The `activate` script calls `hash -r` in two places to make sure the shell picks
up the environment changes the script makes. Before that, it checks to
see if the shell running the script is bash or zsh.
`hash -r` is specified by POSIX and is not exclusive to bash and zsh.
This guard prevents the script from calling `hash -r` in other
`#!/bin/sh`-compatible shells like dash.
Anthony Shaw [Tue, 28 Nov 2023 01:58:53 +0000 (12:58 +1100)]
gh-112292 : Catch import error conditions with readline hooks (gh-112313)
Prevents a segmentation fault in registered hooks for the readline library, but only when the readline module is loaded inside an isolated sub interpreter. The module is single-phase init so loading it fails, but not until the module init function has already run, where the readline hooks get registered.
The readlinestate_global macro was error-prone to PyImport_FindModule returning NULL and crashing in about 18 places. I could reproduce 1 easily, but this PR replaces the macro with a function and adds error conditions to the other functions.
Jacob Walls [Mon, 27 Nov 2023 18:05:55 +0000 (13:05 -0500)]
gh-76912: Raise OSError from any failure in getpass.getuser() (#29739)
* bpo-32731: Raise OSError from any failure in getpass.getuser()
Previously, if the username was not set in certain environment variables, ImportError escaped on Windows systems, and it was possible for KeyError to escape on other systems if getpwuid() failed.
Alex Waygood [Mon, 27 Nov 2023 16:34:44 +0000 (16:34 +0000)]
gh-111874: Call `__set_name__` on objects that define the method inside a `typing.NamedTuple` class dictionary as part of the creation of that class (#111876)
Mark Dickinson [Sun, 26 Nov 2023 14:29:52 +0000 (14:29 +0000)]
gh-112358: Fix Python 3.12 regression with subclassing struct.Struct. (#112424)
Revert commit c8c0afc7137ab9f22bf59d591084948ca967c97c (PR #94532),
which moved `struct.Struct` initialisation from `Struct.__init__` to `Struct.__new__`.
This caused issues with code in the wild that subclasses `struct.Struct`.
Barney Gale [Sat, 25 Nov 2023 17:19:38 +0000 (17:19 +0000)]
GH-112361: Speed up pathlib by removing some temporary objects. (#112362)
Construct only one new list object (using `list.copy()`) when creating a
new path object with a modified tail. This slightly speeds up
`with_name()` and `with_suffix()`
Ned Deily [Wed, 22 Nov 2023 08:29:28 +0000 (03:29 -0500)]
Do not build the macOS installer with mimalloc enabled (#112312)
Do not build the macOS installer with mimalloc enabled pending resolution of
weak linking crashes during interpreter startup on macOS 10.9,
10.10, and 10.11 when built on macOS 11 and later.
gh-110950: add upstream Tk fixes to macOS installer. (GH-111041)
Add upstream Tk patches for three problems affecting tkinter users:
- Update macOS installer to include a fix accepted by upstream Tcl/Tk
for a crash encountered after the first :meth:`tkinter.Tk` instance
is destroyed. (gh-92603)
- Update macOS installer to include an upstream Tcl/Tk fix
for the ``ttk::ThemeChanged`` error encountered in Tkinter. (gh-71383)
- Update macOS installer to include an upstream Tcl/Tk fix for the
``Secure coding is not enabled for restorable state!`` warning
encountered in Tkinter on macOS 14 Sonoma. (gh-110950)
Guido van Rossum [Mon, 20 Nov 2023 19:25:32 +0000 (11:25 -0800)]
gh-112287: Speed up Tier 2 (uop) interpreter a little (#112286)
This makes the Tier 2 interpreter a little faster.
I calculated by about 3%,
though I hesitate to claim an exact number.
This starts by doubling the trace size limit (to 512),
making it more likely that loops fit in a trace.
The rest of the approach is to only load
`oparg` and `operand` in cases that use them.
The code generator know when these are used.
For `oparg`, it will conditionally emit
```
oparg = CURRENT_OPARG();
```
at the top of the case block.
(The `oparg` variable may be referenced multiple times
by the instructions code block, so it must be in a variable.)
For `operand`, it will use `CURRENT_OPERAND()` directly
instead of referencing the `operand` variable,
which no longer exists.
(There is only one place where this will be used.)
Guido van Rossum [Mon, 20 Nov 2023 18:08:53 +0000 (10:08 -0800)]
gh-106529: Make FOR_ITER a viable uop (#112134)
This uses the new mechanism whereby certain uops
are replaced by others during translation,
using the `_PyUop_Replacements` table.
We further special-case the `_FOR_ITER_TIER_TWO` uop
to update the deoptimization target to point
just past the corresponding `END_FOR` opcode.
Terry Jan Reedy [Sun, 19 Nov 2023 06:39:26 +0000 (01:39 -0500)]
IDLE: Fix test_debugger bug and buildbot failures (#112258)
Missing "requires('gui')" causes Tk() to fail when no gui.
This caused CI Hypothesis test to fail, but I did not understand
the its error message. Then buildbots failed.
IdbTest failed on draft Bdb replacement because so different.
Simplified version works on old and new.
Anthony Shaw [Sun, 19 Nov 2023 04:20:38 +0000 (15:20 +1100)]
gh-79871: IDLE - Fix and test debugger module (#11451)
Add docstrings to the debugger module. Fix two bugs: initialize Idb.botframe (should be in Bdb); In Idb.in_rpc_code, check whether prev_frame is None before trying to use it. Make other code changes.
Expand test_debugger coverage from 19% to 66%.
--------- Co-authored-by: Terry Jan Reedy <tjreedy@udel.edu>
Barney Gale [Sat, 18 Nov 2023 17:06:10 +0000 (17:06 +0000)]
GH-110109: Test pure functionality of `pathlib.Path` user subclasses (#112242)
Add `PurePathTest` as a superclass of `PathTest`, and therefore also
`PathSubclassTest`. This adds coverage of pure functionality in user
subclasses of `pathlib.Path`.
Remove `PosixPathAsPureTest` and `WindowsPathAsPureTest`, as they
now duplicate `PosixPathTest` and `WindowsPathTest`.
This makes the MROs of test unit classes match the MROs of pathlib
classes.