view relnotes/5.6 @ 46067:cc0b332ab9fc

run-tests: stuff a `python3.exe` into the test bin directory on Windows

Windows doesn't have `python3.exe` as part of the python.org distribution, and
that broke every script with a shebang after c102b704edb5.  Windows itself
provides a `python3.exe` app execution alias[1], but it is some sort of reparse
point that MSYS is incapable of handling[2].  When run by MSYS, it simply prints

    $ python3 -V
     - Cannot open

That in turn caused every `hghave` check, and test that invokes shebang scripts
directly, to fail.  Rather than try to patch up every script call to be invoked
with `$PYTHON` (and regress when non Windows developers forget), copying the
executable into the test binary directory with the new name just works.  Since
this directory is prepended to the system PATH value, it also overrides the
broken execution alias.  (The `_tmpbindir` is used instead of `_bindir` because
the latter causes python3.exe to be copied into the repo next to hg.exe when
`test-run-tests.t` runs.  Something runs with this version of the executable and
subsequent runs of `run-tests.py` inside `test-run-tests.t` try to copy over it
while it is in use, and fail.  This avoids the failures and the clutter.)

I didn't conditionalize this on py3 because `python3.exe` needs to be present
(for the shebangs) even when running py2 tests.  It shouldn't matter to these
simple scripts, and I think the intention is to make the test runner use py3
always, even if testing a py2 build.  For now, still supporting py2 is helping
to clean up the mess that is py3 tests.

[1] https://stackoverflow.com/a/57168165
[2] https://stackoverflow.com/questions/59148628/solved-unable-to-run-python-3-7-on-windows-10-permission-denied#comment104524397_59148666

Differential Revision: https://phab.mercurial-scm.org/D9543

== New Features ==

 * `hg mv -A` can now be used with `--at-rev`. It behaves just like
   `hg cp -A --at-rev`, i.e. it marks the destination as a copy of the
   source whether or not the source still exists (but the source must
   exist in the parent revision).

 * New revset predicate `diffcontains(pattern)` for filtering revisions
   in the same way as `hg grep --diff pattern`.

 * The memory footprint per changeset and per file during pull/unbundle
   operations has been significantly reduced.


== New Experimental Features ==



== Bug Fixes ==



== Backwards Compatibility Changes ==



== Internal API Changes ==

 * `merge.update()` is now private (renamed to `_update()`). Hopefully
   the higher-level functions available in the same module cover your
   use cases.

 * `phases.registernew` now takes a set of revisions instead of a list
   of nodes. `phases.advanceboundary` takes an optional set of revisions
   in addition to the list of nodes. The corresponeding members of the
   `phasecache` class follow this change.

 * The `addgroup` member of `revlog` classes no longer keeps a list of
   all found nodes. It now returns True iff a node was found in the group.
   An optional callback for duplicated nodes can be used by callers to keep
   track of all nodes themselve.

 * The `_chaininfocache` of `revlog` classes has been changed from a dict
   to a LRU cache.