Pierre-Yves David <pierre-yves.david@octobus.net> [Fri, 02 Jul 2021 02:17:18 +0200] rev 47493
dirstate: add an explicit `possibly_dirty` parameter to `_addpath`
This lets says what we mean instead of using magic value. The lower level can
then decide how to express that.
Differential Revision: https://phab.mercurial-scm.org/D10932
Pierre-Yves David <pierre-yves.david@octobus.net> [Fri, 02 Jul 2021 02:01:10 +0200] rev 47492
dirstate: add an explicit `from_p2` parameter to `_addpath`
This lets says what we mean instead of using magic value. The lower level can
then decide how to express that.
Differential Revision: https://phab.mercurial-scm.org/D10931
Pierre-Yves David <pierre-yves.david@octobus.net> [Fri, 02 Jul 2021 02:05:47 +0200] rev 47491
dirstate: add default value to _addpath
We want to remove the magic value usage from the higher level, so lets stop
passing them explicitely when possible.
Differential Revision: https://phab.mercurial-scm.org/D10930
Pierre-Yves David <pierre-yves.david@octobus.net> [Fri, 02 Jul 2021 01:02:46 +0200] rev 47490
dirstate: move the _rangemask filtering closer to its storage
The goal is to get ready to move these kind of processing at a lower level. We start with move
move _rangemask filtering close to where it is sent to the lower level to make the future move trivial.
Differential Revision: https://phab.mercurial-scm.org/D10929
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 01 Jul 2021 19:15:30 +0200] rev 47489
dirstate: introduce a symbolic constant for the AMBIGUOUS_TIME marker
This is going to be clearer and easier to track than -1. Ultimately I would
like to get ride of this special value everywhere but in the lower level,
however we need to clarify the API first. This changeset is part of such
clarification.
Differential Revision: https://phab.mercurial-scm.org/D10928
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 01 Jul 2021 19:15:22 +0200] rev 47488
dirstate: introduce a symbolic constant for the NONNORMAL marker
This is going to be clearer and easier to track than -1. Ultimately I would
like to get ride of this special value everywhere but in the lower level,
however we need to clarify the API first. This changeset is part of such
clarification.
Differential Revision: https://phab.mercurial-scm.org/D10927
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 01 Jul 2021 19:04:11 +0200] rev 47487
dirstate: introduce a symbolic constant for the FROM_P2 marker
This is going to be clearer and easier to track than -2. Ultimately I would
like to get ride of this special value everywhere but in the lower level,
however we need to clarify the API first. This changeset is part of such
clarification.
Differential Revision: https://phab.mercurial-scm.org/D10926
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 01 Jul 2021 18:59:29 +0200] rev 47486
dirstate: split a not-so-one-liner
This is shorter and simpler to read.
Differential Revision: https://phab.mercurial-scm.org/D10925
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 01 Jul 2021 18:57:35 +0200] rev 47485
dirstate: split a not-so-one-liner
This is simpler to read.
Differential Revision: https://phab.mercurial-scm.org/D10924
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 01 Jul 2021 18:56:34 +0200] rev 47484
dirstate: split a not-so-one-liner
This is shorter and simpler to read.
Differential Revision: https://phab.mercurial-scm.org/D10923
Simon Sapin <simon.sapin@octobus.net> [Mon, 28 Jun 2021 16:50:19 +0200] rev 47483
dirstate-tree: Keep a counter of descendant nodes that have an entry
… and change the `DirstateMap::has_dir` method to be based on this counter
being non-zero instead of the mere presence of a node.
A node with zero descendent with an entry currently should be removed from
the tree, but soon we’ll make the dirstate track additional nodes.
(Specifically, for non-ignored directories in order to keep track of their
mtime and optimize status by doing fewer `read_dir` calls.)
Differential Revision: https://phab.mercurial-scm.org/D10922
Simon Sapin <simon.sapin@octobus.net> [Mon, 28 Jun 2021 15:52:10 +0200] rev 47482
dirstate: Removed unused instances of `DirsMultiset`
… in Rust-backed dirstatemap.
The Python class `dirstatemap` had cached properties `_dirs` and `_alldirs`
that were not used for `hastrackeddir` and `hasdir` since they were redundant
with corresponding fields for the Rust `DirstateMap` struct.
`dirfoldmap` is modified to reuse instead the directory iterator introduced
in 3b9914b28133c0918186b6e8b9e4f1916e21338d.
Differential Revision: https://phab.mercurial-scm.org/D10921
Simon Sapin <simon.sapin@octobus.net> [Mon, 28 Jun 2021 15:41:50 +0200] rev 47481
dirstate-v2: Use 32-bit integers instead of 64-bit for offsets
This saves 12 bytes per node. (Nodes representing files or directories.)
These are offsets to other parts of the file. This would only be a limitation
for a `.hg/dirstate` file larger than 4 GiB, which would only happen for a
repository with dozens of millions of files and directories.
Differential Revision: https://phab.mercurial-scm.org/D10920
Simon Sapin <simon.sapin@octobus.net> [Thu, 24 Jun 2021 21:54:14 +0200] rev 47480
status: Extend read_dir caching to directories with ignored files
See code comments
Differential Revision: https://phab.mercurial-scm.org/D10909
Simon Sapin <simon.sapin@octobus.net> [Fri, 04 Jun 2021 15:26:38 +0200] rev 47479
dirstate-v2: Drop cached read_dir results after .hgignore changes
Soon we’ll want the status algorithm to be able to skip `std::fs::read_dir` in
more cases, notabling when listing unknown files but not ignored files.
When ignore patterns change (which we detect by their hash, added to the
dirstate-v2 format in a previous changeset), a formerly-ignored file could
become unknown without changing its parent directory’s modification time.
Therefore we remove any directory mtime from the dirstate, effictively
invalidating the existing caches.
Differential Revision: https://phab.mercurial-scm.org/D10907