333d6f4 (origin/emacs-26) More changes in the Emacs manual
52ca0d1 * lisp/vc/vc.el (vc-deduce-backend): Use ignore-errors.
69e8046 Don't signal error in vc-deduce-backend
f568c91 * doc/misc/tramp.texi: Index more user options.
92ca881 Minor change in the Emacs manual
e055a12 NEWS update about 'string-trim'
96b6e24 Clarify TRAMP process-environment interaction (Bug#30419)
4fa467e * lisp/progmodes/grep.el (grep-num-matches-found): New variable.
a22820a Avoid aborts in cm.c due to too small TTY frame
26f6441 Another set of improvements in the Emacs manual
cef3b42 Fix help in mode-line-mode-menu
f8a493c Improve documentation of desktop restoring
10637af Improve the Emacs manual's chapter "Frames"
d924953 Fix unbound mm-uu-entry in mm-uu
cb2487b Improve Emacs user manual in fixit.texi
66e9527 Another minor change in the manual
6a1c03d More minor changes in the Emacs manual
32fb8c4 Avoid printing garbled error message from image.el
Conflicts:
etc/NEWS
lisp/image.el
* src/nsterm.m (ns_set_represented_filename): Remove unused variable
'view'.
* src/nsfns.m (x_implicitly_set_name, syms_of_nsfns): The type of a
variable defined with DEFVAR_BOOL is 'bool', not 'Lisp_Object'.
* doc/lispref/processes.texi (Misc Network): Omit discussion of
internal--daemon-sockname, as non-Emacs code shouldn't rely on it.
* src/process.c (syms_of_process): Rename internal-daemon-sockname
to internal--daemon-sockname. All uses changed.
* lisp/server.el (server--external-socket-initialized): Rename
from server-external-socket-initialised, since it should be
private and Emacs uses American spelling. All uses changed.
* src/emacs.c, src/lisp.h: Revert previous changes, as the
initialization is now done in src/process.c, which already
includes the relevant files.
* src/process.c (union u_sockaddr): Move decl to top level.
(external_sock_name, Fget_external_sockname): Remove, replacing
with Vinternal__external_sockname. All uses changed.
(init_process_emacs): Deduce socket name ourselves rather than
have main.c do it. Use conv_sockaddr_to_lisp instead of doing
it by hand. Conditionalize it on HAVE_GETSOCKNAME.
* lisp/server.el: (server-external-socket-initialised): New
(server-name): Compute server name from `get-external-sockname'.
(server-socket-dir): Compute socket dir from
`get-external-sockname'.
(server-start): Don't check for existing server when an
uninitialised external socket has been passed to Emacs.
* src/emacs.c: (main): Obtain socket name via getsockname and pass
to `init_process_emacs'.
* src/lisp.h: (init_process_emacs): Add second parameter.
* src/process.c: (external_sock_name): New.
(get-external-sockname): New.
(init_process_emacs): Set `external_sock_name' to `sockname'
parameter.
* src/frame.c (frame_windows_min_size): Limit TTY frames to a
minimum height large enough to allow for a menu bar, the mode
line, one text line and one echo-area line. This avoids aborts in
cm.c:cmcheckmagic. (Bug#30320)
* src/nsterm.m (EV_MODIFIERS_OPTION):
(EV_MODIFIERS_CONTROL):
(EV_MODIFIERS_COMMAND): New macros.
(ev_modifiers_helper): New function.
(EV_MODIFIERS2): Use new helper function.
(EmacsView::keyDown): Remove duplicate functionality and handle fn key
correctly.
* lisp/term/ns-win.el (frame-title-format, icon-title-format): Change
default format to just the filename.
* src/nsfns.m (ns-use-proxy-icon): New variable.
(ns_set_name_as_filename): Remove function.
(x_implicitly_set_name): Get rid of special handling of
frame-title-format.
* src/nsterm.h (ns_set_represented_filename): Use new definition.
* src/nsterm.m (ns_set_represented_filename): Move some of the logic
from ns_set_name_as_filename into this function.
Latest Terminfo introduces terminal definitions that support direct
color mode. The "Co"/"colors" capability is set to 0x1000000 on these
terminals and Emacs is already compatible with them.
However, if used Terminfo library hasn't been compiled with 32-bit
value support, "Co"/"colors" is truncated to 0x7fff. In this case
direct color mode support can be detected from the "RGB" capability
flag.
There are some minor problems if the color count isn't corrected from
0x7fff. First eight standard colors defined in xterm-standard-colors
are shown correctly. However, their RGB values match the terminal
settings, not the RGB values defined in xterm-standard-colors. Bright
versions of these colors are shown incorrectly. They are interpreted
as pixels #000008 - #000015, which are very dark shades of blue.
* src/term.c (init_tty): Force terminal color count to 0x1000000 if
"RGB" capability is present.
* src/tparam.h: Define prototype for tigetflag. (Bug#30308)
* doc/misc/efaq.texi (Colors on a TTY): Add information about direct
mode terminals supported by Terminfo.
Move calculation of the mode line default help echo from
note_mode_line_or_margin_highlight to display_mode_lines
(Bug#30182). Fix cursor type for dragging the mode line.
Normalize FRAME argument of Fx_show_tip before assigning it to
tip_last_frame and handle the transition from GTK+ to Emacs
tooltips and vice-versa in x_hide_tip (Bug#30399).
* src/window.h (struct window): New Lisp member
mode_line_help_echo.
(wset_mode_line_help_echo): New function.
* src/w32fns.c (Fx_show_tip): Normalize the FRAME argument
bevore storing it in tip_last_frame (Bug#30399).
* src/xdisp.c (display_mode_lines): Calculate mode line
default help echo string here and store it in the window's
mode_line_help_echo slot (Bug#30182).
(note_mode_line_or_margin_highlight): Use value in window's
mode_line_help_echo slot as mode line default help echo. When
the window is resizable show a vertical drag cursor instead of
the vertical scroll bar cursor.
* src/xfns.c (x_hide_tip): Rewrite the GTK+ part to correctly
handle the transition from GTK+ system to Emacs tooltips and
vice-versa (Bug#30399).
(Fx_show_tip): Normalize the FRAME argument bevore storing it
in tip_last_frame (Bug#30399).
0276743 (origin/emacs-26) Doc string fix in latin-alt.el
c9269af Minor improvements in package.texi
898a3e4 Minor improvement in Emacs user manual
8f7d718 * doc/lispref/anti.texi (Antinews): Replace 25.2 with 25.3.
989d0af Improve documentation of mini-window resizing
923e415 More changes in the Emacs manual
565adf2 Clarify that text-quoting-style doesn't affect *Info* (Bug#24...
2dd273b Mention that shell quoting of % on w32 may fail (Bug#19350)
19fa6d5 Fix --no-build-details for w32 and macOS
979944c * doc/lispref/anti.texi (Antinews): Add some more antinews.
c2727e3 Fix crashes when run with --no-build-details
dc08490 More changes for the manual
04c5bd5 Doc fixes for process functions (Bug#30349)
e209034 Fix shr and CSS 4 color maps
5fe8192 Yet another round of improvements in the manual
Conflicts:
doc/emacs/files.texi
test/lisp/subr-tests.el
* src/nsterm.m (keyDown:): Distinguish between shift-like and
control-like modifier keys. Allow treating ⌘ as shift-like
modifier (e.g. for the Gujarati – QUERTY input method, where ⌘
switches to QUERTY.)
* lisp/cus-start.el (standard): Change nil to none for
ns-command-modifier; update description.
* etc/NEWS: Add NEWS entry.
This fixes a bug where (file-directory-p FOO) would fail if FOO
had an inode number out of range for ‘stat’.
* src/fileio.c (file_directory_p): Accept a Lisp string instead of
a C string. All callers changed. On non-MS-Windows hosts, use
openat with O_PATH|O_DIRECTORY if available, otherwise
file_accessible_directory_p unless it fails due to EACCESS,
otherwise stat.
* src/lread.c (Fload): Don't defer to module-load immediately when
encountering a module, but use the normal loading machinery to
properly set up load-history, check for recursive loads, print
messages, etc.
* test/src/emacs-module-tests.el (module/load-history): New test.
(module/describe-function-1): Adapt test.
* etc/NEWS: Mention fixed behavior.
22922c7 (origin/emacs-26) * doc/emacs/entering.texi (Entering Emacs):...
59657c4 Document 'window-at-side-p' in the Elisp manual
2b35ed0 Document external-debugging-output in the Elisp Manual (Bug#2...
db6564c Fix scroll-margin docstring (Bug#13791)
732d1b9 Clarify that `ansi-term' is almost the same as `term' (Bug#18...
f706c59 Update manual description of locate-file (Bug#23650)
1602262 Clarify effect of print-gensym (Bug#27776)
53e9fa2 * lisp/custom.el (defcustom): Fix docstring (Bug#27891).
607cc4e Define cl-type-definition button type as needed (Bug#28899)
9e6889c Emphasize that GPG passphrase caching is temporary (Bug#29907)
4b5711b Fix @examples in cc-mode.info, where lines were getting glued...
71961f1 Minor change in "Mode Line" section of Emacs manual
55a2b76 More fixes in the Emacs manual
6415b2d Allow read-passwd to hide characters inserted by C-y. (Secur...
8cb4ffb * etc/PROBLEMS: Document issues with double-buffering. (Bug#...
fd10070 * lisp/window.el (window-largest-empty-rectangle): Fix grammar.
e1a4403 Minor changes in the Emacs manual
372fda6 Improve the "Files" chapter of the Emacs manual
9afc86d Improve the "Search" chapter of the Emacs manual
22f98b3 * src/indent.c (Findent_to): Doc fix. (Bug#30260)
085ee43 Improve documentation of 'edebug-defun'
4dd1b33 Mention crashes due to Noto Serif Kannada fonts
That commit did not make the corresponding change to printing, thus
breaking the (eq (read (prin1-to-string SYM)) SYM) invariant for those
symbols. It's too late in the release cycle to change printing
behavior, therefore revert the reader change.
Don't merge to master, the print function will be updated there (see
"Fix round tripping of read->print for symbols with strange quotes").
Since 2017-07-22 "Signal error for symbol names with strange
quotes (Bug#2967)", symbol names beginning with certain quote
characters require an escaping backslash. However, the corresponding
change for printing missed, so that (eq (read (prin1-to-string SYM))
SYM) does not give `t' for such symbols.
* src/character.c (confusable_symbol_character_p): New function,
extracted from test `read1'.
* src/lread.c (read1): Use it.
* src/print.c (print_object): Use it to print a backslash for symbols
starting with characters that `read1' requires to be escaped.
* test/src/print-tests.el (print-read-roundtrip): New test.
* etc/NEWS.26:
* etc/NEWS: Clarify the announcement for the earlier reader
change (Bug#30217).
* src/regex.h (RE_DUP_MAX): Raise limit to 2^16-1.
* etc/NEWS: Announce it.
* doc/lispref/searching.texi (Regexp Backslash): Document it.
* test/src/regex-tests.el (regex-repeat-limit): Test it.
* src/regex.h (reg_errcode_t): Add REG_ESIZEBR code.
* src/regex.c (re_error_msgid): Add corresponding entry.
(GET_INTERVAL_COUNT): Return it instead of the more generic REG_EBADBR
when encountering a repetition greater than RE_DUP_MAX.
* lisp/isearch.el (isearch-search): Don't convert errors starting with
"Invalid" into "incomplete". Such errors are not incomplete, in the
sense that they cannot be corrected by appending more characters to
the end of the regexp. The affected error messages are:
- REG_BADPAT "Invalid regular expression"
- \\(?X:\\) where X is not a legal group number
- \\_X where X is not < or >
- REG_ECOLLATE "Invalid collation character"
- There is no code to throw this.
- REG_ECTYPE "Invalid character class name"
- [[:foo:] where foo is not a valid class name
- REG_ESUBREG "Invalid back reference"
- \N where N is referenced before matching group N
- REG_BADBR "Invalid content of \\{\\}"
- \\{N,M\\} where N < 0, M < N, M or N larger than max
- \\{NX where X is not a digit or backslash
- \\{N\\X where X is not a }
- REG_ERANGE "Invalid range end"
- There is no code to throw this.
- REG_BADRPT "Invalid preceding regular expression"
- We never throw this. It would usually indicate a "*" with no
preceding regexp text, but Emacs allows that to match a literal
"*".
463f96b481 * doc/lispref/searching.texi: Document regexp repetition l...
08a6195571 ; test/README: Document TEST_LOAD_EL parameter.
7bbea90b1a * src/syntax.c (char-syntax): Warn about ignoring text pro...
50fcbb5f61 ; * src/process.c (Fprocess_contact): Fix docstring typo.
81ae9c8c05 Load mm-util as needed for url-file and url-data (Bug#30258)
5a1ee67ae1 Another minor copyedit in the manual's "Scroll Bars"
226a651e9e Minor fix in documentation of 'equal'
b26786c8d9 * lisp/dired-x.el (dired-guess-shell-alist-user): Doc fix....
5699a824f0 Minor rewording in Emacs manual's "Help Mode" node
f35ff0156e Fixes for Emacs manual in frames.texi
6cd4e8dcc5 * doc/misc/cl.texi (Efficiency Concerns): Fix 2012-10-27 t...
1412cf3edd Fix a few issues with latest GTK scaling changes
59db8dca03 Use scaled coordinates when calling into GTK
2892f05792 Scale monitor dimensions obtained from GTK
This is part two of a two part fix for the GTK scaling
problems. See the thread starting at
http://lists.gnu.org/archive/html/emacs-devel/2018-01/msg00372.html
for an explanation of why it has been added to Emacs 26.
* src/gtkutil.c (xg_set_geometry): Scale down the coordinates that we
pass to gtk_window_move and to gtk_window_parse_geometry.
* src/xterm.c (x_set_offset): Likewise.
This is part one of a two part fix for the GTK scaling
problems. See the thread starting at
http://lists.gnu.org/archive/html/emacs-devel/2018-01/msg00372.html
for an explanation of why it has been added to Emacs 26.
* src/xfns.c (Fx_display_monitor_attributes_list): Take scaling factor
into account when using GTK to query the monitor dimensions, which
allows frameset-restore to position frames correctly.
