mm-possibly-verify-or-decrypt should only issue errors when decrypting
* lisp/gnus/mm-decode.el (mm-possibly-verify-or-decrypt): Only do
the warning about failed decryption when we're handing encrypted
parts, not when doing verification.
Stefan Kangas [Mon, 16 Sep 2019 01:48:47 +0000 (03:48 +0200)]
Add new function to clear tags in tabulated list
* lisp/emacs-lisp/tabulated-list.el (tabulated-list-clear-all-tags):
New function to clear all tags from padding area in current buffer.
* doc/lispref/modes.texi (Tabulated List Mode): Document it.
* etc/NEWS: Announce it.
Don't indent unrelated widgets following widget of type 'other
* lisp/wid-edit.el (widget 'other): Use \n instead of the %n escape in the
:format property of this widget. If %n is used at the end of the
format string, unrelated widgets get indented. (Bug#12533)
* test/wid-edit-tests.el (widget-test-indentation-after-%n)
(widget-test-indentation-after-newline)
(widget-test-newline-and-indent-same-widget): New tests.
Michael Albinus [Thu, 26 Sep 2019 07:33:47 +0000 (09:33 +0200)]
Avoid errors in tramp-tests with a remote ksh shell
* test/lisp/net/tramp-tests.el (tramp--test-ksh-p): New defun.
(tramp-test41-utf8, tramp-test41-utf8-with-stat)
(tramp-test41-utf8-with-perl, tramp-test41-utf8-with-ls):
Skip, when remote shell is ksh.
* lisp/emacs-lisp/rx.el:
* test/lisp/emacs-lisp/rx-tests.el:
* doc/lispref/searching.texi (Rx Constructs):
Rewrite rx for correctness, clarity, and performance. The new
implementation retains full compatibility and has more comprehensive
tests.
* lisp/emacs-lisp/re-builder.el (reb-rx-font-lock-keywords):
Adapt to changes in internal variables in rx.el.
* src/search.c (Fregexp_quote): Only allocate a new string if needed.
* doc/lispref/searching.texi (Regexp Functions):
* etc/NEWS (Incompatible Lisp Changes): Document.
* lisp/wid-edit.el (widget--should-indent-p): New function, to decide
whether to indent or not.
(widget-checklist-value-add-item, widget-radio-add-item)
(widget-editable-list-format-handler)
(widget-editable-list-entry-create)
(widget-group-value-create): Use it (bug#7851).
Alan Mackenzie [Tue, 24 Sep 2019 17:04:12 +0000 (17:04 +0000)]
New edebug-spec, nested-backquote-form. This fixes bug #31090
* lisp/emacs-lisp/edebug.el: (nested-backquote-form): a new edebug-spec which
handles nested backquote structures without a , or ,@ "between" the outer and
inner backquotes.
(backquote-form): Use nested-backquote-form.
Pip Cet [Tue, 24 Sep 2019 16:35:15 +0000 (18:35 +0200)]
Allow a :stride argument so XBM boolvecs are in the right format.
* src/image.c (xbm_image_p): Explicitly specify the right stride if a
bool vector is used as argument.
* doc/lispref/display.texi (XBM Images): Describe bool vectors
accurately.
* etc/NEWS: Document the change (bug#36337).
Eli Zaretskii [Tue, 24 Sep 2019 07:28:44 +0000 (10:28 +0300)]
Resurrect support for negative frame geometry parameters on MS-Windows
* src/w32fns.c (my_create_window): Avoid assertion violations
in XFIXNUM when the 'top' or 'left' frame parameters are
neither fixnums nor 'unbound', in which case
f->size_hint_flags are set by gui_figure_window_size.
(Bug#37415)
Wilson Snyder [Mon, 23 Sep 2019 22:49:25 +0000 (18:49 -0400)]
Verilog-Mode collected updates.
* lisp/progmodes/verilog-mode.el (verilog-auto-inst-param): Add regexp
paramter to AUTOINSTPARAM to select which parameters to export. Reported
by Vish S.
(verilog-build-defun-re): Fix `verilog-goto-defun' to find automatic
function/tasks, bug1492. Reported by Enze Chi.
(verilog-preprocessor): Update default.
Paul Eggert [Mon, 23 Sep 2019 18:40:51 +0000 (11:40 -0700)]
Bring back mention of NNTP gmane
Problem reported by Michael Albinus at:
https://lists.gnu.org/r/emacs-devel/2019-09/msg00519.html
* admin/notes/emba: Reintroduce mention of nttp+news.gmane.org,
since that still works.
* lisp/epa.el (epa-verify-region): Signal an error if the region
couldn't be verified (bug#14720).
(epa-verify-file): Say that the file couldn't be verified instead
of just "...done".
Make "Show Hidden Headers" in Message buffers more obvious
* lisp/gnus/message.el (message-widen-and-recenter): New command
to ensure that we show most of the headers (bug#23252).
(message-mode-field-menu): Use it.
Hong Xu [Mon, 23 Sep 2019 11:09:36 +0000 (13:09 +0200)]
Add different faces for different citation levels in Message mode
* message.el (message-font-lock-keywords)
(message-font-lock-make-cited-text-matcher): Add support for
different faces for different citation levels. The faces are
defined in the faces named `message-cited-text-N': N of the
Mth citation level will be M mod 4.
(message-cited-text-1, message-cited-text-2)
(message-cited-text-3, message-cited-text-4): Add customization
for the faces of 4 different citation level. In the future, the
number of faces may increase, as the code is flexible enough to
automatically deal with that.
(message-cite-level-function): Add a function to customize the
determination of cite levels given the prefix of the cited text
(bug#25022).
* lisp/mail/sendmail.el (mail-citation-prefix-regexp): Don't
regard "]" as a citation prefix character (bug#25150). This
allows some motion commands (like forward-sexp) to behave more as
expected in Message mode buffers.
Paul Eggert [Mon, 23 Sep 2019 06:53:39 +0000 (23:53 -0700)]
Update some URLs
This mostly changes http: to https: in URLs. It also updates
some URLs that have moved, removes some URLs that no longer
work, recommends against using procmail (procmail.org no
longer works), and removes some mentions of the
no-longer-existing Gmane, LPF and VTW.
It doesn't update all URLs, just the ones I had time for.
* GNUmakefile (help):
* admin/admin.el (manual-doctype-string):
* admin/charsets/Makefile.in (${charsetdir}/ALTERNATIVNYJ.map):
* admin/charsets/mapconv:
* lisp/net/soap-client.el (soap-create-envelope):
* lisp/org/org.el (org-doi-server-url):
* lisp/textmodes/bibtex.el (bibtex-generate-url-list):
Prefer https: to http: un URLs.
Stephen Gildea [Mon, 23 Sep 2019 03:27:53 +0000 (20:27 -0700)]
Expand time-stamp unit tests to cover all formatting options
* time-stamp-tests.el: Expand unit tests to cover all formatting options.
These tests validate time-stamp-pattern formatting that has existed
since at least Emacs 22 (released in 2007). The tests cover both
documented behavior and behavior implemented to support future migrations.
* time-stamp.el (time-stamp-string): Add a second argument (TIME) to
open a testing seam. Have the unit tests call this public function.
* time-stamp.el (time-stamp-string, time-stamp-string-preprocess):
Remove the second pass through time-string--format. (Previously both
functions called it.) It was used only to handle "%", but this is now
handled by having time-stamp-string-preprocess not double it.
Not doubling the "%" in time-stamp-string-preprocess fixes the padding
of "%2%", which was discovered by the new unit tests to be wrong.
The regexp in this list used a capitalized “React” because it actually
should be capitalized like that. Otherwise, the following code would
produce a false positive match: import Thing from './react/Thing'
* lisp/progmodes/js.el (js-jsx-regexps): Update docstring.
(js-jsx--detect-and-enable): Match case-sensitively when determining
whether JSX should be enabled.
Paul Eggert [Sun, 22 Sep 2019 17:43:21 +0000 (10:43 -0700)]
Avoid crashes when casifying noncontiguous regions
This is a followon fix for Bug#37477.
* lisp/simple.el (region-extract-function):
Use setq here, since the var is now defined in C code.
* src/casefiddle.c (casify_pnc_region): New function.
(Fupcase_region, Fdowncase_region, Fcapitalize_region)
(Fupcase_initials_region): Use it.
(Fupcase_initials_region): Add region-noncontiguous-p flag
for consistency with the others. All uses changed.
(syms_of_casefiddle): Define Qbounds, Vregion_extract_function.
* src/insdel.c (prepare_to_modify_buffer_1):
* src/keyboard.c (command_loop_1):
Use Vregion_extraction_function.
* src/insdel.c (syms_of_insdel): No need to define
Qregion_extract_function.
* test/src/casefiddle-tests.el (casefiddle-oldfunc): New var.
(casefiddle-loopfunc, casefiddle-badfunc): New functions.
(casefiddle-invalid-region-extract-function): New test.
* lisp/image/exif.el (exif-parse-buffer): New function.
(exif-orientation): Ditto.
(exif-error): New error symbol, and adjust all error signalling to
only use that signal.
Eric Abrahamsen [Sun, 21 Apr 2019 18:04:20 +0000 (11:04 -0700)]
New option for making Gnus window layouts atomic
* lisp/gnus/gnus-win.el (gnus-use-atomic-windows): New boolean
customization option.
(gnus-configure-windows): When removing old window layouts, check
for and remove atomicity.
(gnus-configure-windows): When gnus-use-atomic-windows is non-nil,
make Gnus window layouts atomic.
* doc/misc/gnus.texi (Window Layout): Document.
Juri Linkov [Sat, 21 Sep 2019 23:00:01 +0000 (02:00 +0300)]
Support rectangular regions in capitalize-region and capitalize-dwim.
* lisp/simple.el (capitalize-dwim): Add arg region-noncontiguous-p
in capitalize-region call.
* src/casefiddle.c (Fcapitalize_region): Add arg region-noncontiguous-p.
If non-nil, operate on multiple chunks. (Bug#37477)
(Fdowncase_region): Use builtin symbol Qregion_extract_function
rather than calling intern.
Paul Eggert [Sat, 21 Sep 2019 18:27:46 +0000 (11:27 -0700)]
Revert too-picky file-access tests
Problem reported by Andreas Schwab (Bug#37475).
* doc/lispref/files.texi (Writing to Files)
(Testing Accessibility, Kinds of Files):
Document that accessibility and file-type predicates return nil
if there is trouble determining accessibility or type.
* etc/NEWS: Adjust, and list the affected primitives.
* src/callproc.c (init_callproc): Go back to Ffile_exists_p.
* src/fileio.c (PICKY_EACCES, file_test_errno):
Remove. All uses removed.
(Ffile_name_case_insensitive_p, Ffile_exists_p, Ffile_symlink_p)
(Ffile_directory_p, Ffile_regular_p): Document that these
functions return nil if there is trouble.
(Ffile_name_case_insensitive_p, check_file_access)
(Ffile_writable_p, Ffile_symlink_p, Ffile_directory_p)
(Ffile_accessible_directory_p, Ffile_regular_p)
* src/lread.c (Fload):
Go back to treating trouble in determining the answer as if the
file were missing.
* src/fileio.c (Ffile_newer_than_file_p): Use file_attribute_errno
not file_test_errno, since returning nil is not appropriate when
there are two files to test; e.g., in the rare cases where both
file timestamps have overflowed then neither t nor nil is correct.