mercurial: Changelog

largefiles: drop "rel" prefix from filename variables The prefixes were meant to indicate that these paths are repo-relative as opposed to absolute. However, that's what the majority of paths in our code base are, so "rel" made me think they were instead cwd-relative. Let's just drop the prefixes. Differential Revision: https://phab.mercurial-scm.org/D5847

branchmap: add some clarifications and clean up flow - Remove indentation where it is not needed. - Swap the subset test branches to follow along logically and put the 'empty' case last. Differential Revision: https://phab.mercurial-scm.org/D5637

branchmap: updating triggers a write Rather than separate updating and writing, create a subclass that doesn't write on update. This minimises chances we forget to write out updates somewhere. This also makes refactoring and improving the branchmap functionality easier. Differential Revision: https://phab.mercurial-scm.org/D5636

branchmap: make branchcache responsible for reading Encapsulate reading in a classmethod, to make it clear what kind of object is being handled. This is part of a stack of refactoring changes to help performance improvements down the line. Differential Revision: https://phab.mercurial-scm.org/D5635

attr: make some docstrings raw strings This avoids a SyntaxWarning in Python 3.8 due to invalid \ escapes. Differential Revision: https://phab.mercurial-scm.org/D5817

revset: leverage getintrange() helper in relation-subscript operation (API) Now a range expression is parsed by a relation function itself since the upper layer have no knowledge about the default first/last bounds.

revset: allow to parse single integer as a range Even though this wouldn't be any useful in followlines(), it helps to unify range processing.

revset: extract a helper to parse integer range It's getting common. As a first step, this patch adds getintrange() and makes followlines() use it. I wanted to unify the error messages to make the function interface simple, but I failed to phrase it briefly.

commit: if interactive, look elsewhere for whitespace settings (BC) Previously, when doing `commit -i`, we respected `diff.ignorews` and other whitespace-related settings, which is probably unexpected. The primary reason for this is to support hgext.record's commandline options, it's probably accidental that the `[diff]` settings were also considered. See comments on issue6042 and D5490. This can cause problems (issue5839, issue6042). It is assumed by the author that the `[diff]` section is primarily for *viewing* diffs, and that it is unlikely what people intend when attempting to commit or revert. With this change, if a user wants the behavior, they can clone their `[diff]` settings to `commands.commit.interactive.<setting>`. This is thus a mild BC change, but one I suspect is not going to be relied on by anyone. Note: while doing a partial commit/revert, we do not know what command the user is actually running. This means that the split extension, which ends up calling into this code, will respect the `commands.commit.interactive.<setting>` settings, and not a hypothetical `commands.split.interactive.<setting>`. This *also* means that setting `commands.commit.interactive.ignoreblanklines`, for example, will still cause issue5839. Considering the highly unlikely chance that a user actually sets `commands.commit.interactive.<setting>`, the author deems this risk acceptable. Differential Revision: https://phab.mercurial-scm.org/D5834