Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 15 Jul 2021 04:55:57 +0200] rev 47694
dirstate: use `reset_state` in `update_file_p1`
Going through the same API is more consistent and allow us to push
implementation lower down the call stack.
Differential Revision: https://phab.mercurial-scm.org/D11136
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 15 Jul 2021 04:34:51 +0200] rev 47693
dirstate: rename `update_file_reference` to `update_file_p1`
This is a bit clearer on the intended usage of the function. (but still not great)
Differential Revision: https://phab.mercurial-scm.org/D11135
Pierre-Yves David <pierre-yves.david@octobus.net> [Mon, 19 Jul 2021 07:23:55 +0200] rev 47692
dirstate-map: move most of `dirstate.update_file` logic in the dsmap
A new `reset_state` method is introduced to deal with most of that logic. This
move things one layer lower, but the ultimate goal is to deal with most of this
at the DirstateItem level.
This reveal various imperfection with the data passed to update_file by
`mergestate.recordupdates`, however this is orthogonal to this patch and should
be dealt with at a higher level.
Differential Revision: https://phab.mercurial-scm.org/D11134
Pierre-Yves David <pierre-yves.david@octobus.net> [Fri, 16 Jul 2021 22:30:11 +0200] rev 47691
dirstate: replace `update_parent_file_data` with simpler `update_parent` call
This make the dirstate API simpler.
Differential Revision: https://phab.mercurial-scm.org/D11133
Pierre-Yves David <pierre-yves.david@octobus.net> [Fri, 16 Jul 2021 15:07:16 +0200] rev 47690
dirstate: factor out the part retrieve "filedata" out of `normal`
We will need them elsewhere.
Differential Revision: https://phab.mercurial-scm.org/D11132
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 15 Jul 2021 02:19:41 +0200] rev 47689
dirstate-map: factor out the change to _dirs and _alldirs on removing
This logic is complicated enough to deserves its own function. So it now does.
This will make it easier to reuse that logic in later changeset.
Differential Revision: https://phab.mercurial-scm.org/D11131
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 15 Jul 2021 01:58:50 +0200] rev 47688
dirstate-map: factor out the change to _dirs and _alldirs on dropping
This logic is complicated enough to deserves its own function. So it now does.
This will make it easier to reuse that logic in later changeset.
Differential Revision: https://phab.mercurial-scm.org/D11130
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 15 Jul 2021 01:58:31 +0200] rev 47687
dirstate-map: factor out the change to _dirs and _alldirs on adding
This logic is complicated enough to deserves its own function. So it now does.
This will make it easier to reuse that logic in later changeset.
Differential Revision: https://phab.mercurial-scm.org/D11129
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 14 Jul 2021 22:06:13 +0200] rev 47686
dirstate-item: introduce a `dm_otherparent` property
See inline documentation for details.
Differential Revision: https://phab.mercurial-scm.org/D11124
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 14 Jul 2021 21:59:18 +0200] rev 47685
dirstate-item: introduce a `dm_nonnormal` property
See inline documentation for details.
Differential Revision: https://phab.mercurial-scm.org/D11123
Simon Sapin <simon.sapin@octobus.net> [Fri, 16 Jul 2021 16:19:31 +0200] rev 47684
dirstate-v2: Reserve a few bytes of space for future extensions
See doc-comment
Differential Revision: https://phab.mercurial-scm.org/D11101
Simon Sapin <simon.sapin@octobus.net> [Fri, 16 Jul 2021 14:08:26 +0200] rev 47683
dirstate-v2: Separate iterators for dirfoldmap and debugdirstate
`dirstatemap.dirfoldmap` was recently changed to re-use a Rust iterator
that was added for the `hg debugdirstate` command.
That iterator was based on all nodes in the tree dirstate without an entry
only existing to hold child nodes, and therefore being directories.
However to optimize status further we may want to store additional nodes
for unknown or ignored files and directories. At that point the two users
of this iterator will want different things, so let’s make two iterators
instead.
See doc-comments in `dispatch.rs`.
Differential Revision: https://phab.mercurial-scm.org/D11099
Simon Sapin <simon.sapin@octobus.net> [Thu, 15 Jul 2021 23:02:17 +0200] rev 47682
dirstate-v2: Move fixed-size tree metadata into the docket file
Before this changeset, the dirstate-v2 data file contained not only nodes
and paths that may be reused when appending to an existing file,
but also some fixed-size metadata that applies to the entire tree
and was added at the end of the data file for every append.
This moves that metadata into the docket file, so that repeated "append"
operations without meaningful changes don’t actually need to grow any file.
Differential Revision: https://phab.mercurial-scm.org/D11098
Simon Sapin <simon.sapin@octobus.net> [Thu, 08 Jul 2021 19:23:44 +0200] rev 47681
dirstate-v2: Add heuristic for when to create a new data file
… instead of appending to the existing one. This is based on keeping track
of how much of the existing data is not used anymore.
Differential Revision: https://phab.mercurial-scm.org/D11097
Simon Sapin <simon.sapin@octobus.net> [Thu, 15 Jul 2021 10:31:43 +0200] rev 47680
dirstate-v2: Reuse existing paths when appending to a data file
When writing a dirstate in v2 format by appending to an existing data file,
filenames / paths that are borrowed from the previous on-disk representation
can be reused.
Differential Revision: https://phab.mercurial-scm.org/D11096
Simon Sapin <simon.sapin@octobus.net> [Thu, 15 Jul 2021 08:53:03 +0200] rev 47679
dirstate-v2: Reuse existing nodes when appending to a data file
When writing a dirstate in v2 format by appending to an existing data file,
nodes that are still "unparsed" from the previous on-disk representation
have been unchanged and can therefore be reused.
This makes the new data point to previously-existing data for tree nodes.
Differential Revision: https://phab.mercurial-scm.org/D11095
Simon Sapin <simon.sapin@octobus.net> [Tue, 13 Jul 2021 17:18:23 +0200] rev 47678
dirstate-v2: Support appending to the same data file
For now we’re still writing the entire data every time, so appending is not
useful yet. Later we’ll have new nodes pointing to some existing data for
nodes and paths that haven’t changed.
The decision whether to append is pseudo-random in order to make tests exercise
both code paths. This will be replaced by a heuristic based on the amount
of unused existing data.
Differential Revision: https://phab.mercurial-scm.org/D11094
Simon Sapin <simon.sapin@octobus.net> [Tue, 13 Jul 2021 09:44:44 +0200] rev 47677
dirstate-v2: shrink on-disk path lengths to 16-bits
Differential Revision: https://phab.mercurial-scm.org/D11091
Simon Sapin <simon.sapin@octobus.net> [Mon, 12 Jul 2021 23:05:56 +0200] rev 47676
dirstate-v2: Rename Header to Root, move it to the end of the data file
Now that they don’t have to be at the start, a given data file may contain
multiple "roots". A docket only points to one of them, and previous ones
are left unused to allow allow append-only in-place writing to an existing
data file.
Differential Revision: https://phab.mercurial-scm.org/D11090
Simon Sapin <simon.sapin@octobus.net> [Mon, 12 Jul 2021 22:46:52 +0200] rev 47675
dirstate-v2: Enforce data size read from the docket file
The data file may not be shorter than its size given by the docket.
It may be longer, but additional data is ignored.
Differential Revision: https://phab.mercurial-scm.org/D11089
Simon Sapin <simon.sapin@octobus.net> [Thu, 08 Jul 2021 12:18:21 +0200] rev 47674
dirstate-v2: Introduce a docket file
.hg/dirstate now only contains some metadata to point to a separate data file
named .hg/dirstate.{}.d with a random hexadecimal identifier. For now every
update creates a new data file and removes the old one, but later we’ll
(usually) append to an existing file.
Separating into two files allows doing the "write to a temporary file then
atomically rename into destination" dance with only a small docket file,
without always rewriting a lot of data.
Differential Revision: https://phab.mercurial-scm.org/D11088
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 15 Jul 2021 17:24:09 +0200] rev 47673
dirstate: replace a dead conditional branch with an assert in `update_file`
This is a case we never meet, so lets trim it away to simplify the code before
more changes.
Differential Revision: https://phab.mercurial-scm.org/D11128
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 14 Jul 2021 23:32:35 +0200] rev 47672
dirstate-map: do not use `size` to gate copy dropping during remove_file
This get us close to moving the block right above withing the DirstateItem
object. Doing so will help us getting rid of magic constant at the dirstatemap
level.
Differential Revision: https://phab.mercurial-scm.org/D11127
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 15 Jul 2021 00:27:29 +0200] rev 47671
dirstate: drop a duplicated assert
This very case is checked in 3 line above that one.
Differential Revision: https://phab.mercurial-scm.org/D11126
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 15 Jul 2021 00:22:49 +0200] rev 47670
dirstate: drop duplicated check
This is covered by the `@requires_parents_change` decorator that this function use.
Differential Revision: https://phab.mercurial-scm.org/D11125
Pierre-Yves David <pierre-yves.david@octobus.net> [Mon, 19 Jul 2021 06:21:04 +0200] rev 47669
dirstate: add a `set_possibly_dirty` in `fakedirstatewritetime`
Differential Revision: https://phab.mercurial-scm.org/D11122
Pierre-Yves David <pierre-yves.david@octobus.net> [Mon, 19 Jul 2021 06:20:00 +0200] rev 47668
dirstatemap: use `set_possibly_dirty` in `clearambiguoustimes`
lets put this new method to use
Differential Revision: https://phab.mercurial-scm.org/D11121
Pierre-Yves David <pierre-yves.david@octobus.net> [Mon, 19 Jul 2021 06:30:04 +0200] rev 47667
dirstate-item: use `set_possibly_dirty` in `pure.pack_dirstate`
Lets make use of the new function.
Differential Revision: https://phab.mercurial-scm.org/D11120
Pierre-Yves David <pierre-yves.david@octobus.net> [Mon, 19 Jul 2021 06:29:30 +0200] rev 47666
dirstate-item: add a `set_possibly_dirty` method
See inline documentation for details.
The pushes the AMBIGUOUS_TIME implementation further down the line within the
DirstateItem only. When this cleanup is done we will be able to stop using this
representation internally.
Differential Revision: https://phab.mercurial-scm.org/D11119
Pierre-Yves David <pierre-yves.david@octobus.net> [Tue, 13 Jul 2021 13:06:50 +0200] rev 47665
dirstate-item: add a `from_v1_data` constructor
This class method is dedicated to building a DirstateItem from the data
available in the "dirstate-v1" format. Since that format is frozen, this
constructor will never change (unlike the `__init__` one).
Differential Revision: https://phab.mercurial-scm.org/D11118