* lisp/emacs-lisp/bytecomp.el (byte-compile-docstring-style-warn):
Make the code slightly cheaper. It's still dominated by the
huge cost of `byte-compile--wide-docstring-p`, though.
* test/lisp/progmodes/cperl-mode-tests.el: Add a hint to the
commentary that the tests are intended to run on older Emacsen.
(cperl-test-bug-10483): Replace skip-when by skip-unless followed
by a negation.
(cperl-test-bug-37127): Replace skip-when by skip-unless followed
by a negation.
c-forward-sws and c-backward-sws were scanning over spaces and
linefeeds marked with the string-fence syntax-table text
property. Fix this by (i) removing the WS text properties
c-in-sws and c-is-sws from characters when setting the
string-fence on them; (ii) checking the syntax of "space"
characters when scanning over them.
* lisp/progmodes/cc-defs.el (c-skip-ws-chars-forward)
c-skip-ws-chars-backward, c-put-string-fence): New macros.
* lisp/progmodes/cc-awk.el
(c-awk-set-string-regexp-syntax-table-properties): Use
c-put-string-fence.
* lisp/progmodes/cc-engine.el (c-beginning-of-statement-1):
Correct the determination of macro-start.
(c-forward-sws, c-backward-sws): Replace skip-chars-forward by
c-skip-ws-chars-forward and skip-chars-backward by
c-skip-ws-chars-backward.
(c-unmark-<>-around-region, c-after-change-unmark-ml-strings)
(c-propertize-ml-string-opener): Use c-put-string-fence.
* lisp/progmodes/cc-mode.el (c-put-syn-tab): Use
c-put-string-fence when appropriate.
Another case that was too optimistic. Better use \> or \< rather
than \b if you want your regexp to be handled efficiently.
* src/regex-emacs.c (mutually_exclusive_aux) <wordbound>: Cancel optimization.
* test/src/regex-emacs-tests.el (regexp-tests-backtrack-optimization):
New test.
* lisp/ls-lisp.el (ls-lisp-verbosity): Add `modes` on GNU & Unix
(ls-lisp-format): When `modes` is not in `ls-lisp-verbosity',
keep just the 1st character of `drwxrwxrwx`.
* doc/emacs/android.texi (Android Windowing): Document
`android-keyboard-bell-duration'.
* java/org/gnu/emacs/EmacsService.java (ringBell): New argument
DURATION.
* src/android.c (android_init_emacs_service): Adjust
correspondingly.
(android_bell): Provide the duration of the vibration.
* src/androidfns.c (syms_of_androidfns)
<android_keyboard_bell_duration>: New variable.
rather than `buffer-read-only`. Applied throughout the file.
(todo-date-pattern, todo-edit-item--header, todo-convert-legacy-date-time)
(todo-read-date): Remove redundant "" arg to `mapconcat`.
* lisp/dired-aux.el (dired-do-replace-regexp-as-diff): New command.
* lisp/misearch.el (multi-file-diff-unsaved): New user option.
(multi-file-replace-as-diff): New function.
(multi-file-replace-regexp-as-diff): New command.
(replace-regexp-as-diff): New command.
(multi-file-diff-no-select): New function.
The optimization was incorrect in a particular corner case.
In this fix I just disable it conservatively for more cases because
it's not obvious how to fix it while preserving the "good" cases.
We may find a better fix by using an approach like the one
in `mutually_exhaustive_aux`, but for now this is good enough,
especially since \{..\} repetitions are not used very frequently.
* test/src/regex-resources/PTESTS: New test.
The treesit-language-at-point function is only suppose to query the
host language.
* lisp/progmodes/elixir-ts-mode.el
(elixir-ts--indent-rules): Add missing rules.
(elixir-ts--treesit-language-at-point): Update function to only query
the host language.
* test/lisp/progmodes/elixir-ts-mode-resources/indent.erts: Add test
for inline docs.
This lets us simplify the logic for how we reset
'eshell-current-command' and 'eshell-last-async-procs', as well as
improving correctness of Eshell command forms in a few esoteric
scenarios. Additionally, this helps set the stage for better support
of background commands in Eshell (bug#66164).
* lisp/eshell/esh-cmd.el (eshell-cmd-initialize): Remove addition to
'eshell-post-command-hook'; this is handled in 'eshell-resume-command'
and 'eshell-resume-eval' now.
(eshell-resume-command): Handle resetting the prompt as needed.
(eshell-resume-eval): Use 'unwind-protect' to ensure that we set
'eshell-last-async-procs' and 'eshell-current-comment' at the right
times.
(eshell-parse-command, eshell-trap-errors, eshell-manipulate): Use
'unwind-protect'.
(eshell-do-eval): Allow 'eshell-defer' to pass through
'unwind-protect' forms without actually calling the unwinding forms
(yet).
* lisp/eshell/esh-proc.el (eshell-kill-process-function)
(eshell-reset-after-proc): Make obsolete. The behavior is now handled
in 'eshell-resume-command'.
(eshell-gather-process-output, eshell-sentinel)
(eshell-interrupt-process, eshell-kill-process, eshell-quit-process)
(eshell-stop-process, eshell-continue-process): Run 'eshell-kill-hook'
directly.
* test/lisp/eshell/esh-cmd-tests.el (esh-cmd-test/throw): New test.
* src/regex-emacs.c (mutually_exclusive_p): Don't pretend that pattern
position 0 has been checked already.
* test/src/regex-emacs-tests.el (regexp-tests-backtrack-optimization):
Add a corresponding regression test plus some other related tests
I had around.
* lisp/vc/vc-annotate.el (vc-annotate-lines): Check whether a symbol
is a face, not just its existence. Previously we would sometimes end
up using non-faces resulting in complaints from the display system.
* lisp/vc/ediff-util.el (ediff-toggle-multiframe): Remove duplicate
cond clause. Fix parentheses so that trailing 'message' is
contained in fallback clause. Fix indentation.
* src/xterm.c (x_draw_fringe_bitmap): Save clip rectangle from
x_clip_to_row, and draw only the intersection between it and the
fringe bitmap, for if the bitmap is overlaid, the clip mask will
override the clip rectangle.
(x_clip_to_row): New argument *RECT_RETURN. All callers
changed.
Instead of scanning vector_free_lists from the appropriate size until
we find a nonempty bucket, start at the last bucket where we last put
something in. This may favour splitting larger vectors than necessary
but in general saves a lot of time in the allocation of small vectors.
Original patch by Ihor Radchenko.
* src/alloc.c (last_inserted_vector_free_idx): New variable.
(setup_on_free_list): Set it.
(allocate_vector_from_block): Use it.
(sweep_vectors): Reset it.
The latter half of vector_free_lists was never used in any meaningful
way but it did require traversal during allocation and GC. Reduce it
to sizes we actually allocate, with a bucket for bigger ones.
* src/alloc.c (VECTOR_MAX_FREE_LIST_INDEX): Rename to...
(VECTOR_FREE_LIST_ARRAY_SIZE): ... this and adjust its value.
(vector_free_lists): Use new, smaller size.
(setup_on_free_list, allocate_vector_from_block):
Adapt to new vector_free_lists size.
(pseudovector_nbytes): New function extracted from...
(vectorlike_nbytes): ...here.
* lisp/emacs-lisp/macroexp.el (macroexp-parse-body):
Return an empty body even when there are declarations present.
Previously, the last declaration was considered part of the body,
which is only correct if the input consists of a single string.
Reported by Jens Schmidt.
* doc/lispref/os.texi (Desktop Notifications): Revise
documentation for android-notifications-notify to reflect
changes.
* java/org/gnu/emacs/EmacsDesktopNotification.java (display1):
Convert notification importance to a legacy priority between
Android 7.1 and 4.1.
* java/org/gnu/emacs/EmacsPixmap.java (EmacsPixmap): Remove
immutable bitmap constructor, as the underlying Android API
functions are erroneously implemented.
* src/android.c (android_init_emacs_pixmap): Cease searching for
deleted constructor.
(android_create_pixmap_from_bitmap_data): Create a pixmap, then
fill it with the contents of the bitmap, in lieu of employing
the aforementioned constructor.
* src/androidselect.c (Fandroid_notifications_notify): Revise
doc string.
Flymake's normal behaviour implies catching the errors of misbehaving
backends. This behavior is tested by Flymake's automated tests, built
on top of ERT, which means that debug-on-error is always t in the
bodies of said tests (I don't know the rationale for this, but it's
been like this for some time)
Flymake used to shun usage of 'condition-case-unless-debug' because of
this. But since that macro is pretty useful, I started using it
again, and as a consequence, tests started breaking.
The solution is to:
1. stop requiring the deprecated legacy backend 'flymake-proc' (which
by design, always errors, except in rare circumstances).
2. deliberately set debug-on-error to nil around the "dummy backends""
test.
* lisp/progmodes/flymake.el (flymake-proc): Don't require it by
default.
* test/lisp/progmodes/flymake-tests.el:
(dummy-backends): Make robust to ert's debug-on-error setting.
See Bug#66106
* lisp/gnus/gnus-msg.el (gnus-summary-reply): Not erase-buffer. The
intention was to narrow the buffer to the headers, and then delete
just those headers. But erase-buffer doesn't respect narrowing, so
use (delete-region (point-min) (point-max)) instead.
