, Any], None], Callable[[Any, Any], Any]] File "/mnt/c/Users/Matt/hg/mercurial/wireprotoserver.py", line 253, in _availableapis: No attribute '__iter__' on Callable[[Any, Any, Any, Any, Any], None] [attribute-error] In Union[Callable[[Any, Any, Any, Any, Any], None], Callable[[Any, Any], Any]] Differential Revision: https://phab.mercurial-scm.org/D11930

pytype: stop excluding pycompat.py Whatever issues were here seem to have been previously fixed. Differential Revision: https://phab.mercurial-scm.org/D11929

procutil: avoid an uninitialized variable usage on tempfile exception If `pycompat.unnamedtempfile()` raises an exception, it would have called `stdin.close()` in the `finally` block without it being initialized first. Differential Revision: https://phab.mercurial-scm.org/D11928

pytype: stop excluding procutil.py This avoids these false warnings: File "/mnt/c/Users/Matt/hg/mercurial/utils/procutil.py", line 78, in <module>: No attribute 'register' on Type[io.BufferedIOBase] [attribute-error] File "/mnt/c/Users/Matt/hg/mercurial/utils/procutil.py", line 117, in <module>: No attribute 'register' on Type[io.IOBase] [attribute-error] File "/mnt/c/Users/Matt/hg/mercurial/utils/procutil.py", line 770, in runbgcommandpy3: No attribute 'close' on int [attribute-error] In Union[IO[Union[bytes, str]], int] Differential Revision: https://phab.mercurial-scm.org/D11927

pytype: stop excluding chgserver.py This teaches pytype about some lazy initialization, and avoids the following: File "/mnt/c/Users/Matt/hg/mercurial/chgserver.py", line 666, in _checkextensions: No attribute '_hashstate' on chgunixservicehandler [attribute-error] File "/mnt/c/Users/Matt/hg/mercurial/chgserver.py", line 672, in _checkextensions: No attribute '_hashstate' on chgunixservicehandler [attribute-error] File "/mnt/c/Users/Matt/hg/mercurial/chgserver.py", line 677, in _bind: No attribute '_realaddress' on chgunixservicehandler [attribute-error] File "/mnt/c/Users/Matt/hg/mercurial/chgserver.py", line 683, in _bind: No attribute '_realaddress' on chgunixservicehandler [attribute-error] File "/mnt/c/Users/Matt/hg/mercurial/chgserver.py", line 686, in _createsymlink: No attribute '_baseaddress' on chgunixservicehandler [attribute-error] File "/mnt/c/Users/Matt/hg/mercurial/chgserver.py", line 686, in _createsymlink: No attribute '_realaddress' on chgunixservicehandler [attribute-error] File "/mnt/c/Users/Matt/hg/mercurial/chgserver.py", line 688, in _createsymlink: No attribute '_baseaddress' on chgunixservicehandler [attribute-error] File "/mnt/c/Users/Matt/hg/mercurial/chgserver.py", line 689, in _createsymlink: No attribute '_realaddress' on chgunixservicehandler [attribute-error] File "/mnt/c/Users/Matt/hg/mercurial/chgserver.py", line 690, in _createsymlink: No attribute '_baseaddress' on chgunixservicehandler [attribute-error] Differential Revision: https://phab.mercurial-scm.org/D11926

pytype: stop excluding webcommands.py I have no idea why, but asserting that each value added to `emptydirs` is not None didn't fix this: File "/mnt/c/Users/Matt/hg/mercurial/hgweb/webcommands.py", line 621, in dirlist: Function bytes.join was called with the wrong arguments [wrong-arg-types] Expected: (self, iterable: Iterable[bytes]) Actually passed: (self, iterable: List[None]) Differential Revision: https://phab.mercurial-scm.org/D11924

tests: drop an obsolete comment about a pytype failure Differential Revision: https://phab.mercurial-scm.org/D11923

pytype: stop excluding patch.py The underlying `email.generator.BytesGenerator` is documented as requiring an `fp` that accepts bytes, so I'm not sure why pytype is getting confused: File "/mnt/c/Users/Matt/hg/mercurial/patch.py", line 112, in msgfp: Function Generator.__init__ was called with the wrong arguments [wrong-arg-types] Expected: (self, outfp: TextIO, ...) Actually passed: (self, outfp: io.BytesIO, ...) Differential Revision: https://phab.mercurial-scm.org/D11922

pytype: stop excluding statprof.py This seems to have worked fine before (at least on Linux). We could just add suppression comments, but this file already imports from the mercurial package, which seems to prevent this from running as a standalone program because of the relative import of `pycompat`. PyCharm isn't happy either way. File "/mnt/c/Users/Matt/hg/mercurial/statprof.py", line 501, in display: Function TextIO.write was called with the wrong arguments [wrong-arg-types] Expected: (self, s: str) Actually passed: (self, s: bytes) Called from (traceback): line 1091, in main File "/mnt/c/Users/Matt/hg/mercurial/statprof.py", line 501, in display: Function TextIO.write was called with the wrong arguments [wrong-arg-types] Expected: (self, s: str) Actually passed: (self, s: bytes) Called from (traceback): line 431, in profile File "/mnt/c/Users/Matt/hg/mercurial/statprof.py", line 522, in display: Function TextIO.write was called with the wrong arguments [wrong-arg-types] Expected: (self, s: str) Actually passed: (self, s: bytes) Called from (traceback): line 1091, in main File "/mnt/c/Users/Matt/hg/mercurial/statprof.py", line 522, in display: Function TextIO.write was called with the wrong arguments [wrong-arg-types] Expected: (self, s: str) Actually passed: (self, s: bytes) Called from (traceback): line 431, in profile File "/mnt/c/Users/Matt/hg/mercurial/statprof.py", line 523, in display: Function TextIO.write was called with the wrong arguments [wrong-arg-types] Expected: (self, s: str) Actually passed: (self, s: bytes) Called from (traceback): line 1091, in main File "/mnt/c/Users/Matt/hg/mercurial/statprof.py", line 523, in display: Function TextIO.write was called with the wrong arguments [wrong-arg-types] Expected: (self, s: str) Actually passed: (self, s: bytes) Called from (traceback): line 431, in profile File "/mnt/c/Users/Matt/hg/mercurial/statprof.py", line 524, in display: Function TextIO.write was called with the wrong arguments [wrong-arg-types] Expected: (self, s: str) Actually passed: (self, s: bytes) Called from (traceback): line 1091, in main File "/mnt/c/Users/Matt/hg/mercurial/statprof.py", line 524, in display: Function TextIO.write was called with the wrong arguments [wrong-arg-types] Expected: (self, s: str) Actually passed: (self, s: bytes) Called from (traceback): line 431, in profile File "/mnt/c/Users/Matt/hg/mercurial/statprof.py", line 796, in _write: Function TextIO.write was called with the wrong arguments [wrong-arg-types] Expected: (self, s: str) Actually passed: (self, s: bytes) Differential Revision: https://phab.mercurial-scm.org/D11921

statprof: convert a few exception byte strings to str That way these display without the b'' prefix. Differential Revision: https://phab.mercurial-scm.org/D11920

pytype: stop excluding wireprotov2server.py This fixes: File "/mnt/c/Users/Matt/hg/mercurial/wireprotov2server.py", line 584, in _capabilitiesv2: unsupported operand type(s) for item assignment: Set[bytes] [unsupported-operands] No attribute '__setitem__' on Set[bytes] File "/mnt/c/Users/Matt/hg/mercurial/wireprotov2server.py", line 611, in _capabilitiesv2: No attribute 'append' on dict [attribute-error] In Union[List[bytes], List[nothing], dict] Called from (traceback): line 543, in httpv2apidescriptor Differential Revision: https://phab.mercurial-scm.org/D11919

pytype: stop excluding stringutil.py This fixes the following: File "/mnt/c/Users/Matt/hg/mercurial/utils/stringutil.py", line 267, in prettyrepr: Function bytes.startswith expects 2 arg(s), got 3 [wrong-arg-count] Expected: (self, prefix) Actually passed: (self, prefix, _) File "/mnt/c/Users/Matt/hg/mercurial/utils/stringutil.py", line 695, in escapestr: No attribute 'escape_encode' on module 'codecs' [module-attr] File "/mnt/c/Users/Matt/hg/mercurial/utils/stringutil.py", line 699, in unescapestr: No attribute 'escape_decode' on module 'codecs' [module-attr] Differential Revision: https://phab.mercurial-scm.org/D11918

pytype: stop excluding cmdutil.py Whatever was broken here seems to have been previously fixed. Differential Revision: https://phab.mercurial-scm.org/D11917

pytype: stop excluding copies.py I can't prove that `targetrev` is always in `all_copies`, but it would have been a runtime error before too if it's not. Differential Revision: https://phab.mercurial-scm.org/D11916

copies: fix some documentation typos Differential Revision: https://phab.mercurial-scm.org/D11915

copies: drop an unused variable Differential Revision: https://phab.mercurial-scm.org/D11914

pytype: stop excluding lsprof.py Whatever was previously flagged in this file appears to have been fixed. Differential Revision: https://phab.mercurial-scm.org/D11913

pytype: stop excluding changegroup.py The false positives that were detected seem to be related to what happens to the variables after the local methods are used: File "/mnt/c/Users/Matt/hg/mercurial/changegroup.py", line 353, in ondupchangelog: No attribute 'append' on range [attribute-error] In Union[List[nothing], range] File "/mnt/c/Users/Matt/hg/mercurial/changegroup.py", line 357, in onchangelog: No attribute 'update' on None [attribute-error] Differential Revision: https://phab.mercurial-scm.org/D11912

pytype: stop excluding upgrade.py The sole failure here was this, which is fixed by simply creating a set like the caller, instead of a dict: File "/mnt/c/Users/Matt/hg/mercurial/upgrade.py", line 73, in upgraderepo: No attribute 'discard' on Dict[nothing, nothing] [attribute-error] In Union[Any, Dict[nothing, nothing]] Differential Revision: https://phab.mercurial-scm.org/D11911

rhg: Skip reading the contents of ambiguous files in some cases If the size of the file in the working directory does not match the length of the filelog data, we know its contents will be different and don’t need to read it. rhg still decodes the filelog revision, which is not needed in some cases. Differential Revision: https://phab.mercurial-scm.org/D11910

rhg: Mark it as expected that the issue6528 bug is not reproduced Differential Revision: https://phab.mercurial-scm.org/D11909

filemerge: remove leftover documentation of removed argument I removed the `premerge` argument in https://phab.mercurial-scm.org/D11859. Differential Revision: https://phab.mercurial-scm.org/D11925

merge with stable

rhg: Print "bad match" errors in rhg status Previously these would only be visible if enabled with some RUST_LOG environment variable. Differential Revision: https://phab.mercurial-scm.org/D11908

share: make it possible to control the working copy format variant A share will use the same format as its source for the store, but there are no reason to not lets it control the working copy variant at creation time. So we make it so. Differential Revision: https://phab.mercurial-scm.org/D11892

share: move the requirement initialisation code around We will make this logic more flexible in the next patch. We start by moving it at the end of the function without any other change for clarity. Differential Revision: https://phab.mercurial-scm.org/D11891

test: use `hg debugrequires` instead of `cat` in some tests This give the "same" result, while taking in account that the requirement file might be in different location, like what `share-safe` is doing. Differential Revision: https://phab.mercurial-scm.org/D11895

test: simplify `debugformat` matching in test-phases.t This will make the test less sensible to unrelated format changing. Differential Revision: https://phab.mercurial-scm.org/D11894

test: simplify `debugformat` matching in test-copies-chain-merge This will make the test less sensible to unrelated format changing. Differential Revision: https://phab.mercurial-scm.org/D11893

status: when extracting arguments from `opts`, use the same default values Sometimes other code, such as commit when using `commands.commit.post-status`, calls `commands.status()` without going through the normal dispatch mechanism that would typically fill in the args to be something besides None. As a "defense in depth" mechanism for a bug where Mercurial would crash if both `commands.commit.post-status` and `experimental.directaccess` were enabled, let's sanitize these values to be identical to the values they would have when the user invoked this method from the commandline. Differential Revision: https://phab.mercurial-scm.org/D11884

directaccess: fix uses of commands.status() that don't go through flag parsing When `commands.commit.post-status` is enabled, after commit/amend, commands.status() is called without any revs argument, which means that status gets None instead of an empty list like it would receive if the user had invoked this on the commandline. With the `experimental.directaccess` config enabled, this gets passed to `unhidehashlikerevs`, which didn't previously handle None, but now should. Differential Revision: https://phab.mercurial-scm.org/D11883

test: add test for a former race resulting in bad dirstate In 6.0 this used to result in the size being stored in the dirstate is wrong. This was fixed by other change to the mtime gathering logic. Differential Revision: https://phab.mercurial-scm.org/D11749

rhg: fix a crash on non-generaldelta revlogs Differential Revision: https://phab.mercurial-scm.org/D11882

rhg: centralize index header parsing Centralize index header parsing, parse the generaldelta flag, and leave breadcrumbs to relate the code to python. Differential Revision: https://phab.mercurial-scm.org/D11881

rhg: demonstrate that rhg breaks on non-generaldelta revlogs Differential Revision: https://phab.mercurial-scm.org/D11880

tests: add a short `sleep` in test-status.t With dirstate-v2 and rhg both enabled, this test would sometimes fail for me with: ``` --- tests/test-status.t +++ tests/test-status.t#dirstate-v2.err @@ -943,7 +943,7 @@ $ rm subdir/unknown $ hg status $ hg debugdirstate --all --no-dates | grep '^ ' - 0 -1 set subdir + 0 -1 unset subdir ``` Meaning that `status` did not write a directory mtime in the dirstate as expected. This can happen if the observed mtime of the directory is the same as "current time" at the start of `status`. This current time is obtained by creating a temporary file and checking its mtime. Even with ext4 on my system being able to store nanosecond precision, identical mtime for successive but separate operations is still possible becuse the kernel may cache the current time: https://stackoverflow.com/a/14393315/1162888 0.1 second should be enough for this cache to be updated, without significantly slowing down the test. Differential Revision: https://phab.mercurial-scm.org/D11900

rhg: Add support for `rhg status --copies` Copy sources are collected during `status()` rather than after the fact like in Python, because `status()` takes a `&mut` exclusive reference to the dirstate map (in order to potentially mutate it for directory mtimes) and returns `Cow<'_, HgPath>` that borrow the dirstate map. Even though with `Cow` only some shared borrows remain, the still extend the same lifetime of the initial `&mut` so the dirstate map cannot be borrowed again to access copy sources after the fact: https://doc.rust-lang.org/nomicon/lifetime-mismatch.html#limits-of-lifetimes Additionally, collecting copy sources during the dirstate tree traversal that `status()` already does avoids the cost of another traversal or other lookups (though I haven’t benchmarked that cost). Differential Revision: https://phab.mercurial-scm.org/D11899

rhg: refactor relativize_path into a struct + method … instead of a function that takes an iterator and a callback. Differential Revision: https://phab.mercurial-scm.org/D11898

rhg: refactor display_status_paths with a struct for common arguments Differential Revision: https://phab.mercurial-scm.org/D11897

rhg: Add support for ui.ignore and ui.ignore.* config This fixes some but not all failures in `tests/test-hgignore.t` when running with `rhg status` enabled. Differential Revision: https://phab.mercurial-scm.org/D11907

doc: remove a spurious ^L in some doc This was wrongly introduced in D11784. Differential Revision: https://phab.mercurial-scm.org/D11896

upgrade: drop some dead code Everything done by this block is now done by earlier code. So we don't need it anymore. Differential Revision: https://phab.mercurial-scm.org/D11872

upgrade: only process revlogs that needs it by default We have more and more requirement that does not affect revlog or that only affect some of them. It is silly to force a full processing of all revlog to juste move the requirement around, or to simply rewrite the dirstate. So now, only the revlog that needs to be touched will be touched. Unless the --changelog & al flags are used. Differential Revision: https://phab.mercurial-scm.org/D11871

upgrade: issue a message when a revlog type has to be upgraded This is more explicite and prepare for a smoother transition to smarter picking of the revlog we will process. Differential Revision: https://phab.mercurial-scm.org/D11870

upgrade: explicitly warn when a `--no-xxx` flag is overwritten Some format upgrade/downgrades -needs- revlog to be recomputed. So we now detect that individually and warn when it contradict explicitly passed flag. This is part of a larger series to make `debugupgraderepo` smarter about which revlog it picks by default. Differential Revision: https://phab.mercurial-scm.org/D11869

upgrade: make the list of explicitly specified revlog a dict This makes various logic simpler and will help making future patch clearer. Differential Revision: https://phab.mercurial-scm.org/D11868

upgrade: move the revlog selection code lower down the chain We about about to make revlog section smarter. Moving the code around will make the next changesets clearer. Differential Revision: https://phab.mercurial-scm.org/D11867

rhg: Set second_ambiguous as needed in post-status fixup This fixes an intermittent bug that manifested only in test-revert.t, and unfortunately not on CI. On a fast enough machine we could have: 1. A file is modified 2. `rhg status` writes an updated dirstate-v1 3. The same file is modified again … all within the same integer second. Because the dirstate-v1 file format does not store sub-second precision, step 2 must write the file’s mtime as "unknown" because of the possibility of step 3. However, most of the code now handles timestamps with nanosecond precision in order to take advantage of it in dirstate-v2. `second_ambiguous` must be set for timestamps that become ambiguous if sub-second precision is dropped (such as through serialization in dirstate-v1 format). Differential Revision: https://phab.mercurial-scm.org/D11889

dirstate: Document Timestamp.second_ambiguous Differential Revision: https://phab.mercurial-scm.org/D11888

rhg: support the new extension suboptions syntax See inline comments Differential Revision: https://phab.mercurial-scm.org/D11874

branching: merge stable into default

filemerge: remove `premerge` argument from `_makebackup()` We now always pass `True`, so there's no need to pass it. Differential Revision: https://phab.mercurial-scm.org/D11863

filemerge: inline `_filemerge()` into `filemerge()` The caller is now trivial, so the indirection is pointless. Differential Revision: https://phab.mercurial-scm.org/D11860

filemerge: make `_filemerge()` do both premerge and merge This patch removes the `premerge` argument from `_filemerge()` and makes it do both the "premerge" and "merge" steps without the caller having to call it twice. Differential Revision: https://phab.mercurial-scm.org/D11859

filemerge: merge `premerge()` into `filemerge()` Differential Revision: https://phab.mercurial-scm.org/D11858

mergestate: inline `_resolve()` into `resolve()` The caller is now trivial, so the indirection is pointless. Differential Revision: https://phab.mercurial-scm.org/D11857

mergestate: make `_resolve()` do both preresolve and resolve This patch removes the `preresolve` argument from `_resolve()` and makes it do both the "preresolve" and "resolve" steps without the caller having to call it twice. Differential Revision: https://phab.mercurial-scm.org/D11856

mergestate: merge `preresolve()` into `resolve()` The separation between `preresolve()` and `resolve()` was something added for the "merge driver" feature. Since we removed that in 32ce4cbaec4b (mergedriver: delete it, 2020-09-17), we don't need the separation anymore. More cleanups will follow. Differential Revision: https://phab.mercurial-scm.org/D11855

tests: remove unnecessary explicit args from `failfilemerge.py` The `failfilemerge()` function always raises an exception without looking at its arguments, so let's use `*args` and `**kwargs`. Differential Revision: https://phab.mercurial-scm.org/D11854

simplemerge: set `conflicts` earlier for :union tool Differential Revision: https://phab.mercurial-scm.org/D11853

typo: s/unkown/unknown across the codebase Differential Revision: https://phab.mercurial-scm.org/D11807

rhg: Update the dirstate on disk after status Differential Revision: https://phab.mercurial-scm.org/D11840

rhg: Add Repo::write_dirstate This method is not used yet. It saves to disk any mutation that was done to the `Repo`’s dirstate through `Repo::dirstate_map_mut`. It takes care of dirstate-v1 v.s. dirstate-v2, dockets, data files, appending when possible, etc. Differential Revision: https://phab.mercurial-scm.org/D11839

rhg: Add lazy/cached dirstate data file ID parsing on Repo The `dirstate_parents`, `dirstate_data_file_uuid`, and `dirstate_map` members of `Repo` can be access in any order and the `.hg/dirstate` file should only be opened once. Differential Revision: https://phab.mercurial-scm.org/D11838

rhg: Make Repo::dirstate_parents a LazyCell Unify with the same abstraction used for other lazily-initialized components Differential Revision: https://phab.mercurial-scm.org/D11837

rust: Add Vfs::write_atomic This method writes to a temporary file then renames in place Differential Revision: https://phab.mercurial-scm.org/D11836

rhg: Initial repository locking Initial Rust implementation of locking based on the `.hg/wlock` symlink (or file), with lock breaking when the recorded pid and hostname show that a lock was left by a local process that is not running anymore (as it might have been killed). Differential Revision: https://phab.mercurial-scm.org/D11835

rust: Serializing a DirstateMap does not mutate it anymore Differential Revision: https://phab.mercurial-scm.org/D11834

rhg: don't run `blackbox` if not activated You currently have no way of turning off blackbox. Aside from being a bug, this can be annoying if you use `rhg` in your shell prompt, but also because the current implementation of blackbox is quite slow due to `user` querying. Differential Revision: https://phab.mercurial-scm.org/D11813

rust: add function to check if an extension is enabled This will be used in the next patch Differential Revision: https://phab.mercurial-scm.org/D11812

dirstate: remove unused method Its last usage was inlined in `merge.py`. Differential Revision: https://phab.mercurial-scm.org/D11849

python2-compat: fix extensions' tests on python2 They are minor output different to be taken in account. Differential Revision: https://phab.mercurial-scm.org/D11861

rhg: add support for narrow clones and sparse checkouts This adds a minimal support that can be implemented without parsing the narrowspec. We can parse the narrowspec and add support for more operations later. The reason we need so few code changes is as follows: Most operations need no special treatment of sparse because some of them only read dirstate (`rhg files` without `-r`), which bakes in the filtering, some of them only read store (`rhg files -r`, `rhg cat`), and some of them read no data at all (`rhg root`, `rhg debugrequirements`). `status` is the command that might care about sparse, so we just disable rhg on it. For narrow clones, `rhg files` clearly needs the narrowspec to work correctly, so we fall back. `rhg cat` seems to work consistently with `hg cat` if the file exists. If the file is hidden by narrow spec, the error message is different and confusing, so that's something that we should improve in follow-up patches. Differential Revision: https://phab.mercurial-scm.org/D11764

update: reformat the `commands.update.check` help This make the various option clearer. Differential Revision: https://phab.mercurial-scm.org/D11852

update: properly overwrite `check` config with `--no-` variants Before this change `--merge` override `commands.update.check=abort/noconflict`, and `--check` "override" `commands.update.check=none/linear`. However the `--no-merge` and `--no-check` version where not doing anything. The `--no-` two flags now behave properly. Differential Revision: https://phab.mercurial-scm.org/D11851

update: add a test for `--merge` overriding the "abort" config I don't see this behavior tested anywhere. Differential Revision: https://phab.mercurial-scm.org/D11850

update: use long version of the flag in the test This helps readability and searchability. Differential Revision: https://phab.mercurial-scm.org/D11848

backout: backed out changeset 5d83a40cd3f8 This is the re-send of the backout in D11757 that I accidentally queued. I'll rewrite the original commit message: ...sigh We do not have Python 3 packaging for thg on Windows where the vast majority of of users use the thg installer for Mercurial. Until this is done (hopefully this cycle), we're keeping Python. (end of the old commit message) On top of this, we have a solid lead to have competent people take care of this packaging issue for us in time for the 6.1 release, which is really the main reason for us to wait. We're trying our best to make this work, so please bear with us. Differential Revision: https://phab.mercurial-scm.org/D11770

status: keep second-ambiguous mtimes during fixup Now that we support the feature, we can keep "second ambiguous" mtime during the fixup phase. These are the mtime that would be ambiguous if we did not had sub-second précions. See the v2 format documentation for details. Differential Revision: https://phab.mercurial-scm.org/D11847

dirstate-item: make sure we load `mtime-second-ambiguous` from disk Now that we support the associated logic, we can safely load it from it. It is no longer necessary to ignore the stored mtime when the flag is encountered. Differential Revision: https://phab.mercurial-scm.org/D11846

dirstate-item: make sure we set the mtime-second-ambiguous on v2 write We want to preserve the second-ambiguity alongside the ambiguous mtime. So we use the decimated flag for that. note: the C code was already doing so. No change was needed to it. Differential Revision: https://phab.mercurial-scm.org/D11845

dirstate-item: ignore mtime to write v1 when `mtime-second-ambiguous` is set We cannot preserve that information in the v1 format so that mtime is ambiguous. Differential Revision: https://phab.mercurial-scm.org/D11844

dirstate-item: implement the comparison logic for mtime-second-ambiguous If the flag is set we now process it properly. We "just" need to actually set it and persist it. Differential Revision: https://phab.mercurial-scm.org/D11843

dirstate-item: add a "second_ambiguous` flag in the mtime tuple This will be used to support the `mtime-second-ambiguous` flag from dirstate v2. See format documentation for details. For now, we only make it possible to store the information, no other logic have been added. Differential Revision: https://phab.mercurial-scm.org/D11842

dirstate: drop comparison primitive on the timestamp class All comparison are now managed without using operator : - the status mtime comparisons is handled by the DirstateItem, - the fixup reliability check, - the update "hack". So we no longer needs the operator and should discourage its usage. Differential Revision: https://phab.mercurial-scm.org/D11841

largefile: use the proper "mtime boundary" logic during fixup This will prevent ambiguous cache entry to be used in racy situation. This fix flakiness in test and some real live misbehavior. Differential Revision: https://phab.mercurial-scm.org/D11800

status: move the boundary comparison logic within the timestamp module Some extensions will need it too. So lets isolate the logic. It also makes things clearer. Differential Revision: https://phab.mercurial-scm.org/D11799

tests: remove potential mtime ambiguity in a dirstate test If the test was fast enough, some mtime where not stored. We now wait long enough to ensure the mtime is no longer ambiguous. Differential Revision: https://phab.mercurial-scm.org/D11798

dirstate: cleanup remaining of "now" during write Since the whole `need_delay` have been removed, we no longer need this. Differential Revision: https://phab.mercurial-scm.org/D11797

dirstate: remove need_delay logic Now that all¹ stored mtime are non ambiguous, we no longer need to apply the `need_delay` step. The need delay logic was not great are mtime gathered during longer operation could be ambiguous but younger than the `dirstate.write` call time. So, we don't need that logic anymore and can drop it This make the code much simpler. The code related to the test extension faking the dirstate write is now obsolete and associated test will be migrated as follow up. They currently do not break. [1] except the ones from `hg update`, but `need_delay` no longer help for them either. Differential Revision: https://phab.mercurial-scm.org/D11796

dirstate: remove `lastnormaltime` mechanism This is now redundant with the new, simpler `mtime_boundary` one. Differential Revision: https://phab.mercurial-scm.org/D11795

status: use filesystem time boundary to invalidate racy mtime We record the filesystem time at the start of the status walk and use that as a boundary to detect files that might be modified during (or right after) the status run without the mtime allowing that edition to be detected. We currently do this at a second precision. In a later patch, we will use nanosecond precision when available. To cope with "broken" time on the file system where file could be in the future, we also keep mtime for file over one day in the future. See inline comment for details. Large file tests get a bit more confused as we reduce the odds for race condition. As a "side effect", the win32text extension is happy again. Differential Revision: https://phab.mercurial-scm.org/D11794

test: use a different timestamp for the updated file In the test we want to trigger a write after the underlying dirstate changed. To do so, we need a write. And as we are about to make dirstate update smarter we need to meddle with the script a bit to make sure there will be a write. Differential Revision: https://phab.mercurial-scm.org/D11793

win32text: drop associated dirstate cache information on revert Otherwise the could get size from one version of the file while the on-disk version is still clean but with another size. This fix the previously introduced error. Differential Revision: https://phab.mercurial-scm.org/D11792

dirstate: stop gathering parentfiledata in update_file Gathering information here assume that they are valid cache information for a clean file. It is true most of the time, but not garanteed. Accurate data can still be explicitly provided. We drop the spontaneous and will let the next `hg status` call record actual information. Differential Revision: https://phab.mercurial-scm.org/D11791

dirstate: stop gathering parentfiledata in update_file_p1 Gathering information here assume that they are valid cache information for a clean file. It is true most of the time, but not garanteed. So we drop this and will let the next `hg status` call record actual information. Differential Revision: https://phab.mercurial-scm.org/D11790

dirstate: make it mandatory to provide parentfiledata in `set_clean` Gathering the mode, size and mtime, independently from determining that the file is clean is a race-machine. So we just make these information required arguments. (note that the data is still gathered in a racy way in practice, but at least the API is no longer encouraging it.) Differential Revision: https://phab.mercurial-scm.org/D11789

dirstate: do no use `set_clean` in revert The current `set_clean` usage is racy (the file might be modified between its restoration and the `set_clean` call). So we simply leave the file as ambiguous and the next status will fix that. We still have to make sure the copy information is dropped, so we teach dirstate how to do that. The win32txt extension is confused after this because current logic is broken in more location. However this series will ultimately fix that so we "ignore" it for now. Fixing it now is complicated without some extra fix landing later. Differential Revision: https://phab.mercurial-scm.org/D11788

status: adapt the "keyword" extensions to gather stats at lookup time See main core code for details. We don't factor the code in a common function yet, because we will have to adapt a bit more things in the keyword case at the end of the series. Differential Revision: https://phab.mercurial-scm.org/D11787

status: adapt largefile to gather stats at lookup time See the core code for details of why we are doing this. We don't factor the code in a common function yet, because we will have to adapt a bit more things in the largefile case at the end of the series. Differential Revision: https://phab.mercurial-scm.org/D11786

status: gather fixup info at comparison time This is still racy, but on a much small windows. In addition, the API now make it possible for it to not be racy. This also unlock other cleanups that we are about to do regarding mtime ambiguity at gathering time. Differential Revision: https://phab.mercurial-scm.org/D11785

update: filter the ambiguous mtime in update directly Right now, this filtering is done by `dirstate.write` using the time of `dirstate.write` method call. However that filtering is done "too late" It works "fine" as most command are "fast enough", and race rare enough. We are about to change the mtime filtering logic in the dirstate to be more accurate and reliable. However `hg update` will still need such filtering (mostly because it is actually quite racy, even with the existing filtering). So we explicitly implement a similar logic here. Before removing the older one later in the series. Differential Revision: https://phab.mercurial-scm.org/D11784

dirstate: move "get fs now" in the timestamp utility module We will need it during update. Differential Revision: https://phab.mercurial-scm.org/D11783

dirstate-item: allow mtime to be None in "parentdata" This will be useful to filter out unreliable mtime. Differential Revision: https://phab.mercurial-scm.org/D11782

dirstate: add a comment about a racy piece of code during updates This is a bit that is not really correct but works "fine" in practice. Let us write the details down so that people stop wondering how that logic might be correct… It is not. Differential Revision: https://phab.mercurial-scm.org/D11781

tests: add missing `head` for when things go wrong See comment above the changed line, you can get a millions of line of output in case of failure. When don't need to have them all. Differential Revision: https://phab.mercurial-scm.org/D11780

tests: ensure a status will have non ambiguous mtime in some race test For the test to work, we need some mtime to be recorded. For them to be recorded, they need to be "clearly in the past", otherwise edit with the same mtime would be possible. Strictly speaking there might be file system with a minimal mtime increment longer than one second. However it is unlikely that we will run the test on them for now. We can be smarter about this in the future if it becomes necessary. Differential Revision: https://phab.mercurial-scm.org/D11779