Eli Zaretskii [Fri, 3 Jan 2020 07:28:35 +0000 (09:28 +0200)]
Fix redisplay when mode-line-format changes mode-line's height
* lisp/frame.el (top-level): Add mode-line-format,
tab-line-format, and header-line-format to the list of variables
that should trigger an immediate redisplay of the buffer's
window. This fixes redisplay of windows when the mode line
changes its height.
* src/xdisp.c (window_box_height): Use the window's
mode_line_height, tab_line_height, and header_line_height fields
in preference to CURRENT_MODE_LINE_HEIGHT,
CURRENT_TAB_LINE_HEIGHT, and CURRENT_HEADER_LINE_HEIGHT,
respectively. This fixes display of vertical scroll bar when the
height of the window's mode line changes.
* src/dispnew.c (adjust_glyph_matrix): When resizing a window's
matrix, reset the mode_line_p flag of the previous mode-line row,
so that the window_box_height, CURRENT_MODE_LINE_HEIGHT, and their
ilk won't use stale info. (Bug#38828)
Eli Zaretskii [Wed, 1 Jan 2020 17:55:35 +0000 (19:55 +0200)]
Fix compilation with GTK versions older than 3
* src/xfns.c (x_get_net_workarea): Change a recently moved #ifndef
so that GTK builds which need this function will compile it.
Reported by John <jpff@codemist.co.uk>.
Eli Zaretskii [Wed, 1 Jan 2020 17:49:57 +0000 (19:49 +0200)]
Fix reverting customizations
* lisp/custom.el (custom-push-theme): Don't use setcar to modify
the recorded value of PROP; instead, cons a new property list by
deleting the old value and adding the new one. (Bug#38812)
Occasionally, loading cl-generic.el from source requires
max_specpdl_size > 1600 when bootstrapping, and thus fails.
In any case we are very close to the limit.
Paul Eggert [Wed, 1 Jan 2020 01:03:14 +0000 (01:03 +0000)]
Pacify gcc -Wunused-function on Ubuntu 18.04.3
This improves on 2019-12-25T20:01:07Z!eggert@cs.ucla.edu,
by fixing a GCC warning on Ubuntu 18.04.3
“‘x_get_net_workarea’ defined but not used”.
* src/xfns.c (x_get_net_workarea) [!USE_GTK]: Do not define.
João Távora [Wed, 25 Dec 2019 17:57:20 +0000 (17:57 +0000)]
Don't force completion recalculation in icomplete-fido-ret
Besides the adverse effect of delaying completions, it tripped up the
useful logic of icomplete-force-complete-and-exit in the case where a
default was available, but no completions calculated yet.
* lisp/icomplete.el (icomplete-fido-ret): Don't force calculation
of completions.
* test/lisp/comint-tests.el (comint-test-no-password-function)
(comint-test-password-function-with-value)
(comint-test-password-function-with-nil):
Call accept-process-output as many times as needed, with a slightly
more generous timeout (100 ms), after sending the Password: prompt to
the process, since there must be time for some back-and-forth
communication. Also clear the process-query-on-exit flag, since it
doesn't go well with noninteractive tests.
Eli Zaretskii [Mon, 30 Dec 2019 16:17:03 +0000 (18:17 +0200)]
Fix tar-mode reading the oldgnu Tar format
This makes sure an entry for a long file name will not
accidentally appear as a directory, and thus its size will be
disregarded, causing corrupted file headers for all the subsequent
entries. The original problem happened because the long file name
truncated to 99 bytes happened to end in a slash, which tar-mode
takes to be the indication of a directory.
* lisp/tar-mode.el (tar-header-block-tokenize): Accept an
additional argument DISABLE-SLASH; if non-nil, don't set the
link-type field of the descriptor to 5 (meaning a directory) just
because the name ends in a slash. Use this argument when calling
itself recursively, to read the entry of the file with a long
name. Set the link-type to 5 if the long name ends in a slash.
(Bug#38777)
Eli Zaretskii [Mon, 30 Dec 2019 15:53:52 +0000 (17:53 +0200)]
Fix mini-window resizing under resize-mini-windows = t
* src/window.c (grow_mini_window): Fix resizing of mini-window
when the echo area is cleared, and the caller asks for the
mini-window to have zero lines. (Bug#38791)
Avoid unbounded growth of cl-random-state components (bug#38753)
* lisp/emacs-lisp/cl-extra.el (cl-random):
Perform the modulo 2**23 operation before updating the state instead
of after. The result is always the same, but it prevents the state
from growing into arbitrary large bignums.
Dmitry Gutov [Sun, 29 Dec 2019 12:22:11 +0000 (15:22 +0300)]
Fix up requires
* lisp/progmodes/project.el (project-files):
Remove (require 'xref).
(project--files-in-directory): Add it here instead.
(project-find-regexp, project-or-external-find-regexp): And here.
For 'xref--show-xrefs'.
Dmitry Gutov [Sun, 29 Dec 2019 12:44:08 +0000 (15:44 +0300)]
Rename xref-collect-references and xref-collect-matches
* lisp/progmodes/xref.el (xref-references-in-directory): Rename
from xref-collect-references. Update the sole caller.
(xref-matches-in-directory): Rename from xref-collect-matches.
Update all callers (all of them are in the /tests/ dir).
* test/lisp/progmodes/xref-tests.el (xref-tests-data-dir):
Don't use the EMACS_TEST_DIRECTORY env var. It doesn't work when
running interactively.
Dmitry Gutov [Sat, 28 Dec 2019 16:28:30 +0000 (19:28 +0300)]
Do some renames for clarity
* lisp/progmodes/xref.el (xref--last-file-buffer):
Rename from xref--last-visiting-buffer. Update users.
(xref--find-file-buffer): Rename from xref--find-buffer-visiting
to match the standard function that gets called inside. Ditto.
Eli Zaretskii [Sat, 28 Dec 2019 13:19:05 +0000 (15:19 +0200)]
Rearrange NEWS, add missing documentation
* etc/NEWS: Rearrange and mark entries whether documented or not.
* doc/lispref/streams.texi (Output Variables): Document the new
default of 'print-quoted'.
* doc/lispref/keymaps.texi (Functions for Key Lookup): Document
that KEYMAP arg to 'lookup-key' can also be a list.
* doc/lispref/customize.texi (Variable Definitions): Document the
:local keyword of 'defcustom'.
* doc/lispref/numbers.texi (Float Basics): Document changes in
'logb'.
* doc/lispref/hooks.texi (Standard Hooks): Document
'comint-password-function'.
* doc/emacs/display.texi (Text Scale): Document text-size
adjustment using the mouse wheel.
* doc/emacs/frames.texi (Mouse Commands): Document image scaling
with the mouse wheel.
* doc/emacs/windows.texi (Window Convenience): Document
'global-tab-line-mode'.
* doc/emacs/search.texi (Repeat Isearch, Symbol Search)
(Isearch Yank): Document the new support for numeric arguments in
Isearch commands.
(Special Isearch): Document 'M-s M->' and 'M-s M-<'.
(Search Customizations): Document 'isearch-lazy-count'. Improve
indexing.
(Not Exiting Isearch): Document the new value of
'isearch-allow-scroll'.
* doc/emacs/maintaining.texi (Xref Commands): Document the new 'g'
key binding.
* doc/emacs/package.texi (Package Installation): Document changes
in 'package-check-signature'.
* doc/emacs/maintaining.texi (VC Change Log): Document
'vc-log-search'.
* doc/emacs/dired.texi (Operating on Files): Document
'dired-vc-rename'.
Eli Zaretskii [Sat, 28 Dec 2019 09:13:25 +0000 (11:13 +0200)]
Fix documentation of define-obsolete-* functions
* doc/lispref/functions.texi (Obsolete Functions): Make
documentation of functions that obsolete symbols match the
advertised calling conventions.
* lisp/emacs-lisp/byte-run.el (define-obsolete-function-alias)
(define-obsolete-variable-alias, define-obsolete-face-alias):
State in the doc strings that WHEN is a mandatory argument, to be
consistent with the advertised calling convention. (Bug#38675)
* lisp/gnus/gnus-start.el (gnus-group-change-level): Fix previous
patch that made info nil when adding new groups (thereby making
gnus-newsrc-alist invalid).
Dmitry Gutov [Fri, 27 Dec 2019 16:17:48 +0000 (19:17 +0300)]
(xref--find-buffer-visiting): Speed up by using get-file-buffer
* lisp/progmodes/xref.el (xref--find-buffer-visiting):
Speed up by using get-file-buffer.
(xref--collect-matches): Remove the condition on remote-id.
(https://lists.gnu.org/archive/html/emacs-devel/2019-12/msg00753.html)
Dmitry Gutov [Fri, 27 Dec 2019 15:18:41 +0000 (18:18 +0300)]
project--vc-list-files: Recurse into submodules
* lisp/progmodes/project.el (project-try-vc): Do not treat a Git
submodule as a project root, go up to the parent repo.
(project--git-submodules): New function.
(project--vc-list-files): Use it. Recurse into submodules.
Eli Zaretskii [Fri, 27 Dec 2019 14:19:09 +0000 (16:19 +0200)]
Minor improvements of buffer documentation
* doc/emacs/buffers.texi (Buffers): Say that buffers are killed
when no longer needed. Suggested by Jean-Christophe Helary
<jean.christophe.helary@traduction-libre.org>.
* lisp/calc/calc-alg.el (math-simplify-only, calc-simplify-mode)
(math-expand-formulas, calc-poly-div-remainder)
(math-living-dangerously, math-simplifying, calc-angle-mode)
(calc-prefer-frac, math-poly-base-variable):
Declare dynamic variables.
* test/lisp/calc/calc-tests.el (calc-poly-div):
Add test for at least one bug caused by missing declarations.
Dmitry Gutov [Thu, 26 Dec 2019 15:39:48 +0000 (17:39 +0200)]
Speed up dired-do-find-regexp
* lisp/dired-aux.el (dired-do-find-regexp): Speed up (bug#36857).
Previously, 'find' was called for every marked file (for plain
files and directories both). Now 'find' is only called for
directories.
Paul Eggert [Wed, 25 Dec 2019 20:01:07 +0000 (12:01 -0800)]
Port x_get_monitor_attributes_fallback to !HAVE_GTK3
* src/xfns.c (x_get_net_workarea) [!HAVE_GTK3]:
Define in this case too, since x_get_monitor_attributes_fallback
contains a call to it regardless whether HAVE_GTK3 is defined.
Wolfgang Scherer [Mon, 26 Aug 2019 22:45:48 +0000 (00:45 +0200)]
Improve vc--add-line, vc--remove-regexp
* lisp/vc/vc.el (vc--add-line): Create file if it does not exist.
Use existing buffer to avoid discrepancies with filesytem. Make sure
that the file ends with a newline.
(vc--remove-line): Do not create file if it does not exist. Use
existing buffer to avoid discrepancies with filesytem. (bug#37185)
Don't default to showing X-Faces externally in Gnus
* lisp/gnus/gnus-art.el (gnus-article-x-face-command): It's seldom
useful to display the face in an external command, so don't ever
default to that. This also fixes the problem of Emacs running as
a server in mixed graphical/non-graphical environments.
Clarify base64 requirements and say what {en,de}code_coding_region does
* src/coding.c (Fencode_coding_region): Clarify what this does.
(Fdecode_coding_region): Ditto.
* src/fns.c (Fbase64_decode_region): Clarify that this function
returns bytes, not text (bug#38587).
(Fbase64_encode_region): Clarify that this function takes bytes,
not text.
* src/image.c (imagemagick_initialize): New function,
with body taken from imagemagick_load_image.
(imagemagick_load_image, Fimagemagick_types): Use it.
Copyright-paperwork-exempt: yes
Philipp Stephani [Mon, 23 Dec 2019 14:37:49 +0000 (15:37 +0100)]
Document and verify that emacs_limb_t doesn’t have padding bits.
This is a useful property when doing further bit-twiddling with the
magnitude array before/after calling extract_big_integer or
make_big_integer. For example, constructing an emacs_limb_t object
using repeated shift-and-add should work as expected, but relies on
the type not having padding bits. Since the C standard already
guarantees that unsigned integers use a pure binary representation,
not having padding bits is enough to guarantee that the type has
unique object representations in the sense of C++’s
std::has_unique_object_representations.
* doc/lispref/internals.texi (Module Values): Document that
emacs_limb_t doesn’t have padding bits.
* src/emacs-module.c: Verify that emacs_limb_t doesn’t have padding
bits.
Fix extension of underline, overline, and strike-through attributes
* src/xdisp.c (extend_face_to_end_of_line): Don't return early
if face attributes beyond background color and box are set,
since that means these attributes need to be extended past the
EOL.
* etc/NEWS:
* lisp/comint.el (comint-password-function): New variable.
(comint-send-invisible): Use it.
* test/lisp/comint-tests.el (comint-test-no-password-function,
comint-test-password-function-with-value,
comint-test-password-function-with-nil): Test new variable.
Juri Linkov [Sun, 22 Dec 2019 23:02:22 +0000 (01:02 +0200)]
* lisp/files.el (save-buffers-kill-emacs): Mark "*Process List*" as dedicated.
Using the 'dedicated' prop for display-buffer--maybe-at-bottom marks the
window of the buffer "*Process List*" as dedicated, so it won't be replaced
by another buffer that might be visited following links to process buffers
from the buffer "*Process List*".
João Távora [Sun, 22 Dec 2019 21:51:08 +0000 (22:51 +0100)]
Fix bug in flex completion style's sorting and simplify
This previous commit targetting this function introduced a bug whereby
the completion table's sorting function wouldn't be called. That is
fixed by this commit, which also simplifies the function further: it
now skips re-sorting the completions completely if there is no
minibuffer input at all (in other words, when flex isn't doing
anything useful).
Stefan Monnier [Sun, 22 Dec 2019 21:09:31 +0000 (16:09 -0500)]
* lisp/gnus/gnus-start.el: Use lexical-binding
(gnus-group-change-level, gnus-make-hashtable-from-newsrc-alist):
Use gnus-info-make.
(gnus-make-hashtable-from-newsrc-alist): Prefer `gnus-info-group`
to `car` when applied to a `gnus-info` object.
(gnus-make-hashtable-from-killed): Remove unused vars `lists` and `list`.
(gnus-gnus-to-quick-newsrc-format): Extract common code from if branches.
* lisp/gnus/gnus.el (gnus-info-make): New constructor.
João Távora [Sun, 22 Dec 2019 11:52:17 +0000 (12:52 +0100)]
Improve sorting of flex completion style with non-nil minibuffer-default
This affects the behaviour of flex completion when there is a default
completion and the user hasn't entered any input pattern to flex-match
against. It is most visible when icomplete-mode or fido-mode are
being used in conjunctio.
When using M-x man, for instance, the default completion is picked
from text around point. Say it is "emacs" (for Emacs's man page). It
will not match the intended completion, "emacs(1)", exactly. If the
user hasn't yet given any input to the completion prompt, that
completion should bubble to top so that
icomplete-force-complete-and-exit will select it, but it didn't.
This new approach uses 'string-prefix-p' instead of 'equal' to find
the default to bubble to the top. This strategy could eventually be
improved, most naturally by flex-matching the default string to all
the candidates and picking the highest scoring one.
Additionally, the new strategy only considers minibuffer-default if
there is no input in the minibuffer, which seems sensible and produces
a small but noticeable speedup.
projects / emacs.git / log