* lisp/vc/vc-git.el (vc-git-stash-map): Give `vc-git-stash' a
keystroke, because it's a useful command.
When there are no stashes, there's no way to use that keystroke (since
it's only bound on the stash lines), so perhaps that should also be
fixed somehow.
* lisp/progmodes/m4-mode.el (m4-font-lock-keywords): "dnl"
comments should override any previous syntax-based highlighting
(37132). This makes "dnl foo#bar" be highlighted correctly.
Move description of value to syntax-ppss function.
* lisp/emacs-lisp/syntax.el (syntax-ppss): Move the description of
the return value from...
* src/syntax.c (Fparse_partial_sexp): ... here because
`syntax-ppss' is what's called over the place, and jumping through
an indirection to get to the value description is inconvenient.
Juri Linkov [Wed, 30 Oct 2019 00:00:02 +0000 (02:00 +0200)]
Rename read-char-with-history to read-char-from-minibuffer (bug#10477)
* lisp/simple.el (read-char-from-minibuffer-history):
Rename from read-char-with-history--history.
(read-char-from-minibuffer-map):
Rename from read-char-with-history--map.
(read-char-from-minibuffer): Rename from read-char-with-history.
(read-char-from-minibuffer-self-insert): New command.
(zap-to-char): Use read-char-from-minibuffer.
Juri Linkov [Tue, 29 Oct 2019 23:01:11 +0000 (00:01 +0100)]
Reimplement read-char-with-history based on read-from-minibuffer
* lisp/simple.el (read-char-with-history): Reimplement based on
read-from-minibuffer.
* lisp/simple.el (read-char-with-history--map): New variable
(bug#10477).
Juri Linkov [Tue, 29 Oct 2019 22:31:11 +0000 (00:31 +0200)]
* lisp/tab-bar.el: Store point position and frame buffer-lists in tab.
* lisp/tab-bar.el (tab-bar--tab): Store additionally point-marker,
frame parameters buffer-list and buried-buffer-list, both for wc and ws.
Use seq-filter.
(tab-bar-select-tab): Restore point-marker after restoring
window-configuration. Also restore frame parameters buffer-list
and buried-buffer-list both for window-configuration and window-state.
(tab-bar-history-limit, tab-bar-history--minibuffer-depth): New variables.
(tab-bar-history-current): Rename from tab-bar-history--pre-change.
(tab-bar-history--pre-change): Set tab-bar-history-current.
(tab-bar--history-change): Use seq-take for tab-bar-history-limit.
(tab-bar-history-back, tab-bar-history-forward): Restore point-marker.
(tab-bar-list-noselect): Use seq-remove.
Default exotic image formats (like .webp) to image-mode
* doc/lispref/errors.texi (Standard Errors): Mention the new error.
* lisp/files.el (auto-mode-alist): Add a bunch of image suffixes
to the list (bug#37972) based on the output from "gm convert -list
format" (i.e., graphicsmagick).
* lisp/image-mode.el (image-mode): Rewrite to possibly notify the
user about image-use-external-converter.
(image-mode--setup-mode): Factor out into own function and don't
run under `condition-case' as there's nothing here that should
error.
* lisp/image.el (unknown-image-type): New error.
(image-type): Signal that error so that image-mode can offer
sensible feedback to the user.
Stefan Monnier [Tue, 29 Oct 2019 20:17:14 +0000 (16:17 -0400)]
* lisp/minibuffer.el: Tweak and undo parts of recent changes
(completion-metadata): Always return a fresh new cons cell.
(completion--nth-completion): Don't bother calling adjust-metadata
if the result won't be used.
(completion-pcm--hilit-commonality): Revert recent change which had
removed support for `completions-first-difference` in `substring` and
`partial-completion` styles.
(completion--flex-adjust-metadata): Treat the arg as immutable.
* src/xfaces.c (merge_face_vectors) : Pass ATTR_FILTER to
merge_face_ref to stop recursion merging in advance when possible.
(merge_named_face) : add more conditions to potentially reduce calls to
merge_face_vectors in more unneeded situations.
* src/xfaces.c (merge_face_vectors) : Add extra parameter ATTR_FILTER
with the same functionality than in merge_named_face. Update all
occurrences.
(merge_named_face): Remove condition to avoid call
`merge_face_vectors`.
João Távora [Tue, 29 Oct 2019 00:18:55 +0000 (00:18 +0000)]
Make icomplete-exhibit actually work when navigating up directories
Sometimes, when finding files with icomplete-mode, backward-deleting
the previous word or sexp (to move up a directory) doesn't actually
refresh the file list of the new directory. Forcing redisplay in
icomplete-exhibit misteriously fixes the problem.
* lisp/icomplete.el (icomplete-exhibit): Add call to redisplay.
João Távora [Tue, 29 Oct 2019 00:12:16 +0000 (00:12 +0000)]
New commands for making icomplete behave more like ido
This tries (not particularly hard) to emulate Ido's magic C-k, C-d,
RET and DEL. For now, the new commands have to be explicitly added to
an active map by the user, using something like this:
Juri Linkov [Mon, 28 Oct 2019 21:55:16 +0000 (23:55 +0200)]
* lisp/tab-bar.el: Check for minibuffer-depth in tab-bar-history-mode.
* lisp/tab-bar.el (tab-bar-history--pre-change): Store also minibuffer-depth.
(tab-bar-history-change): Check stored minibuffer-depth for zero.
(tab-bar-history-mode): Use prefix -- for internal functions.
* lisp/progmodes/python.el (python-rx): Use `rx-let' instead of
`rx-constituents'. This allows for some slight redundancy reduction,
since `rx-let' definitions are expanded inside `not' (bug#37849).
Reorder some `or' forms for more efficient matching.
* lisp/erc/erc.el (erc-hide-current-message-p): Make
erc-channel-hide-list work (bug#37879) by getting matching on the
channel name instead of the ERC client name.
Fix `G c' in Gnus group buffers for non-ASCII text
* lisp/gnus/gnus-cus.el (gnus-group-customize): Decoding is
superfluous here -- everything is utf-8 already.
(gnus-group-customize-done): Don't double-encode the text (bug#37901).
Koichi Arakawa [Mon, 28 Oct 2019 08:49:59 +0000 (09:49 +0100)]
Reorder command-line switches in Tramp
* lisp/net/tramp-sh.el (tramp-maybe-open-connection):
`tramp-encoding-command-interactive' must be the last command-line
switch, at least for bash. (Bug#37953)
Juri Linkov [Sun, 27 Oct 2019 23:04:47 +0000 (01:04 +0200)]
* lisp/tab-bar.el: Add tab-bar-history-mode with arrow buttons for navigation.
* lisp/tab-bar.el (tab-bar-back-button, tab-bar-forward-button):
New variables.
(tab-bar-make-keymap-1): Show these buttons in tab-bar-history-mode.
(tab-bar--tab): Add history-back and history-forward.
(tab-bar-select-tab): Restore history-back and history-forward.
(tab-bar-history-omit, tab-bar-history-back)
(tab-bar-history-forward, tab-bar-history-pre-change): New variables.
(tab-bar-history-pre-change, tab-bar-history-change)
(tab-bar-history-back, tab-bar-history-forward)
(tab-bar-history-mode): New functions.
Stefan Monnier [Sun, 27 Oct 2019 21:00:55 +0000 (17:00 -0400)]
* .gitignore: Don't ignore .rej files
Left over rejected hunks that we forgot to merge by hand should not be
silently ignored. Better explicitly remove/rename .rej files when
done with them.
Stephen Gildea [Sun, 27 Oct 2019 15:20:13 +0000 (08:20 -0700)]
time-stamp-time-zone: update customization
* time-stamp.el (time-stamp-time-zone): Support customization with
an integer offset (a new possible value of the ZONE argument to
format-time-string in Emacs 27).
Update the safe-local-variable predicate from string-or-null-p
(describing time-stamp-time-zone's domain before 2015) to new
predicate time-stamp-zone-type-p (describing the current domain).
* time-stamp-tests.el (time-stamp-test-helper-zone-type-p): New test.
Eli Zaretskii [Sun, 27 Oct 2019 15:07:47 +0000 (17:07 +0200)]
Fix point position after revert-buffer in tabulated-list mode
* lisp/emacs-lisp/tabulated-list.el (tabulated-list-print):
Don't use count-screen-lines, as it is unreliable when lines
are truncated and the region ends before a newline (the root
cause is in vertical-motion, but is very hard to fix there).
Instead, use vertical-motion directly, as the problems with
counting/not counting the newline that ends the region are not
relevant to what we need in this function, which is simply the
number of the current window line. (Bug#37941)
João Távora [Sun, 27 Oct 2019 13:20:56 +0000 (13:20 +0000)]
Optimize lisp/icomplete.el when default completion available
Often, when using icomplete with icomplete-show-matches-on-no-input
and commands like C-h f, icomplete-exhibit will take a long time (like
several seconds) to show all the completions.
However, if there is a non-nil default value, like when the C-h f is
performed exactly on a function name, the minibuffer prompt will
reflect that immediately. If the user immediately presses RE
(icomplete-force-complete-and-exit) we take that to mean "use the
default" and avoid the long wait.
As an extra performance tweak, this commit also removes an unneeded
call to icomplete-exhibit in icomplete-minibuffer-setup.
* lisp/icomplete.el (icomplete-force-complete-and-exit): Short-circuit
possibility of non-nil minibuffer-default.
(icomplete-minibuffer-setup): No need to explicitly icomplete-exhibit
here since icomplete-post-command-hook will call it unconditionally.
João Távora [Sun, 27 Oct 2019 00:33:54 +0000 (01:33 +0100)]
Improve scoring algorithm for flex-style completions
The previous algorithm had two problems: it considered non-matches in
the beginning and end of the string as matching "holes" and failed to
penalize larger holes, making flex-score-match-tightness only
effective in some corner cases.
The new formula, which is described in code and in pseudo-code in the
comments, fixes these problems.
As a result, by default, C-h f flex now correctly bubbles up
"company-search-flex-regexp" to the top, in front of "file-exists-p".
With a flex-score-match-tightness smaller than 1.0, the situation is
reversed.
Juri Linkov [Sat, 26 Oct 2019 23:20:15 +0000 (02:20 +0300)]
dired-dwim-target uses most recently visited window instead of next window.
* doc/emacs/dired.texi (Operating on Files): Document behavior change.
* lisp/dired-aux.el (dired-dwim-target-directories): New function.
(dired-dwim-target-directory, dired-dwim-target-defaults): Use it
to get the most recently used window instead of the next window (bug#35385).
* lisp/dired.el (dired-dwim-target): Doc fix.
* test/lisp/dired-tests.el: Remove unnecessary require and pacify
byte-compiler.
Juri Linkov [Sat, 26 Oct 2019 22:38:32 +0000 (01:38 +0300)]
Customizable tab-line-tabs-function and new buffer-local tab-line-mode.
* lisp/tab-line.el (tab-line-tabs-function): Turn defvar into defcustom
with default of tab-line-tabs-window-buffers.
(tab-line-tabs-mode-buffers): New function.
(tab-line-tabs-window-buffers): Rename from tab-line-tabs.
(tab-line-select-tab, tab-line-switch-to-prev-tab)
(tab-line-switch-to-next-tab): Handle the value tab-line-tabs-window-buffers
of tab-line-tabs-function specially.
(tab-line-mode): Rename from global-tab-line-mode and change scope
to buffer-local.
(global-tab-line-mode): New globalized-minor-mode.
(tab-line-mode--turn-on): New function.
(tab-line-exclude-modes): New defcustom.
(tab-line-exclude): New buffer-local variable.
https://lists.gnu.org/archive/html/emacs-devel/2019-10/msg00988.html
* lisp/dired.el (dired-mark-pop-up): Set tab-line-exclude to nil.
Don't set tab-line-format. (Bug#37699)
* lisp/speedbar.el (speedbar-mode): Set tab-line-exclude to nil.
Don't set tab-line-format.
Juri Linkov [Sat, 26 Oct 2019 22:16:10 +0000 (01:16 +0300)]
* lisp/tab-bar.el (tab-bar-switch-to-recent-tab): New command.
(tab-recent): Alias to tab-bar-switch-to-recent-tab.
(tab-bar--tab-index-recent): New internal function.
(tab-bar-close-tab-select): Add new default option 'recent'.
(tab-bar-close-tab): Handle it.
Robert Cochran [Thu, 24 Oct 2019 00:34:24 +0000 (17:34 -0700)]
Add customization option for what do when the last tab is closed
* lisp/tab-bar.el (tab-bar-close-last-tab-choice): New custom
variable.
(tab-bar-close-tab): Handle closing the last tab specially, referring to
tab-bar-close-last-tab-choice.
Stephen Gildea [Sat, 26 Oct 2019 21:31:22 +0000 (14:31 -0700)]
time-stamp-tests: add name prefix to tests of formatting
* time-stamp-tests.el: rename all the time-stamp-string formatting tests
to have the word "format" in their name, to make room in the namespace
for other, future tests.
João Távora [Sat, 26 Oct 2019 20:47:36 +0000 (21:47 +0100)]
Restore default value of face completions-common-part
The previous commit titled "Rework face hints for partial-string
completion styles" contained a potentially controversial
backwards-incompatible change to this face's default value.
Arash Esbati [Sat, 26 Oct 2019 11:29:14 +0000 (13:29 +0200)]
Improve support for biblatex and cleveref macros
* lisp/textmodes/reftex-cite.el (reftex-figure-out-cite-format):
Extend regexp to match additional cite commands from biblatex.sty.
* lisp/textmodes/reftex-dcr.el (reftex-view-crossref): Extend
regexp to match additional cite commands from biblatex.sty and
reference commands from cleveref.sty.
Arash Esbati [Sat, 26 Oct 2019 11:25:19 +0000 (13:25 +0200)]
Avoid infloop's by doing a case-sensitive match
* lisp/textmodes/reftex-ref.el (reftex-format-special): Be
case-sensitive when checking the actual reference macro against
\ref. This avoids infloop when \Ref is chosen.
João Távora [Sat, 26 Oct 2019 13:31:38 +0000 (14:31 +0100)]
Rework face hints for partial-string completion styles
Don't use completions-first-difference for the 'substring', 'flex' and
'partial-completion' styles, since there can be really no reasonable
definition of a "first" difference there.
Make completions-common-part inherit from 'underline' so that it is
useful by default for all completion styles.
* lisp/minibuffer.el (completions-common-part): Adjust
description and change default value.
(completions-first-difference): Ajust docstring.
(completion-pcm--hilit-commonality): Don't use
completions-first-difference.
João Távora [Sat, 26 Oct 2019 13:13:08 +0000 (14:13 +0100)]
Allow completion styles to adjust completion metadata
This commit re-does the now-reverted commit with the same title. That
version relied on generic functions, which cannot be used yet in files
such as lisp/minibuffer.el. This version uses a symbol property
completion--adjust-metadata instead.
The new facility allows completion styles to have a say in metadata
properties such as cycle-sort-function and display-sort-function.
This is especially useful for completion styles such as 'flex', which
generally produce many matches, including some potentially "obscure"
ones. The default sorting strategy would often bubble the latter to
the top of the list.
The sorting function for 'flex' considers pre-computed matching scores
and is thus much better than the default for this particular style.
Additionally, it overrides the completion table's cycle-sort-function
or display-sort-function properties if they exist, although it still
uses them to pre-sort the result, so that they are still relevant for
resolving ties.
* lisp/minibuffer.el (completion--nth-completion)
(completion--flex-adjust-metadata): New helper.
(flex): Put completion--adjust-metadata property.
João Távora [Fri, 25 Oct 2019 22:57:44 +0000 (23:57 +0100)]
Allow completion styles to adjust completion metadata
The new facility, realized in the completion-adjust-metadata-for-style
generic, allows completion styles to have a say in metadata properties
such as cycle-sort-function and display-sort-function. This is
especially useful for completion styles such as 'flex', which
generally produce many matches, including some potentially "obscure"
ones. The default sorting strategy would often bubble the latter
to the top of the list.
The sorting function for 'flex' considers pre-computed matching scores
and is thus much better than the default for this particular style.
Additionally, it overrides the completion table's cycle-sort-function
or display-sort-function properties if they exist, although it still
uses them to pre-sort the result, so that they are still relevant for
resolving ties.
* lisp/minibuffer.el (completion--nth-completion): Call
completion-adjust-metadata-for-style.
(completion-adjust-metadata-for-style): New generic.
(completion-adjust-metadata-for-style 'flex): New method.
memeplex [Sat, 19 Oct 2019 03:13:15 +0000 (00:13 -0300)]
Replay key if kboard is interrupted while initializing (Bug#37782)
The problem with the original fix for bug#5095 is that it drops
the current event, which is a valid character event and not -2.
Thus, the first ESC character sent by the terminal after turning
on focus tracking is lost and we get '[' and 'I' events separately
inserted into the buffer afterwards.
* src/keyboard.c (read_key_sequence): Add key as mock_input and replay
sequence using new keyboard, when the key is not -2.
Tassilo Horn [Fri, 25 Oct 2019 20:07:39 +0000 (22:07 +0200)]
Adapt doc-view to latest image-mode changes
The functions image-scroll-down, image-scroll-up, image-next-line, and
image-previous-line now return pixel values instead of multiples of
character height as of commit 9c66b09950.
* lisp/doc-view.el (doc-view-scroll-up-or-next-page):
(doc-view-scroll-down-or-previous-page):
(doc-view-next-line-or-next-page):
(doc-view-previous-line-or-previous-page):
(doc-view-insert-image): Adapt to image-scroll-up/down and
image-next/previous-line now returning pixel values instead of
multiples of character heights. Fixes bug#37874.
Alan Mackenzie [Fri, 25 Oct 2019 20:11:48 +0000 (20:11 +0000)]
CC Mode: Fix positioning of point whilst inserting comments without non-ws
* lisp/progmodes/cc-cmds.el (c-guess-fill-prefix): When determining a new
block comment prefix (i.e. there isn't one already there to copy), and that
prefix is hard up against a comment closer, ensure there are at least two
spaces before the closer.
(c-indent-new-comment-line): Amend the strategy for ensuring that point isn't
left hard up against the comment closer after M-j.
* lisp/gnus/gnus-sum.el (gnus-thread-hide-subtree): Make the
"Non-nil" predicate do what it's supposed to (i.e., return non-nil
on non-nil values (excepting predicates)) (bug#37916).
Stefan Monnier [Fri, 25 Oct 2019 03:06:23 +0000 (23:06 -0400)]
* lisp/cedet/mode-local.el: Clean up name space
Mostly renaming functions by adding `mode-local--` to their name
and leaving an obsolete alias behind, just in case.
(define-child-mode): Make obsolete.
(mode-local--set-parent): Rename from set-mode-local-parent.
(mode-local--new-bindings): Rename from new-mode-local-bindings.
Use `obarray-make`.
(mode-local--activate-bindings): Rename from activate-mode-local-bindings.
(mode-local--deactivate-bindings): Rename from
deactivate-mode-local-bindings.
(make-obsolete-overload): Rename properties with a `mode-local--` prefix.
Adjust all users.
(mode-local--overload-obsoleted-by): Rename from overload-obsoleted-by.
(mode-local--overload-that-obsolete): Rename from overload-that-obsolete.
(mode-local--function-overload-p): Rename from function-overload-p.
(mode-local-read-function): Mark obsolete.
(mode-local--overload-docstring-extension): Rename from
overload-docstring-extension.
(mode-local--describe-overload): Rename from describe-mode-local-overload.
* lisp/cedet/semantic/grammar-wy.el (semantic-grammar-wy--install-parser):
* lisp/cedet/semantic/bovine/grammar.el (bovine-grammar-mode):
* lisp/cedet/semantic/texi.el (semantic-default-texi-setup):
* lisp/cedet/semantic/wisent/grammar.el (wisent-grammar-setupcode-builder)
(wisent-grammar-mode):
* lisp/cedet/semantic/html.el (semantic-default-html-setup):
Make the `semantic-` prefix explicit to ease grep search.
(html-helper-mode): Remove obsolete setting.
* lisp/cedet/semantic/wisent/javascript.el: Fix js-mode/javascript-mode
mixup so we don't need define-child-mode any more.
(semantic-get-local-variables, semantic-ctxt-current-symbol)
(semantic-tag-protection, semantic-analyze-scope-calculate-access):
Use `js-mode` rather than `javascript-mode` as the mode name since
that's the real mode's name.
* lisp/cedet/semantic/wisent/python.el (python-2-mode, python-3-mode):
Remove child declaration for non-existing modes.
Paul Eggert [Thu, 24 Oct 2019 21:32:06 +0000 (14:32 -0700)]
Update from Gnulib
This incorporates:
2019-10-23 nstrftime: speed up integer overflow checking
2019-10-23 port better to GCC under macOS
2019-10-15 inttypes: use more-robust test for int range
2019-10-14 update-copyright: use en dashes in .texi ranges
* build-aux/update-copyright, lib/intprops.h, lib/inttypes.in.h:
* lib/nstrftime.c, lib/verify.h:
Copy from Gnulib.
Stefan Monnier [Wed, 23 Oct 2019 21:48:41 +0000 (17:48 -0400)]
* lisp/cedet/mode-local.el: Use lexical-binding and `declare`
(with-mode-local-symbol, with-mode-local, setq-mode-local)
(defvar-mode-local, defconst-mode-local)
(define-overloadable-function, define-mode-local-override):
Use `declare` for indent and edebug specs.
(xref-mode-local-find-overloadable-regexp): Simplify regexp.
(mode-local-setup-edebug-specs): Delete.
(edebug-setup-hook): Don't use any more.
Juri Linkov [Wed, 23 Oct 2019 20:58:30 +0000 (23:58 +0300)]
* lisp/tab-bar.el: Allow to specify interactively where to add a new tab.
* lisp/tab-bar.el (tab-bar-new-tab-to): Rename from tab-bar-new-tab.
Add optional arg TO-INDEX.
(tab-bar-new-tab): New implementation to use relative ARG.
(tab-new-to): Alias to tab-bar-new-tab-to.
Juri Linkov [Wed, 23 Oct 2019 20:53:08 +0000 (23:53 +0300)]
* lisp/tab-bar.el: Rename tab-bar-swap-tabs to tab-bar-move-tab-to.
* lisp/tab-bar.el (tab-bar-move-tab-to): Rename from tab-bar-swap-tabs.
Change logic to push the moved tab between existing tabs.
(tab-bar-move-tab): Call tab-bar-move-tab-to instead of tab-bar-swap-tabs.
(tab-move-to): Rename alias from tab-swap.
Juri Linkov [Wed, 23 Oct 2019 20:39:31 +0000 (23:39 +0300)]
Implement tab close undo feature for tab-bar.
* lisp/tab-bar.el (tab-bar-closed-tabs): New variable.
(tab-bar-close-tab): Add closed tab to tab-bar-closed-tabs.
Disable tab-bar-mode on closing the last tab.
(tab-bar-close-other-tabs): Add all closed tabs to tab-bar-closed-tabs.
(tab-bar-undo-close-tab): New command.
(tab-undo): Alias to tab-bar-undo-close-tab.
Completion tables are not supposed to signal errors:
emacs -q
M-x semantic-mode
; visit a file supported by semantic, such as a C file
; put cursor in a blank space
M-x completion-at-point
It will error with: "Nothing to complete"
(semantic-analyze-possible-completions): Return nil instead of
signaling errors.