Previously, 'quit' could cause remote endpoints to never get a
reply and thus sometimes hang. Ensure we always reply. Also,
give the application a chance to signal jsonrpc-error with the
served code=32000, meaning "no error".
* doc/lispref/text.texi (JSONRPC Overview): Rework section on
request dispatchers.
* lisp/jsonrpc.el (jsonrpc-connection-receive): Rework.
The info returned from `garbage-collect` is really handy to
track the evolution of the heap size, but sadly it's available only
at the cost of running a full GC, which has two big downsides:
it's slow, it affects what we're measuring, and it can't be used
in `post-gc-hook`.
So, this patch makes it available without running the GC.
* src/alloc.c (Fgarbage_collect_heapsize): New function, extracted from
`Fgarbage_collect`.
(Fgarbage_collect): Use it.
(syms_of_alloc): defsubr it.
* doc/lispref/internals.texi (Garbage Collection): Extract
documentation for it from that of `garbage-collect`.
* doc/lispref/internals.texi (Module Values, C Integer Types):
Mention that the example assumes Emacs was built with the
GMP library, not with mini-gmp. Mention stdckdint.h for
integer overflow checking, and mention Emacs integers
for values outside machine range.
Previously, if 'calc-string-maximum-character' wasn't a valid
character 'math-vector-is-string' would throw an error in the
comparison, leading to an incomplete display of the stack and a
cryptic error message. Instead, have 'math-vector-is-string'
return nil, which effectively disables the display of strings.
Refines feature introduced in bug#78528.
* doc/misc/calc.texi (Customizing Calc): Update description of
behavior for invalid 'calc-string-maximum-character'.
* lisp/calc/calccomp.el (math-vector-is-string): Return nil when
'calc-string-maximum-character' doesn't represent a character.
* test/lisp/calc/calc-tests.el (calc-math-vector-is-string): Correct
and simplify tests.
* doc/misc/dbus.texi (Top): Add "Inhibitor Locks" submenu.
Remove trailing period from chapter and section titles.
(Inhibitor Locks): New node.
* etc/NEWS: New D-Bus functions to support systemd inhibitor locks.
Presentational fixes and improvements.
* src/dbusbind.c (xd_registered_inhibitor_locks): New variable.
(Fdbus_make_inhibitor_lock, Fdbus_close_inhibitor_lock)
(Fdbus_registered_inhibitor_locks): New DEFUNs. (Bug#79963)
(syms_of_dbusbind_for_pdumper): Initialize
`xd_registered_inhibitor_locks'.
(syms_of_dbusbind): Declare subroutines
`Sdbus_make_inhibitor_lock', `Sdbus_close_inhibitor_lock' and
`Sdbus_registered_inhibitor_locks'. Declare symbol `Qdbus_call_method'.
staticpro `xd_registered_inhibitor_locks'.
* test/lisp/net/dbus-tests.el (dbus--test-systemd-service)
(dbus--test-systemd-path, dbus--test-systemd-manager-interface):
New defconsts.
(dbus-test10-inhibitor-locks): New test.
* doc/lispref/display.texi (Customizing Bitmaps): Fix grammar.
(XBM Images): Fix reference to incorrect number of properties.
(Multi-Frame Images, Other Image Types): Mention multi-frame WebP
support.
* src/image.c: Don't include pdumper.h; it hasn't been needed since
commit of 2019-05-17 "Clean up and simplify image-type setup". Fix
some commentary.
(image_pix_container_create_from_bitmap_data): Remove no longer used
frame parameter (bug#80191); this continues from commit of
2022-11-25 "Remove unused parameter from
image_create_pix_container". All callers updated.
(x_create_xrender_picture, initialize_image_type): Simplify.
(xbm_image_p): Remove redundant conditional branch. Update
commentary.
* etc/NEWS:
* doc/emacs/maintaining.texi (Projects):
* lisp/progmodes/project.el (project-mode-line): Update
documentation due to changes in bug#78545.
Add an optional 'context' argument to 'make-progress-reporter'
which 'progress-reporter-echo-area' consults to inhibit updates
if the context is 'async' and the echo area is busy.
* lisp/subr.el (make-progress-reporter): Add the optional
'context' argument.
(progress-reporter-context): New defun accessor.
(progress-reporter-echo-area): Consult
'progress-reporter-context'.
* doc/lispref/display.texi: Document context.
* etc/NEWS: Announce context.
* lisp/textmodes/ispell.el (ispell-save-corrections-as-abbrevs):
New user option.
(ispell--abbrev-saving-allowed)
(ispell--save-correction-as-abbrev): New variables.
(ispell--maybe-save-correction-abbrev): New function.
(ispell-word, ispell-process-line): Use them to save corrections
as abbrevs when appropriate (bug#79985).
(ispell-command-loop): Add C-u as command character to toggle
abbrev saving for an immediately following replacement command.
(ispell-help): Document C-u binding.
* doc/emacs/fixit.texi (Spelling): Document new feature.
* doc/misc/dbus.texi (Signals): All registered signal handlers are
called.
(Synchronous Methods, Signals, Monitoring Messages): Add function
result in examples.
* src/dbusbind.c (xd_store_event): New function.
(xd_read_message_1): Use it. Call all registered handlers per
signal. (Bug#80168)
* test/lisp/net/dbus-tests.el (dbus--test-signal-handler): Adapt defun.
(dbus--test-signal-handler1, dbus--test-signal-handler2): New defuns.
(dbus-test05-register-signal-several-handlers): New test.
(dbus-test04-register-method)
(dbus-test04-call-method-authorizable)
(dbus-test05-register-signal)
(dbus-test05-register-signal-with-nils)
(dbus-test06-register-property-emits-signal): Adapt tests.
A unique frame id is assigned to a new or cloned frame, and
reused on an undeleted frame.
The id facilitates unambiguous identification among frames that
share identical names or titles, deleted frames where a live
frame object no longer exists that we can resurrect by id, for
example via 'tab-bar-undo-close-tab'. It also aids debugging at
the C level using the frame struct member id.
Rewrite 'clone-frame' and 'undelete-frame' to not let bind
variables that 'make-frame' uses to avoid conflicts with nested
'make-frame' calls, for example via
'after-make-frame-functions'.
* lisp/frame.el (clone-frame, undelete-frame): Use
'frame--purify-parameters' to supply parameters explicitly.
(undelete-frame--save-deleted-frame): Save frame id for
restoration.
(undelete-frame): Restore frame id.
(frame--purify-parameters): New defun.
(make-frame): Assign a new id for a new or cloned frame, reuse
for undeleted frame.
* src/frame.h (struct frame): Add id member.
(frame_next_id): New extern.
* src/frame.c (frame_next_id): New global counter.
(frame_set_id, frame_set_id_from_params): New function.
(Fframe_id): New DEFUN.
(syms_of_frame <Sframe_id>): New defsubr.
(syms_of_frame <Qinternal_id>): New DEFSYM.
(syms_of_frame <frame_internal_parameters>): Add 'Qinternal_id'.
* src/androidfns.c (Fx_create_frame):
* src/haikufns.c (Fx_create_frame):
* src/nsfns.m (Fx_create_frame):
* src/pgtkfns.c (Fx_create_frame):
* src/w32fns.c (Fx_create_frame):
* src/xfns.c (Fx_create_frame): Call 'frame_set_id_from_params'.
* doc/lispref/frames.texi: Add documentation.
* etc/NEWS: Announce frame id.
This documents how to use LSP multiplexer programs like Rassumfrassum
to connect multiple language servers to a single buffer.
* doc/misc/eglot.texi (Top): Add "Multi-server support" menu entry.
(Multi-server support): New chapter.
(Using Rassumfrassum, Design rationale): New sections documenting
how to use the Rassumfrassum multiplexer program with Eglot, with
practical examples for C++, Python, and multi-language files.
(Performance): Mention Rassumfrassum as solution for JSONRPC traffic
performance issues.
(Reporting bugs): Add guidance for troubleshooting multiplexer-related
bugs. Improve project description guidance. Fix various typos.
* lisp/progmodes/eglot.el (eglot-server-programs): Add a couple
of rass entries.
* etc/EGLOT-NEWS: Announce support for LSP server multiplexers via
Rassumfrassum.
* lisp/term/xterm.el (xterm-update-cursor): New user option.
(xterm--init): Use it.
(xterm--post-command-hook): New function for all xterm
functionality installed in 'post-command-hook'.
(xterm--init-frame-title): Install it.
(xterm--init-update-cursor, xterm--set-cursor-type)
(xterm--update-cursor-type, xterm--update-cursor-color): New
functions.
(xterm--cursor-type-to-int): New constant.
* doc/emacs/display.texi (Cursor Display):
* etc/NEWS: Document the new feature.
* lisp/emacs-lisp/package.el (package-review-policy)
(package-review-directory, package-review-diff-command): Add
new options.
(package--review-p): Add new function to consult
'package-review-policy'.
(package-review): Add new function.
(package-unpack): Use new functions.
(package-install-from-archive): Return package descriptors of
installed packages.
(package-download-transaction): Handle failure of a incomplete
transaction.
(package-install): Report if a package installation failed.
(package-upgrade): Anticipate a failed package transaction by
not deleting a package beforehand.
(package-install-from-buffer): Handle the failure to download
dependencies or a rejection during the actual package review.
* doc/emacs/package.texi: Document feature.
* etc/NEWS: Mention new feature.
The only alternative I've found is
M-: ( i n f o " ( e g l o t ) " ) RET
which isn't really very user friendly.
* lisp/progmodes/eglot.el (eglot-manual): Unobsolete.
* doc/misc/eglot.texi (Top): Mention M-x eglot-manual.
* etc/EGLOT-NEWS: Mention unobsoletion.
* doc/misc/tramp.texi (External methods): Mention, how external
methods are used for multi-hops.
(Ad-hoc multi-hops): Change requirement of method.
* doc/misc/trampver.texi:
* lisp/net/trampver.el: Change version to "2.8.2-pre".
* etc/NEWS: External Tramp methods can be used in multi-hops.
Presentational fixes and improvements.
* lisp/net/tramp-sh.el (tramp-sh-handle-copy-directory): Check for
`tramp-method-out-of-band-p' explicitly. Don't flush directory
properties.
(tramp-method-out-of-band-p): There shouldn't be a multi-hop.
* lisp/net/tramp-smb.el (tramp-smb-handle-copy-directory): Don't
flush directory properties.
* lisp/net/tramp.el (tramp-methods): Adapt docstring.
(tramp-barf-if-file-missing)
(with-parsed-tramp-file-name): Adapt debug spec.
(tramp-skeleton-copy-directory): Flush directory properties.
(tramp-add-hops): Check for `tramp-login-args' property.
* doc/emacs/files.texi: (Time Stamps): Expand the examples and add
node-end summaries. Thanks to Ruth for her help with the clarity of
the documentation.
* doc/lispref/positions.texi (Point): Define "buffer point".
* doc/lispref/windows.texi (Window Point): Say that buffer point
can change in unpredictable ways when windows get selected
(Bug#79927).
* src/window.c (Fwindow_cursor_info): New function.
(syms_of_window): Defsubr it.
* doc/lispref/windows.texi (Window Point): Document it.
* etc/NEWS: Announce new function. (Bug#80023)
* lisp/frame.el (clone-frame): Set the frame parameter
'cloned-from'.
(undelete-frame): Set the frame parameter 'undeleted'.
* src/frame.c (syms_of_frame): <Qcloned_from> <Qundeleted>: New
DEFSYM.
(syms_of_frame): Add 'Qcloned_from' and 'Qundeleted' to
'frame_internal_parameters'.
* doc/lispref/frames.texi: Document these frame parameters.
* etc/NEWS: Announce the new frame parameters.
* lisp/tab-bar.el (tab-bar--reusable-frames): Implement support for the
`reusable-frames' value 0, for better consistency with other buffer
display actions. (bug#80092)
* doc/lispref/windows.texi: Document the display actions
`display-buffer-in-tab' and `display-buffer-in-new-tab' and the
associated alist entries `tab-name' and `tab-group'.
* lisp/tab-bar.el (display-buffer-in-tab): Restructure docstring for
clarity. Describe the `reusable-frames' behaviour first (as if it finds
a frame with a suitable tab then the `tab-name' entry is not used).
(display-buffer-in-new-tab): Use consistent wording.
* lisp/window.el (display-buffer--action-function-custom-type): Add the
display actions as customize choices for user options.
(display-buffer): Document the display actions and alist entries.
The direct cause of the problem in the bug report is that when
user runs treesit-font-lock-recompute-features to add the
emacs-devel feature in c-ts-mode's mode hook, the added query
for emacs-devel aren't compiled.
This change consists of two parts:
1. The immediate fix: validate and compile queries in
treesit-font-lock-recompute-features.
2. To make it more fool-proof, change treesit-font-lock-rules
back to compile the queries and make
treesit--compile-query-with-cache support compiled queries. This
way, as long as the query goes through treesit-font-lock-rules,
it'll be compiled eventually and not cause slow-down. I had to
add some c-level functions, but they're kind of overdue anyway,
so I don't have any problem adding them to the API.
* lisp/treesit.el (treesit--compile-query-with-cache): Support
compiled queries.
(treesit-font-lock-rules): Compile the queries (but not
eagerly).
(treesit-font-lock-recompute-features): Validate and compile
queries.
(treesit-major-mode-setup): Remove call to
treesit-validate-and-compile-font-lock-rules since it's now
called in treesit-font-lock-recompute-features.
* src/treesit.c (Ftreesit_query_eagerly_compiled_p):
(Ftreesit_query_source): New functions.
* doc/lispref/parsing.texi (Pattern Matching): Add manual
entries for the new functions.
Discussed in
https://lists.gnu.org/archive/html/emacs-devel/2025-12/msg00903.html.
* lisp/progmodes/project.el (project-files) <vc>: Look up the
value of 'project-vc-ignores' inside DIR rather than project root.
This allows this variable to be applied in external dirs as well.
(project-ignores) <vc>: Do likewise.
(project--value-in-dir): Ensure that DIR has a trailing slash.
The caller might pass the value without slash if produced
differently.
* test/lisp/progmodes/project-tests.el
(project-vc-ignores-in-external-directory): New test.
(project-vc-supports-files-in-subdirectory): Update the test,
considering the subtle change in behavior: the directory DIR
directly includes a .dir-locals.el which sets project-vc-ignores,
which wasn't previously applied because it wasn't set in the root.
