Paul Eggert [Mon, 4 Mar 2019 08:00:39 +0000 (00:00 -0800)]
Simplify list creation in C code
The main new thing here is that C code can now say
‘list (a, b, c, d, e, f)’ instead of
‘listn (CONSTYPE_HEAP, 6, a, b, c, d, e, f)’,
thus relieving callers of the responsibility of counting
arguments (plus, the code feels more like Lisp). The old
list1 ... list5 functions remain, as they’re probably a bit
faster for small lists.
* src/alloc.c (cons_listn, pure_listn): New functions.
(listn): Omit enum argument.
All callers changed to use either new ‘list’ or ‘pure_list’ macros.
* src/charset.c (Fdefine_charset_internal):
* src/coding.c (detect_coding_system)
(Fset_terminal_coding_system_internal):
* src/frame.c (frame_size_history_add, adjust_frame_size):
* src/gtkutil.c (xg_frame_set_char_size):
* src/keyboard.c (command_loop_1):
* src/nsfns.m (frame_geometry):
* src/widget.c (set_frame_size):
* src/xfaces.c (Fcolor_distance):
* src/xfns.c (frame_geometry):
* src/xterm.c (x_set_window_size_1):
* src/xwidget.c (Fxwidget_size_request):
Prefer list1i, list2i, etc. to open-coding them.
* src/charset.c (Fset_charset_priority):
* src/nsterm.m (append2):
* src/window.c (window_list):
* src/xfaces.c (Fx_list_fonts):
Use nconc2 instead of open-coding it.
* src/eval.c (eval_sub, backtrace_frame_apply):
* src/kqueue.c (kqueue_generate_event):
* src/nsterm.m (performDragOperation:):
* src/pdumper.c (Fpdumper_stats):
* src/w32.c (init_environment):
Prefer list1, list2, etc. to open-coding them.
* src/font.c (font_list_entities):
Parenthesize to avoid expanding new ‘list’ macro.
* src/gtkutil.c (GETSETUP): Rename from MAKE_FLOAT_PAGE_SETUP
to get lines to fit. Move outside the ‘list’ call, since it’s
now a macro.
* src/keymap.c (Fmake_keymap): Simplify.
* src/lisp.h (list, pure_list): New macros.
(list1i): New function.
Merge diff-font-lock-refine and diff-auto-refine-mode into diff-refine
This change was discussed in Bug#32991.
* admin/gitmerge.el (gitmerge-resolve): Bind 'diff-refine'
instead of 'diff-auto-refine-mode' to nil.
* doc/emacs/files.texi (Diff Mode): Explain 'diff-refine'
instead of 'diff-auto-refine-mode' in the documentation of
'diff-hunk-next' and 'diff-hunk-prev'. Mention in the
documentation of 'diff-refine-hunk' that refining is already
done by default.
* etc/NEWS (Diff mode): Explain renamed 'diff-refine' variable
and mention deprecation and disabling of
'diff-auto-refine-mode'.
* lisp/vc/diff-mode.el (diff-font-lock-refine): Rename to
'diff-refine' and allow choices nil, 'font-lock' and 'navigation'.
(diff-auto-refine-mode): Disable it by default, make it
obsolete and make it set 'diff-refine' appropriately to keep
backward compatibility.
(diff-hunk-next, diff-hunk-prev): Adapt to rename of
diff-auto-refine-mode and ensure that refining only happens
when calling these commands interactively.
(diff--font-lock-refined): Adapt to rename of
diff-font-lock-refine.
* lisp/vc/smerge-mode.el (smerge-next, smerge-prev): Check
that 'diff-refine' is set instead of checking
'diff-auto-refine-mode' when deciding whether to refine a
conflict.
Eli Zaretskii [Sun, 3 Mar 2019 17:02:34 +0000 (19:02 +0200)]
Fix starting temacs interactively
* src/dispnew.c (init_display_interactive): Don't call
init_faces_initial if we aren't initialized, since
tty-set-up-initial-frame-faces is not available then. This
restores the ability to start "temacs -nw" and also avoids
bad crashes in "emacs -nw" if the pdumper file is not found.
(Bug#34707)
Paul Eggert [Sun, 3 Mar 2019 07:02:01 +0000 (23:02 -0800)]
Fix typo in previous change
* src/alloc.c (memory_full_cons_threshold):
Move to after definition of struct cons_block.
Problem reported by Basil L. Contovounesios in:
https://lists.gnu.org/r/emacs-devel/2019-03/msg00067.html
Paul Eggert [Sat, 2 Mar 2019 19:05:40 +0000 (11:05 -0800)]
Avoid staticvec duplicates
* src/alloc.c (staticpro) [ENABLE_CHECKING]: Check for duplicates.
* src/keyboard.c (syms_of_keyboard): Define
while-no-input-ignore-events and inhibit--record-char here ...
(syms_of_keyboard_for_pdumper): ... instead of here.
This avoids duplicates in staticvec.
Eli Zaretskii [Sat, 2 Mar 2019 18:07:36 +0000 (20:07 +0200)]
Fix DND on MS-Windows with files from UNC directories
* lisp/dnd.el (dnd-get-local-file-uri): Always return nil on
MS-Windows, as this method cannot possibly work there: URIs
that begin with the local system's name are UNCs, where the
//SERVER part cannot be removed. (Bug#34675)
rx: fix `or' ordering by adding argument to regexp-opt
The rx `or' form may reorder its arguments in an unpredictable way,
contrary to user expectation, since it sometimes uses `regexp-opt'.
Add a NOREORDER option to `regexp-opt' for preventing it from
producing a reordered regexp (Bug#34641).
* doc/lispref/searching.texi (Regular Expression Functions):
* etc/NEWS (Lisp Changes in Emacs 27.1):
Describe the new regexp-opt NOREORDER argument.
* lisp/emacs-lisp/regexp-opt.el (regexp-opt): Add NOREORDER.
Make no attempt at regexp improvement if the set of strings contains
a prefix of another string.
(regexp-opt--contains-prefix): New.
* lisp/emacs-lisp/rx.el (rx-or): Call regexp-opt with NOREORDER.
* test/lisp/emacs-lisp/rx-tests.el: Test rx `or' form match order.
Eli Zaretskii [Sat, 2 Mar 2019 10:17:47 +0000 (12:17 +0200)]
Fix 'end-of-visual-line' with overlay strings with newlines
* src/indent.c (Fvertical_motion): Get out of overlay strings
with embedded newlines even if moving within the same screen
line. See https://github.com/emacs-lsp/lsp-mode/issues/677
for more details.
Fix test for IT_CHARPOS being at the beginning of the
accessible portion of the buffer.
Alan Mackenzie [Fri, 1 Mar 2019 17:35:12 +0000 (17:35 +0000)]
Maintain interval ->position fields correctly in update_interval
Also fix some anomalies in the handling of byte positions in regexp-emacs.c
This fixes bug #34525.
* src/intervals.c (SET_PARENT_POSITION): New macro.
(update_interval): When moving to an interval's parent, set that parent's
->position field, to maintain the consistency of the tree.
* src/intervals.h (struct interval): Amend the comment describing when
->position is valid.
* src/pdumper.c: Update the hash associated with struct interval.
* src/regex-emacs.c: (re_match_2_internal): Only invoke POINTER_TO_OFFSET on a
known character boundary. Only perform arithmetic on character positions, not
on byte positions. Correct the argument to an invocation of
UPDATE_SYNTAX_TABLE_FORWARD by adding 1 to it (in case wordend:).
* src/syntax.c: (update_syntax_table): Remove the now redundant code that set
the ->position field of all parents of the interval found by update_interval.
Glenn Morris [Fri, 1 Mar 2019 17:40:22 +0000 (09:40 -0800)]
Merge from origin/emacs-26
9df1365 Fix a typo in the Calc manual f3dab02 Minor improvement in cross-references of the ELisp manual 560c84b Fix last change on 'compilation-parse-errors'
Glenn Morris [Fri, 1 Mar 2019 17:40:22 +0000 (09:40 -0800)]
Merge from origin/emacs-26
dee3cdc Minor improvement for docs of completion c86d419 ; * src/image.c (imagemagick_load, svg_load): Fix typos in co... 6d46fa9 Disable the timerfd interface on Cygwin 3707ea4 Fix a typo in the doc string of 'regex-opt' d9905d5 Document bash 5.0.0 misbehavior in tramp.texi (Bug#34192)
Paul Eggert [Fri, 1 Mar 2019 17:01:59 +0000 (09:01 -0800)]
Suppress GC stats when obviously not needed
This should help future improvements where these stats can be
bignums that do not fit into intmax_t.
* src/alloc.c (struct gcstat, gcstat): New type and static var,
to package up GC statistics into one C object. It replaces ...
(total_free_intervals, total_intervals, total_strings)
(total_free_strings, total_string_bytes, total_vectors)
(total_vector_slots, total_free_vector_slots): ... these
removed static vars. All uses changed.
(garbage_collect_1): Accept a struct gcstat *, not a void *
which was not used anymore anyway. Return a bool indicating
success, instead of a Lisp object. All callers changed.
(garbage_collect): New function. All C callers of
Fgarbage_collect changed to use it, since none of them use the
return value. Now, only Lisp code uses Fgarbage_collect.
(Fgarbage_collect): No longer noinline. Cons up the return
value here, not in garbage_collect_1.
Evan Moses [Thu, 21 Feb 2019 21:37:01 +0000 (13:37 -0800)]
Add 'breakpoint' to builtins for Python
* lisp/progmodes/python.el (python-font-lock-keywords-level-2)
(python-font-lock-keywords-maxiumum-decoration): Add 'breakpoint'
to the list of builtins, it's new as of Python 3.7.
Eli Zaretskii [Fri, 1 Mar 2019 09:44:52 +0000 (11:44 +0200)]
Be more defensive regarding elements of 'load-history'
* lisp/loadhist.el (file-dependents):
* lisp/apropos.el (apropos-library):
* lisp/help-fns.el (help-fns--autoloaded-p, help--loaded-p):
* lisp/emacs-lisp/package.el (package--list-loaded-files):
Don't assume 'load-history' elements must have a string as their
'car'. (Bug#34462)
Eli Zaretskii [Thu, 28 Feb 2019 19:16:10 +0000 (21:16 +0200)]
Avoid rare crashes in xbacktrace
* src/eval.c (backtrace_thread_p, backtrace_top): Don't
segfault in "xbacktrace" if called before the specpdl
machinery is initialized in pdumped Emacs.
Paul Eggert [Wed, 27 Feb 2019 09:14:27 +0000 (01:14 -0800)]
DEFVAR_INT variables are now intmax_t
Formerly they were fixnums, which led to problems when dealing
with values that might not fit on 32-bit platforms, such as
string-chars-consed or floats_consed. 64-bit counters should
be good enough for these (for a while, anyway...).
While we’re at it, fix some unlikely integer overflow bugs
that have been in the code for a while.
* lib-src/make-docfile.c (write_globals):
* src/data.c (do_symval_forwarding, store_symval_forwarding):
* src/eval.c (restore_stack_limits, call_debugger):
* src/frame.h (struct frame.cost_calculation_baud_rate):
* src/keyboard.c (last_auto_save, bind_polling_period, read_char):
* src/lisp.h (struct Lisp_Intfwd.intvar):
* src/lread.c (defvar_int):
* src/pdumper.c (dump_fwd_int):
* src/thread.h (struct thread_state.m_lisp_eval_depth):
* src/undo.c (truncate_undo_list):
* src/xselect.c (wait_for_property_change)
(x_get_foreign_selection):
* src/xterm.c (x_emacs_to_x_modifiers):
DEFVAR_INT variables now have the C type intmax_t, not EMACS_INT.
* src/data.c (store_symval_forwarding):
* src/gnutls.c (Fgnutls_boot):
* src/keyboard.c (bind_polling_period):
* src/macros.c (pop_kbd_macro, Fexecute_kbd_macro):
* src/undo.c (truncate_undo_list):
Allow any integer that fits into intmax_t, instead of
requiring it to be a Lisp fixnum.
* src/dispnew.c (update_window):
* src/frame.c (x_figure_window_size):
* src/gnutls.c (init_gnutls_functions)
(emacs_gnutls_handle_error):
* src/keyboard.c (make_lisp_event):
* src/nsterm.m (ns_dumpglyphs_image):
* src/profiler.c (make_log):
* src/scroll.c (calculate_scrolling)
(calculate_direct_scrolling):
* src/termcap.c (tputs):
* src/xterm.c (x_draw_image_relief):
Avoid implementation-defined behavior on conversion of
out-of-range integers.
* src/eval.c (when_entered_debugger): Now intmax_t.
(max_ensure_room): New function, that avoids signed integer overflow.
(call_debugger, signal_or_quit): Use it.
* src/fileio.c (Fdo_auto_save):
* src/keyboard.c (make_lisp_event):
* src/term.c (calculate_costs):
* src/xdisp.c (build_desired_tool_bar_string)
(hscroll_window_tree, try_scrolling, decode_mode_spec)
(x_produce_glyphs):
Avoid signed integer overflow.
* src/lisp.h (clip_to_bounds): Generalize to intmax_t.
* src/pdumper.c (dump_emacs_reloc_immediate_emacs_int): Remove, ...
(dump_emacs_reloc_immediate_intmax_t): ... replacing with this
function. All uses changed.
* src/profiler.c (make_log): Omit args. All callers changed.
* src/termcap.c: Include stdlib.h, for atoi.
Include intprops.h.
* src/window.c (sanitize_next_screen_context_lines): New function.
(window_scroll_pixel_based, window_scroll_line_based):
Use it to avoid signed integer overflow.
Paul Eggert [Tue, 26 Feb 2019 01:44:46 +0000 (17:44 -0800)]
Fix mod-test build failure
Problem reported by Glenn Morris in:
https://lists.gnu.org/r/emacs-devel/2019-02/msg00739.html
* test/Makefile.in (HYBRID_MALLOC, LIBEGNU_ARCHIVE):
New macros, taken from ../src/Makefile.in.
(MODULE_CFLAGS): Add -I$(srcdir)/../lib.
($(test_module)): Link $(LIBEGNU_ARCHIVE) too.
Paul Eggert [Mon, 25 Feb 2019 19:33:51 +0000 (11:33 -0800)]
format-time-string: document new '+' flag
* doc/lispref/os.texi (Time Parsing), etc/NEWS:
* src/timefns.c (Fformat_time_string): Document the new
behavior, added for compatibility with POSIX.1-2017.
* lisp/net/tramp-cmds.el (tramp-cleanup-connection-hook)
(tramp-cleanup-all-connections-hook): New hook variables.
(tramp-cleanup-connection): Set `tramp-current-connection' always
to nil.
(tramp-cleanup-connection): Do not call `tramp-recentf-cleanup'.
Run ´tramp-cleanup-connection-hook`.
(tramp-cleanup-all-connections): Do not call
`tramp-archive-cleanup-hash' and ´tramp-recentf-cleanup'. Run
`tramp-cleanup-all-connections-hook'.
pending_signals is often set if no quit is pending. This results in
bugs in module code if the module returns but no quit is actually
pending.
As a better alternative, add a new process_input environment function
for Emacs 27. That function processes signals (like maybe_quit).
* configure.ac: Add module snippet for Emacs 27.
* src/module-env-27.h: New file.
* src/emacs-module.h.in: Add process_input function to environment
interface.
* src/emacs-module.c (module_should_quit): Use QUITP macro to check
whether the caller should quit.
(module_process_input): New function.
(initialize_environment): Use it.
* src/eval.c: Remove obsolete comment.
* test/data/emacs-module/mod-test.c (signal_wrong_type_argument)
(signal_errno): New helper functions.
(Fmod_test_sleep_until): New test module function.
* test/src/emacs-module-tests.el (mod-test-sleep-until): New unit
test.
Stefan Monnier [Sun, 24 Feb 2019 21:19:59 +0000 (16:19 -0500)]
* lisp/term.el: Fix minor compilation issues with cl-lib and lexbind
Remove left-over comment from the bulk comint->term query/replace.
(term-command-function): Rename from term-command-hook.
Give it a global default value.
(term-suppress-hard-newline): Mark it as generally obsolete.
(term-mode): Don't manually make hooks buffer-local.
(term--remove-fake-newlines): Fix assert -> cl-assert.
(term-char-mode): Use add-function.
(term-send-input): Use run-hook-with-args.
(term-dynamic-complete): Use run-hook-with-args-until-success.
(term-dynamic-simple-complete): Completion tables can be plain lists.
(serial-read-name): Simplify and fix misuse of `set`.
John Shahid [Mon, 21 Jan 2019 00:08:17 +0000 (19:08 -0500)]
Adjust line wrapping on window resize and killing text
* lisp/term.el (term-mode): Advice filter-buffer-substring-function to
remove line unwrapping from killed text.
(term-reset-size): Add or remove line unwrapping depending on the new
terminal width.
(term-suppress-hard-newline): Mark obsolete.
(term-unwrap-line): Use text properties to be able to find the
newlines later.
Eric Abrahamsen [Wed, 30 Jan 2019 20:31:49 +0000 (12:31 -0800)]
Make pinyin to Chinese character mapping available to elisp
* leim/Makefile.in: Build the file pinyin.el from pinyin.map.
* lisp/international/titdic-cnv.el (pinyin-convert): New function that
writes the library pinyin.el, containing a new constant
`pinyin-character-map'.
* .gitignore: Ignore the generated pinyin.el file.
Richard Stallman [Sun, 24 Feb 2019 18:45:34 +0000 (10:45 -0800)]
fix rmail armor decryption problems
* lisp/mail/rmail.el (rmail-epa-decrypt): Don't decrypt an armor
that was copied into the message from a message it is a reply to.
(rmail-epa-decrypt-1): Catch and ignore errors in
epa-decrypt-region. Make armor-start and armor-end markers.
Paul Eggert [Sat, 23 Feb 2019 21:47:52 +0000 (13:47 -0800)]
Don’t assume timersub and gettimeofday
POSIX does not specify timersub, and marks gettimeofday as
obsolescent. Avoid porting problems by using timespec.h
functions instead.
* src/editfns.c: Include systime.h instead of sys/time.h.
(EXTRA_CONTEXT_FIELDS): Replace start and max_secs with
time_limit. All uses changed. This removes the need to call
gettimeofday or timersub.
* src/term.c (timeval_to_Time): Remove. Replace all uses with ...
(current_Time): ... this new function, removing the need to
call gettimeofday.
* src/editfns.c (Freplace_buffer_contents): Add two optional arguments
for mitigating performance issues.
* lisp/emacs-lisp/subr-x.el (replace-region-contents): Move from
subr.el. Add the same two arguments as for replace-buffer-contents.
* lisp/json.el (json-pretty-print-max-secs): New variable holding the
default MAX-SECS value json-pretty-print passes to
replace-buffer-contents.
(json-pretty-print): Use it.
* doc/lispref/text.texi (Replacing): Add documentation for
replace-buffer-contents two new optional arguments. Document
replace-region-contents.
Eli Zaretskii [Sat, 23 Feb 2019 13:31:15 +0000 (15:31 +0200)]
Minor improvement for docs of completion
* doc/lispref/minibuf.texi (Completion Commands)
(Completion in Buffers, Programmed Completion): Add to text
that references completion tables a cross-reference to where
"completion table" is described.
(Programmed Completion): Fix the description of
'completion-table-dynamic'. Add more cross-references.
Stefan Monnier [Thu, 21 Feb 2019 17:09:03 +0000 (12:09 -0500)]
* lisp/vc/diff-mode.el: Reduce redundancy in diff-syntax-fontify code
(diff--iterate-hunks): New function extracted from diff--font-lock-refined.
(diff--font-lock-refined, diff--font-lock-syntax): Use it.
(diff--overlay-auto-delete): Rename from diff--font-lock-refine--refresh.
(diff--font-lock-syntax--refresh): Delete.
(diff-syntax-fontify-hunk): Don't completely silence errors.
(diff-syntax-fontify-props): Remove `no-init` arg, testing if `file` is
nil instead. Adjust all callers.
(diff-syntax-fontify-props): Remove redundant code since we don't
modify the buffer.
Correct implementation of `sql-set-product-feature' (Bug#30494).
* lisp.progmodes/sql.el (sql-add-product): Correct argument spec.
(sql-set-product-feature): Handle all cases as intended.
(sql-get-product-feature): Fetch varaiable value by `eval'.
* test/lisp/progmodes/sql-tests.el (sql-test-feature-value-[a-d]):
New test variables.
(sql-test-product-feature-harness): New test macro.
(sql-test-add-product, sql-test-add-existing-product)
(sql-test-set-feature, sql-test-set-indirect-feature)
(sql-test-set-existing-feature)
(sql-test-set-existing-indirect-feature)
(sql-test-set-missing-product, sql-test-get-feature)
(sql-test-get-indirect-feature, sql-test-get-missing-product)
(sql-test-get-missing-feature)
(sql-test-get-missing-indirect-feature): New ERT tests
Glenn Morris [Wed, 20 Feb 2019 06:04:19 +0000 (22:04 -0800)]
Remove .art from the default list of ImageMagick extensions
It seems that .art files can be non-image files that
ImageMagick mistakenly treats as extremely large images.
Real .art images seem rare.
* lisp/image.el (imagemagick-enabled-types): Remove ART. (Bug#22289)
* lisp/emacs-lisp/easy-mmode.el (easy-mmode-define-navigation):
Simplify a one-argument call to "or" and use buffer-narrowed-p instead
of checking that condition by hand.
João Távora [Mon, 18 Feb 2019 20:32:38 +0000 (20:32 +0000)]
switch-to-buffer's completion table uses its own sorting
* src/minibuf.c (Finternal_complete_buffer): Add
Qcycle_sort_function to completion table's metadata.
(syms_of_minibuf): New symbol Qcycle_sort_function.