* lisp/gnus/gnus-topic.el (gnus-topic-line-format)
(gnus-topic-line-format-alist, gnus-topic-prepare-topic)
(gnus-topic-insert-topic-line, )
(gnus-topic-display-missing-topic, )
(gnus-topic-update-topic-line):
Provide number of groups in subtopics for topic line display and
add 'G' element to be used for 'gnus-topic-line-format' variable
* doc/misc/gnus.texi (Topic Variables):
Document 'G' element for 'gnus-topic-line-format' variable (bug#49068).
Correct the previous fix (did not correctly handle qualified types).
Also fix the following issues:
- remove module name (+ dot) from tags, as prefixing module name is
often inconsistent in code and may cause tags to be too specific.
- now tag 0-arity predicates and functions (':- func foo_14.')
- now tag one-word declarations (':- interface.')
* lib-src/etags.c (mercury_pr): Pass the correct NAME and NAMELEN
arguments to 'make_tag'.
(mercury_decl): Return more information about the declaration or
definition it finds. This allows mercury_pr to be smarter.
(Bug#47408)
Since Emacs Lisp is more than Lisp-2, separately output letters
for each namespace: functions, variables, faces, classes.
Use non-letter characters for additional properties
such as advice, obsolete, local.
(help--symbol-completion-table-affixation): Use format "%-4s".
https://lists.gnu.org/archive/html/emacs-devel/2021-06/msg00524.html
* lisp/window.el (display-buffer-override-next-command):
Separate 'postfun' from 'clearfun', so 'clearfun' resets
'display-buffer-overriding-action', whereas 'postfun' calls
'post-function' that can select the right window in 'post-command-hook'.
* lisp/windmove.el (windmove-display-no-select): Add new choice 'ignore'.
Improve docstring.
(windmove-display-in-direction): Use new value 'ignore' of
'windmove-display-no-select'. Improve docstring.
(windmove-display-left, windmove-display-up)
(windmove-display-right, windmove-display-down): Mention
'windmove-display-no-select' in docstrings.
* src/doprnt.c (exprintf, evxprintf):
* src/lisp.h (exprintf, evxprintf): Don't use a pointer-to-const type
for the `nonheapbuf` argument: although it is never dereferenced, GCC
will warn when passing a pointer to uninitialised memory otherwise.
* src/fns.c (sort_vector_copy, realize_face, realize_gui_face)
(realize_tty_face): Use the same signatures in the prototypes as in
the actual function definitions.
* doc/lispref/modes.texi (Search-based Fontification):
Remove paragraph that has been unnecessary and confusing ever since
regexp-opt stopped using capturing brackets more than 21 years ago.
* src/pdumper.c (struct dump_context): Add buf, buf_size, max_offset fields.
(dump_grow_buffer): New function.
(dump_write): Use memcpy, not an actual emacs_write.
(dump_seek): Keep track of maximum seen offset. Don't actually seek.
(Fdump_emacs_portable): Write out the file contents when done.
* lisp/mouse.el (mouse-drag-line): Do the right thing in the
presence of margins (bug#41156).
(mouse-drag-line): Bind left-margin/right-margin in the map, too.
* lisp/emacs-lisp/bytecomp.el (byte-compile--wide-docstring-p):
Don't consider the function signature when determining whether the
doc string is too wide (bug#49007). (The signature is folded
later when displaying help.)
* lisp/mh-e/mh-junk.el (mh-spamassassin-*list): Remove support
for SpamAssassin 2.20. (SpamAssassin 3.0 was released in 2004.)
This change updates both the flags for sa-learn and the comments
about how the current version works.
This change reverts part of a change made in 2003 that added
support for what was even then an "old version of spamassassin."
* lisp/calendar/iso8601.el (iso8601-parse-time): Get fractional
times (with leading zeroes in the fraction part) right (bug#49017).
Fix based on a patch by "J.P." <jp@neverwas.me>.
* lisp/emacs-lisp/bytecomp.el (byte-compile-file): Produce .elc
temporary files only when non native compiling or when native
compiling but `byte+native-compile' is non nil.
* doc/misc/tramp.texi (Auto-save and Backup):
Describe tramp-allow-unsafe-temporary-files.
(Ad-hoc multi-hops): Use proper format.
* etc/NEWS: Mention confirmation for writing sensitive auto-save
or backup remote files to the local temporary directory..
* lisp/net/tramp-cache.el (tramp-dump-connection-properties):
Strengthen test.
* lisp/net/tramp.el (tramp-allow-unsafe-temporary-files): New defcustom.
(tramp-handle-find-backup-file-name)
(tramp-handle-make-auto-save-file-name): Don't expose sensible
auto-save or backup files on local temporary directory. (Bug#45245)
* test/lisp/net/tramp-tests.el (tramp--test-always): New defalias.
(tramp-test10-write-region, tramp-test21-file-links)
(tramp--test--deftest-direct-async-process): Use it.
(tramp-test37-make-auto-save-file-name)
(tramp-test38-find-backup-file-name): Extend tests.
* doc/lispref/windows.texi (Buffers and Windows): Mention the
special handling of side windows and add a reference.
(Buffer Display Action Alists): Say explicitly that
`display-buffer-in-side-window' is dedicating to side by default.
(Dedicated Windows): Add case (4) and explain its meaning, add
a reference.
(Displaying Buffers in Side Windows): Move the paragraph about
`switch-to-(prev|next)-buffer' into a new item to emphasize the
special meaning of dedication for side windows.
* lisp/window.el (set-window-buffer-start-and-point): Restore
side dedication.
(switch-to-prev-buffer, switch-to-next-buffer): Correct return
value that should be nil instead of the same buffer in case of
no change.
(delete-windows-on): Restore side dedication.
(replace-buffer-in-windows): Update the docstring, restore side
dedication.
(quit-restore-window): Rearrange the logic so that strongly
dedicated windows are eventually deleted first. Restore the
side dedication. In the final case try to
`switch-to-prev-buffer' before deleting a window (Bug#48367).
* src/fileio.c (Fsubstitute_in_file_name): Mention that undefined
variables aren't replaced (as opposed to what happens in a shell
substitution) (bug#40949).
* doc/emacs/killing.texi (Clipboard): Document it.
* lisp/simple.el (save-interprogram-paste-before-kill): Extend
range (bug#41168).
(kill-new): Implement it (bug#41168).
