Paul Eggert [Tue, 2 Apr 2019 07:17:37 +0000 (00:17 -0700)]
More regexp advice and clarifications
* doc/lispref/searching.texi (Regexp Special): Simplify style
advice for order of ], ^, and - in character alternatives.
Stick with saying that it’s not a good idea to put ‘-’ after a
range. Remove the special case about raw 8-bit bytes and
unibyte characters, as this documentation is confusing and
seems to be incorrect in some cases. Say that z-a is the
preferred style for reversed ranges, since it’s clearer and is
typically what’s used in practice. Mention some bad styles:
duplicates in character alternatives, ranges that denote <=3
characters, and ‘-’ as the first character.
Paul Eggert [Mon, 1 Apr 2019 18:54:23 +0000 (11:54 -0700)]
Make struct Lisp_Objfwd etc. objects read-only
Initialize these objects statically, and make them constants.
This is a bit safer and more efficient.
* src/data.c (XBOOLFWD, XKBOARD_OBJFWD, XFIXNUMFWD, XOBJFWD):
* src/lisp.h (XBUFFER_OBJFWD):
Return a pointer-to-const instead of an unrestricted pointer.
(lispfwd): fwdptr is now a pointer-to-const instead of an
unrestricted pointer. All uses changed.
(SET_SYMBOL_FWD): Accept pointer-to-const instead of an
unrestricted pointer.
(DEFVAR_LISP, DEFVAR_LISP_NOPRO, DEFVAR_BOOL, DEFVAR_INT)
(DEFVAR_KBOARD): Initialize static structures statically
instead of dynamically, and make them const.
* src/lread.c (defvar_int, defvar_bool, defvar_lisp_nopro)
(defvar_lisp, defvar_kboard): Accept pointer-to-const instead
of an unrestricted pointer; it’s now the caller’s
responsibility to initialize the pointed-to storage. No need
for a separate address argument any more. All callers
changed.
Paul Eggert [Mon, 1 Apr 2019 18:54:23 +0000 (11:54 -0700)]
Fix union Lisp_Fwd * alignment bug
It's not portable to cast (e.g.) struct Lisp_Objfwd * to union
Lisp_Fwd * and then back again, because the compiler can then assume
that the pointer is aligned for union Lisp_Fwd * when accessing
the struct Lisp_Objfwd * components, and this assumption might
be incorrect becase we don't force that alignment.
* src/lisp.h (lispfwd): New type, replacing ...
(union Lisp_Fwd): ... this type, which was removed.
All uses changed.
(SET_SYMBOL_FWD): 2nd arg is now void *, not lispfwd.
All uses changed (casts no longer needed; they were
not portable anyway).
Paul Eggert [Mon, 1 Apr 2019 03:00:30 +0000 (20:00 -0700)]
Use update-game-score more often with Tetris
* lisp/play/gamegrid.el (gamegrid-add-score-with-update-game-score):
Go back to previous way of deciding whether to invoke
update-game-score, except do not try to invoke it if file-modes
fails on it (Bug#35056).
* lisp/textmodes/text-mode.el (text-mode): Do not reset
indent-line-function to its global default value of indent-relative.
* doc/lispref/modes.texi (Example Major Modes):
* etc/NEWS: Document change accordingly.
For discussion, see thread starting at:
https://lists.gnu.org/archive/html/emacs-devel/2019-03/msg01012.html
* lisp/textmodes/text-mode.el (text-mode): Do not reset
indent-line-function to its global default value of indent-relative.
* doc/lispref/modes.texi (Example Major Modes):
* etc/NEWS: Document change accordingly.
* lisp/textmodes/text-mode.el: Use lexical-binding.
(text-mode, paragraph-indent-minor-mode, text-mode-hook-identify):
Use setq-local.
(toggle-text-mode-auto-fill): Quote function symbols as such.
(center-line): Minor simplification.
* doc/lispref/modes.texi (Example Major Modes): Update code example
for these changes to text-mode.
* doc/lispref/text.texi (User-Level Deletion): Document new optional
arguments of delete-indentation.
* lisp/simple.el (delete-indentation): Do not barf if called
interactively when region is inactive. (bug#35021)
Do not skip blank lines. (bug#35036)
Consistently deactivate mark even when no text was changed.
Handle active region spanning a single line.
* test/lisp/simple-tests.el (simple-test--buffer-substrings):
New convenience function.
(simple-test--dummy-buffer, simple-test--transpositions): Use it.
(simple-delete-indentation-no-region)
(simple-delete-indentation-inactive-region): Update commentary.
Call delete-indentation interactively when testing for behavior with
inactive region and region is not explicitly defined.
(simple-delete-indentation-blank-line)
(simple-delete-indentation-boundaries)
(simple-delete-indentation-region)
(simple-delete-indentation-prefix): New tests.
Alan Mackenzie [Sat, 30 Mar 2019 13:19:47 +0000 (13:19 +0000)]
Allow a CC Mode derived mode to have strings delimited by single quotes.
Also fix the bug where the delimiters of '\033', etc. got the error face.
* lisp/progmodes/cc-langs.el (c-single-quotes-quote-strings): Enhance the docr
string.
(c-string-delims): Change doc string to doc comment.
* listp/progmodes/cc-mode.el (c-before-change-check-unbalanced-strings): In
searches and comparisons, take account of the string delimiters possibly
being '. Fix argument in call of c-before-change-check-unbalanced-strings.
(c-parse-quotes-before-change, c-parse-quotes-after-change): Bind
case-fold-search to nil. Analyze escape constructs inside character constants
more accurately, in particular accepting as valid more than one character
after /[0-7], /x, /u, and /U. Amend calculations to account for this extra
length.
* src/search.c (looking_at_1, search_buffer_re): Unfreeze the regexp
buffer before signalling a matcher overflow, since the error
processing may require quite some regexp use as well (Bug#34910).
* src/search.c (compile_pattern): Don't give up if the last regexp
cache entry is busy. Instead, use the last (least recently used)
non-busy entry, and only signal a reentrancy error if there is no free
entry at all (Bug#34910).
Eli Zaretskii [Sat, 30 Mar 2019 09:01:58 +0000 (12:01 +0300)]
Don't run buffer-related hooks in " *code conversion work*" buffers
Note: portions of this change were mistakenly pushed as part
of an unrelated commit a35a1f6a9.
* src/buffer.c (Fget_buffer_create): Set inhibit_buffer_hooks
non-zero for temporary buffers created by coding.c. Don't run
buffer-list-update-hook for such buffers.
(Frename_buffer, Fkill_buffer, record_buffer)
(Fbury_buffer_internal): Don't run hooks for buffers whose
inhibit_buffer_hooks flag is set.
* src/buffer.h (struct buffer): New member
inhibit_buffer_hooks.
* src/pdumper.c (dump_buffer): Dump the new field. Update the
hash value in HASH_buffer_XXX.
* src/coding.c (make_conversion_work_buffer): Function deleted;
code moved to code_conversion_save.
(code_conversion_save): Insert code from
make_conversion_work_buffer, but arrange for unwind-protecting
the current buffer before switching to the work buffer. This
avoids leaving reused_workbuf_in_use set if user presses C-g
during encoding/decoding.
(Vcode_conversion_workbuf_name): Now external variable.
* src/coding.h (Vcode_conversion_reused_workbuf): Declare.
Stefan Monnier [Fri, 29 Mar 2019 12:46:29 +0000 (08:46 -0400)]
Gnus: Automatically render text/calendar in a human-friendly way
* lisp/gnus/mm-decode.el (mm-inline-media-tests): Add text/calendar entry.
Use (fboundp 'device-sound-enabled-p) rather than fishing for features.
(mm-automatic-display): Add text/calendar entry.
* lisp/emacs-lisp/pcase.el (pcase--u1):
Use memql instead of memq to work with bignums (Bug#34781).
* test/lisp/emacs-lisp/pcase-tests.el (pcase-tests-member): Test the above.
Eli Zaretskii [Thu, 28 Mar 2019 16:00:06 +0000 (18:00 +0200)]
Fix display of sliced images on MS-Windows
* src/w32term.c (x_draw_image_foreground): Fix detection of
scaled images for sliced images. Scale the original width of
a slice and its coordinates of origin as well.
Paul Eggert [Thu, 28 Mar 2019 04:51:39 +0000 (21:51 -0700)]
Simpler way to export HAVE_X_WINDOWS to GDB
* src/.gdbinit: Simplify by removing dependency on globals
implementation. This is useful for a future performance
improvement that I have in mind.
* src/alloc.c (enum defined_HAVE_X_WINDOWS, defined_HAVE_X_WINDOWS):
New enum and constant.
(gdb_make_enums_visible) [__GNUC__]: Use it, to make
defined_HAVE_X_WINDOWS visible to GDB.
Paul Eggert [Thu, 28 Mar 2019 04:03:10 +0000 (21:03 -0700)]
Tweak re_registers allocation
* src/regex-emacs.c (re_match_2_internal):
No need to allocate one extra trailing search register;
Emacs does not use it. Avoid quadratic behavior on
reallocation.
Eli Zaretskii [Wed, 27 Mar 2019 18:34:22 +0000 (20:34 +0200)]
Attempt to fix crashes under GDB on Windows 10
* src/pdumper.c (dump_discard_mem)
[VM_SUPPORTED == VM_MS_WINDOWS]: Don't pass NULL pointer as
last argument to VirtualProtect. Reported by Martin Rudalics
<rudalics@gmx.at>.
Paul Eggert [Wed, 27 Mar 2019 17:01:21 +0000 (10:01 -0700)]
Minor xml.el tweaks
* lisp/xml.el (xml-name-start-char-re, xml-name-char-re):
No need to call concat.
(xml-name-char-re): Use \u escapes rather than chars inline,
so that the code matches the comments better and is easier
to audit.
Alan Mackenzie [Wed, 27 Mar 2019 11:50:53 +0000 (11:50 +0000)]
Improve C++ raw string fontification.
Integrate the handling of raw string and ordinary string fontification.
* lisp/progmodes/cc-defs.el (c-font-lock-flush)
(c-search-forward-char-property-without-value-on-char): new macros.
(c-point): In the 'eoll arm, check for eobp.
(c-search-forward-char-property-with-value-on-char): Handle the &optional
limit argument being nil.
(c-clear-char-property-with-value-on-char-function)
(c-clear-char-property-with-value-on-char): Return the position of the first
cleared property.
* lisp/progmodes/cc-engine.el (c-find-decl-prefix-search): Don't spuriously
recognize the change of face at a ) as the start of a string (a
"pseudo match").
(c-old-beg-rs c-old-end-rs): New variables.
(c-raw-string-pos): Analyze raw string delimiters more carefully.
(c-raw-string-in-end-delim): New function.
(c-depropertize-raw-string): Largely rewritten.
(c-before-change-check-raw-strings): New functionality: only remove the
syntax-table text properties from raw strings whose delimiters are about to
change.
(c-propertize-raw-string-id): New function.
(c-after-change-re-mark-raw-strings): Remove, incorporating functionality into
other functions.
(c-propertize-raw-string-opener): Largely rewritten.
(c-after-change-re-mark-raw-strings): Removed.
(c-after-change-unmark-raw-strings, c-after-change-unmark-raw-strings): New
functions.
* lisp/progmodes/cc-langs.el (c-before-font-lock-functions): Replace
c-after-change-re-mark-unbalanced-strings by
c-after-change-mark-abnormal-strings in the t, c+objc, c++ and java sections.
Add c-after-change-unmark-raw-strings and remove
c-after-change-re-mark-raw-strings from the c++ section.
* lisp/progmodes/cc-mode.el (c-old-BEG c-old-END): Remove.
(c-old-END-literality): New variable.
(c-depropertize-CPP): Remove syntax-table properties from raw strings within
macros.
(c-before-change-check-unbalanced-strings): Call
c-truncate-semi-nonlit-pos-cache to preserve the integrity of the cache.
(c-before-change-check-unbalanced-strings): Call
c-truncate-semi-nonlit-pos-cache, largely rewritten.
(c-after-change-re-mark-unbalanced-strings): Renamed to
c-after-change-mark-abnormal-strings. Call c-maybe-re-mark-raw-string.
* src/xterm.c (x_cr_draw_image): Add arguments image_width and
image_height and support scaling. All callers changed.
* src/image.c (Fimage_scaling_p): Return t when USE_CAIRO.
(x_set_image_size) [USE_CAIRO]: Record the scaled dimensions
in the image struct.
* src/dispextern.h (HAVE_NATIVE_SCALING): Define when
USE_CAIRO as well.
* etc/NEWS: Update the announcement of native image scaling.
Paul Eggert [Wed, 27 Mar 2019 02:06:36 +0000 (19:06 -0700)]
2019-03-26 regex cleanup
Problems reported by Mattias Engdegård in:
https://lists.gnu.org/r/emacs-devel/2019-03/msg01028.html
* lisp/align.el (align-rules-list):
* lisp/speedbar.el (speedbar-check-read-only, speedbar-check-vc):
* lisp/vc/diff-mode.el (diff-add-change-log-entries-other-window):
* lisp/woman.el (woman-parse-numeric-arg):
Put "-" at end of character alternatives, since a range was not intended.
* lisp/erc/erc.el (font-lock):
* lisp/mail/footnote.el (cl-seq):
Avoid duplicate character alternatives by using cl-seq API.
* lisp/mail/footnote.el (footnote--current-regexp):
* lisp/textmodes/css-mode.el (css--font-lock-keywords):
Avoid repetition of repetition.
* lisp/net/webjump.el (webjump-url-encode):
Add ~ to character alternatives, and rewrite confusing range.
* lisp/progmodes/verilog-mode.el (verilog-compiler-directives)
(verilog-assignment-operator-re):
Remove duplicate.
* lisp/progmodes/verilog-mode.el (verilog-preprocessor-re):
* lisp/textmodes/css-mode.el (css--font-lock-keywords):
Don’t escape a char that doesn’t need it.
* lisp/textmodes/picture.el (picture-tab-chars): In docstring,
do not say regexp characters will be quoted; merely say in
another way that the syntax is that of character alternatives.
(picture-set-tab-stops, picture-tab-search): Don’t attempt
to regexp-quote picture-tab-chars.
(picture-tab-search): Quote \ in picture-tab-chars for
skip-chars-backwards, which treats \ differently than
regexp character alternatives do.
Support tool bar icon image on GTK+ >= 3.10 with cairo
* src/gtkutil.c (xg_get_image_for_pixmap) [USE_CAIRO]: Use cairo image
surface for GtkImage source.
(xg_tool_item_stale_p, update_frame_tool_bar) [USE_CAIRO]: Use cairo
image surface instead of pixmap for data associated with tool bar item.
Simplify cairo image surface creation and destruction
* src/dispextern.h (struct image) [USE_CAIRO]: Remove member cr_data2.
* src/image.c (set_cairo_image_surface): New function split from
original create_cairo_image_surface. Call cairo_surface_mark_dirty.
(create_cairo_image_surface): Just create image surface and return it.
(x_clear_image): Don't free cr_data2.
(xpm_load, pbm_load, png_load_body, jpeg_load_body, tiff_load)
(gif_load, imagemagick_load_image, svg_load_image) [USE_CAIRO]: Use
new create_cairo_image_surface and cairo_image_surface_get_data
instead of xmalloc. Use set_cairo_image_surface instead of old
create_cairo_image_surface.
(pbm_load) [USE_CAIRO]: Call cairo_surface_destroy for surface instead
of xfree for data.
(gif_load) [USE_CAIRO]: Multiply y-coordinate value by width instead
of subimg_width.
Paul Eggert [Mon, 25 Mar 2019 22:20:20 +0000 (15:20 -0700)]
Port recent ngettext stub to non-glibc
* src/editfns.c: Don’t try to call glibc ngettext;
we’re not ready for that yet.
(Fngettext): Do not restrict integer arguments to fixnums.
Improve doc string a bit.
Paul Eggert [Mon, 25 Mar 2019 21:53:56 +0000 (14:53 -0700)]
Update from Gnulib.
This incorporates:
2019-03-23 Support cross-compilation to musl libc
2019-03-23 noreturn: In C++ mode with clang, use _Noreturn as fallback
2019-03-22 _Noreturn: beware of C's _Noreturn in C++ pre C++11
2019-03-19 Help making signal handlers more reliable
2019-03-18 _Noreturn: clang and MSVC do support [[noreturn]] in C++11
2019-03-17 _Noreturn: GCC 4.7 does not support [[noreturn]] in C++11
2019-03-14 all: Update URLs to msdn.microsoft.com
* doc/misc/texinfo.tex, lib/_Noreturn.h, lib/gettimeofday.c:
* lib/mktime.c, lib/regcomp.c, lib/regexec.c, lib/stat-time.h:
* lib/utimens.c, m4/fdopendir.m4, m4/getgroups.m4:
* m4/gettimeofday.m4, m4/gnulib-common.m4, m4/putenv.m4, m4/utimes.m4:
Update from gnulib.
Paul Eggert [Mon, 25 Mar 2019 21:31:31 +0000 (14:31 -0700)]
Fix alignment bug with pure bignums
Problem found on 32-bit sparc, which has stricter alignment
checking than x86-64.
* src/alloc.c (pure_alloc): When TYPE is negative it now specifies
the negation of the required alignment of the result.
(make_pure_bignum): Specify bignum limb alignment.
Eric Abrahamsen [Mon, 25 Mar 2019 20:27:07 +0000 (13:27 -0700)]
Fix text property bug in gnus-group-list-active
* lisp/gnus/gnus-group.el (gnus-group-list-active): The property value
should be the group name, not the value of gethash. Ie, it should be
the key, not the value.
Paul Eggert [Mon, 25 Mar 2019 15:47:57 +0000 (08:47 -0700)]
Fix some integer issues in regex-emacs
Also, remove some duplicate comments related to thread.h.
* src/regex-emacs.h (struct re_registers):
* src/regex-emacs.c (SIGN_EXTEND_CHAR): Remove.
(TALLOC, RETALLOC): Remove. All uses replaced by usual
allocators, which check for integer overflow.
(extract_number): Redo without using ‘unsigned’.
(CHARSET_RANGE_TABLE_EXISTS_P): Clearly return a boolean.
(print_fastmap, print_partial_compiled_pattern, CHECK_INFINITE_LOOP)
(regex_compile, analyze_first, bcmp_translate, mutually_exclusive_p)
(re_match_2_internal):
Use bool for booleans.
(print_fastmap, regex_compile, execute_charset):
Prefer int to unsigned where either will do.
(print_double_string): Prefer ptrdiff_t to ssize_t, since the
latter can in theory be narrower than the former. Use fwrite
instead of repeated putchar.
(emacs_re_max_failures, fail_stack_type, compile_stack_type)
(re_wctype_parse, regex_compile, re_search, re_search_2)
(re_match_2, re_match_2_internal, re_compile_pattern):
Prefer ptrdiff_t to size_t where either will do.
(union fail_stack_elt, PUSH_FAILURE_REG, POP_FAILURE_REG_OR_COUNT):
Make the integer an intptr_t, not long.
(GET_BUFFER_SPACE, EXTEND_BUFFER, regex_compile):
Use xpalloc to simplify allocation.
(regex_compile): Check for integer overflow when calculating
register numbers.
* src/regex-emacs.c (re_set_registers, re_match_2_internal):
* src/regex-emacs.h (struct re_registers, struct re_pattern_buffer):
* src/search.c (Freplace_match):
Prefer ptrdiff_t to unsigned where either will do.
* src/regex-emacs.h (struct re_pattern_buffer):
Prefer bool_bf to unsigned where either will do.
João Távora [Mon, 25 Mar 2019 14:50:20 +0000 (14:50 +0000)]
Avoid occasional confusion of Flymake C/C++ backend
The regexp would sometimes match non-errors and create a diagnostic
with nil severity, which would have consequences for the Flymake
engine (these should eventually be addressed, too).
* lisp/gnus/gnus-sum.el (gnus-summary-mode-group): Declare before
first use to silence byte-compiler.
(gnus-summary-select-article): Simplify boolean expression.
(gnus-summary-move-article): Do not try to unsuppress article when
duplicate suppression is disabled. (bug#34973, bug#34974)
Stephen Leake [Fri, 22 Mar 2019 23:14:50 +0000 (16:14 -0700)]
Fix bug in delete-indentation when region is inactive
* test/lisp/simple-tests.el: Add tests for delete-indentation.
(simple-delete-indentation-no-region): Works with no region.
(simple-delete-indentation-inactive-region): Was broken with inactive
region; now fixed.
* lisp/simple.el (delete-indentation): Check (use-region-p) before using BEG.
Eric Abrahamsen [Thu, 26 Apr 2018 23:26:27 +0000 (16:26 -0700)]
Change Gnus hash tables into real hash tables
Gnus has used obarrays as makeshift hash tables for groups: group
names are coerced to unibyte and interned in custom obarrays, and
their symbol-value set to whatever value needs to be stored. This
patch replaces those obarrays with actual hash tables.
* lisp/gnus/gnus-util.el (gnus-intern-safe, gnus-create-hash-size):
Remove functions.
(gnus-make-hashtable): Change to return a real hash table.
(gnus-text-property-search): Utility similar to `text-property-any',
but compares on `equal'. Needed because the 'gnus-group text
property is now a string.
* lisp/gnus/gnus.el (gnus-gethash, gnus-gethash-safe, gnus-sethash):
Remove macros.
(gnus-group-list): New variable holding all group names as an
ordered list. Used because `gnus-newsrc-hashtb' used to preserve
`gnus-newsrc-alist' ordering, but now doesn't.
* lisp/gnus/nnmaildir.el (nnmaildir--servers): Change from obarray to
alist.
(nnmaildir--up2-1): Remove function.
* lisp/thingatpt.el (thing-at-point-newsgroup-p): This was making use
of Gnus obarrays, replace with a cond that can handle many different
possibilities.
* lisp/gnus/gnus-bcklg.el (gnus-backlog-articles): Remove
gnus-backlog-hashtb, which wasn't doing anything. Just keep a list
of ident strings in gnus-backlog-articles.
(gnus-backlog-setup): Delete unnecessary function.
(gnus-backlog-enter-article, gnus-backlog-remove-oldest-article,
gnus-backlog-remove-article, gnus-backlog-request-article): Alter
calls accordingly.
* lisp/gnus/gnus-dup.el (gnus-duplicate-list-max-length): Rename from
`gnus-duplicate-list-length', for accuracy.
* lisp/gnus/gnus-start.el (gnus-active-to-gnus-format,
gnus-groups-to-gnus-format, gnus-newsrc-to-gnus-format): Read group
names as strings.
(gnus-gnus-to-quick-newsrc-format): Write `gnus-newsrc-alist' using
the ordering in `gnus-group-list'.
* lisp/gnus/gnus-agent.el:
* lisp/gnus/gnus-async.el:
* lisp/gnus/gnus-cache.el:
* lisp/gnus/gnus-group.el:
* lisp/gnus/gnus-score.el:
* lisp/gnus/gnus-sum.el:
* lisp/gnus/gnus-topic.el:
* lisp/gnus/message.el:
* lisp/gnus/mml.el:
* lisp/gnus/nnagent.el:
* lisp/gnus/nnbabyl.el:
* lisp/gnus/nnvirtual.el:
* lisp/gnus/nnweb.el: In all files, change obarrays to hash-tables,
and swap `gnus-sethash' for `puthash', `gnus-gethash' for `gethash',
`mapatoms' for `maphash', etc.
* test/lisp/gnus/gnus-test-headers.el (gnus-headers-make-dependency-table,
gnus-headers-loop-dependencies): New tests to make sure we're
building `gnus-newsgroup-dependencies' correctly.
Michael Albinus [Fri, 22 Mar 2019 13:38:06 +0000 (14:38 +0100)]
Fix Bug#24394, Bug#34172
* lisp/subr.el (process-file-shell-command):
Use `with-connection-local-variables'. Do not set "/bin/sh" for
remote buffers, trust settings of `shell-file-name'.
* lisp/net/tramp-sh.el (tramp-sh-handle-file-notify-add-watch)
(tramp-maybe-open-connection): Use proper read syntax for function
names.
* lisp/net/tramp.el (tramp-handle-shell-command): Do not use shell
file names from `tramp-methods'.
Respect `async-shell-command-buffer'. (Bug#24394, Bug#34172)
Use `start-file-process-shell-command' and `process-file-shell-command'.
* test/lisp/net/tramp-tests.el (tramp-test32-shell-command):
Let it run partly for tramp-adb.
Nicolas Petton [Fri, 22 Mar 2019 09:06:42 +0000 (10:06 +0100)]
Fix the error message when modifying maps in-place (Bug#34941)
* lisp/emacs-lisp/map.el (map-not-inplace): Fix the message of the
error.
(map-put!): When signaling map-no-inplace error, output the map that
cannot be updated inplace.
projects / emacs.git / log