Raphaël Gomès <rgomes@octobus.net> [Fri, 25 Sep 2020 17:51:34 +0200] rev 45562
rust: add `dirstate_tree` module
Mercurial needs to represent the filesystem hierarchy on which it operates, for
example in the dirstate. Its current on-disk representation is an unsorted, flat
structure that gets transformed in the current Rust code into a `HashMap`.
This loses the hierarchical information of the dirstate, leading to some
unfortunate performance and algorithmic compromises.
This module adds an implementation of a radix tree that is specialized for
representing the dirstate: its unit is the path component. I have made no
efforts to optimize either its memory footprint or its insertion speed: they're
pretty bad for now.
Following will be a few patches that modify the dirstate.status logic to use
that new hierarchical information, fixing issue 6335 in the same swing.
Differential Revision: https://phab.mercurial-scm.org/D9085
Pulkit Goyal <7895pulkit@gmail.com> [Fri, 02 Oct 2020 19:19:01 +0530] rev 45561
Added signature for changeset 07731064ac41
Pulkit Goyal <7895pulkit@gmail.com> [Fri, 02 Oct 2020 19:18:54 +0530] rev 45560
Added tag 5.5.2 for changeset 07731064ac41
Ludovic Chabant <ludovic@chabant.com> [Tue, 29 Sep 2020 14:38:46 -0700] rev 45559
hgdemandimport: exclude more sqlalchemy modules
We could potentially exclude the entire sqlalchemy library.
Raphaël Gomès <rgomes@octobus.net> [Fri, 24 Jul 2020 16:35:02 +0200] rev 45558
rust-dirstatemap: add #[timed] to dirstatemap read for comparison
Differential Revision: https://phab.mercurial-scm.org/D9084
Martin von Zweigbergk <martinvonz@google.com> [Mon, 21 Sep 2020 11:12:58 -0700] rev 45557
merge: make low-level update() private (API)
We have very few callers left that call the low-level `merge.update()`
function. I think it's time to make it private. I'll remove the
remaining callers in coming patches, except for one call from the
`rebase` module. I hope to eventually fix that too, but it's more
complex because it requires teaching `merge.graft()` to work with a
dirty working copy.
Differential Revision: https://phab.mercurial-scm.org/D9065
Martin von Zweigbergk <martinvonz@google.com> [Mon, 21 Sep 2020 10:09:39 -0700] rev 45556
merge: use merge.clean_update() when applicable
We have had this higher-level function (higher than `merge.update()`,
that is) for a while. Let's simply some callers by using it. I don't
know why I didn't do this when I introduced the function.
After this patch, there are no remaining callers that call
`hg.updaterepo()` with `overwrite=True`. We'll clean that up soon.
Differential Revision: https://phab.mercurial-scm.org/D9063
Martin von Zweigbergk <martinvonz@google.com> [Fri, 18 Sep 2020 15:03:06 -0700] rev 45555
rebase: teach in-memory rebase to not restart with on-disk rebase on conflict
When in-memory rebase runs into conflicts, it redoes the whole rebase
operation. This patch teaches it to instead discard just the current
`overlayworkingctx` and redo that node on disk.
I've tested this by enabling in-memory rebase by default and checking
that there are no unexpected differences after this patch.
The next step is to make it so that `hg rebase --continue` can use
in-memory merge.
Differential Revision: https://phab.mercurial-scm.org/D9076
Raphaël Gomès <rgomes@octobus.net> [Thu, 24 Sep 2020 16:30:17 +0200] rev 45554
rust: update Cargo.lock
Differential Revision: https://phab.mercurial-scm.org/D9080
Raphaël Gomès <rgomes@octobus.net> [Thu, 24 Sep 2020 16:30:00 +0200] rev 45553
rust-tests: add test to check if `Cargo.lock` is up-to-date
An out-of-date `Cargo.lock` is the source of rebase conflicts, prevents a
reproductible build, introduces desync between the series that introduces a
change in dependencies and the one that eventually commits the `Cargo.lock`,
and is just a general annoyance.
This commit demonstrates that the test works, the next one fixes the
currently out-of-date `Cargo.lock`.
Differential Revision: https://phab.mercurial-scm.org/D9079