As the constraint logic of the compiler is not trivial and largely
independent from the rest of the code move it into comp-cstr.el to
ease separation and maintainability.
This commit improve the conversion type
specifier -> constraint for generality.
Lastly this should help with bootstrap time as comp.el compilation
unit is slimmed down.
* lisp/emacs-lisp/comp-cstr.el: New file.
(comp--typeof-types, comp--all-builtin-types): Move from comp.el.
(comp-cstr, comp-cstr-f): Same + rename.
(comp-cstr-ctxt): New struct.
(comp-supertypes, comp-common-supertype-2)
(comp-common-supertype, comp-subtype-p, comp-union-typesets)
(comp-range-1+, comp-range-1-, comp-range-<, comp-range-union)
(comp-range-intersection): Move from comp.el.
(comp-cstr-union-no-range, comp-cstr-union): Move from comp.el and
rename.
(comp-cstr-union-make): New function.
(comp-type-spec-to-cstr, comp-cstr-to-type-spec): Move from
comp.el, rename it and rework it.
* lisp/emacs-lisp/comp.el (comp-known-func-cstr-h): Rework.
(comp-ctxt): Remove two fields and include `comp-cstr-ctxt'.
(comp-mvar, comp-fwprop-call): Update for `comp-cstr' being
renamed.
(comp-fwprop-insn): Use `comp-cstr-union-no-range' or
`comp-cstr-union'.
(comp-ret-type-spec): Use `comp-cstr-union' and rework.
* test/lisp/emacs-lisp/comp-cstr-tests.el: New file.
(comp-cstr-test-ts, comp-cstr-typespec-test): New functions.
(comp-cstr-typespec-tests-alist): New defconst to generate tests
on.
(comp-cstr-generate-tests): New macro.
* test/src/comp-tests.el (comp-tests-type-spec-tests): Update.
(ret-type-spec): Initialize constraint context.
* lisp/cus-edit.el (custom-variable-modified-p): Quote the value when
custom form is 'lisp (or 'mismatch) prior to comparing in order to
accommodate `custom-variable-value-create' (bug#44852).
* etc/tutorials/TUTORIAL: Don't keep referring to EDIT as if it were a
common name for the Meta key; since a few decades back it's labelled
Alt (or Option or ⌥ but those keys usually also have 'alt' engraved on
them). Similarly, CTL is practically extinct and not worth
mentioning.
For certain patterns, re-matching the same regexp on the matched
substring does not produce correctly translated match data
(bug#15107 and bug#44861).
Using a new builtin function also improves performance since the
number of calls to string-match is halved.
Reported by Kevin Ryde and Shigeru Fukaya.
* lisp/subr.el (replace-regexp-in-string): Translate the match data
using match-data--translate instead of trusting a call to string-match
on the matched string to do the job.
* test/lisp/subr-tests.el (subr-replace-regexp-in-string):
Add test cases.
* src/search.c (Fmatch_data__translate): New internal function.
(syms_of_search): Register it as a subroutine.
* lisp/progmodes/cc-engine.el (c-full-pp-to-literal): Handle correctly END
being before HERE by using parse-partial-sexp to get the end of the literal
containing HERE.
* lisp/emacs-lisp/eldoc.el (eldoc-display-in-echo-area): Use
'save-excursion' to keep point position in *eldoc* buffer.
Suggested by Andrii Kolomoiets <andreyk.mad@gmail.com>.
* lisp/gnus/gnus-search.el (gnus-search-imap-handle-flag): This allows
"mark:A" to be translated into "ANSWERED" and "-mark:A" to be
translated into "UNANSWERED".
This behavior both better matches the previous nnir behavior, reducing
confusion for new users, and matches behavior when using parsed
queries.
* lisp/gnus/gnus-search.el (gnus-search-imap-search-keys): Make sure
this variable contains all known IMAP search keys.
(gnus-search-run-search): If the search query doesn't start with a
known search key, prepend "TEXT " to the query.
6442cdc0e4 Revert extra focus redirection in do_switch_frame (Bug#24803)
fc4379f1ae Minor cleanup of tramp-tests.el on MS Windows
dea3d6aa18 Fix handling of defcustom :local tag
* src/frame.c (do_switch_frame): Do not also redirect frame
focus when FRAME has its minibuffer window on the selected
frame which was intended to fix Bug#24500. It may cause
Bug#24803 and lead to a nasty state where no active cursor is
shown on any frame, see
https://lists.gnu.org/archive/html/emacs-devel/2020-11/msg01137.html.
* doc/lispref/minibuf.texi (Completion Variables)
(Programmed Completion): Add affixation-function.
* lisp/help-fns.el (help--symbol-completion-table-affixation): New function.
(help--symbol-completion-table): Set affixation-function when
completions-detailed is non-nil.
* lisp/minibuffer.el (completion-metadata): Add affixation-function
to docstring.
(completions-annotations): Inherit from shadow with italic.
(completions-detailed): New defcustom.
(completion--insert-strings): Count string-width on all strings in
completion list. Insert prefix and suffix.
(completion-extra-properties): Add affixation-function to docstring.
(minibuffer-completion-help): Call affixation-function.
(minibuffer-default-prompt-format): Move down closer to its use.
https://lists.gnu.org/archive/html/emacs-devel/2020-11/msg00613.html
* lisp/face-remap.el
(text-scale-remap-header-line-face): New buffer local variable.
(text-scale-mode): Adjust header line if above variable is non-nil.
(face-remap--clear-remappings, face-remap--remap-face): New defuns.
* lisp/face-remap.el: Arrange to watch text-scale-mode-remapping.
(text-scale--refresh): New function.
* lisp/emacs-lisp/tabulated-list.el (tabulated-list-mode): Use
text-scale-remap-header-line. (Bug#41852)
* doc/emacs/killing.texi (Earlier Kills): Document standalone M-y.
* doc/emacs/search.texi (Isearch Yank): Explain standalone M-y.
* doc/lispref/text.texi (Yank Commands): Soften the wording of
yank after another yank.
* lisp/delsel.el: Put 'yank' property on yank-pop and yank-from-kill-ring.
* lisp/isearch.el (isearch-yank-pop): Use with-isearch-suspended
and read-from-kill-ring to read a string from the kill-ring and
append it to the search string.
* lisp/simple.el (yank-pop): Call yank-from-kill-ring and
read-from-kill-ring when last-command is not 'yank' instead of
signaling an error. Remove "*" from interactive spec. Update docstring.
(read-from-kill-ring): New function.
(yank-from-kill-ring): New command.
https://lists.gnu.org/archive/html/emacs-devel/2020-11/msg00801.html
For discussion, see the following emacs-devel thread:
https://lists.gnu.org/r/emacs-devel/2020-11/msg00734.html
* lisp/custom.el (custom-declare-variable): Delay call to
make-variable-buffer-local until after user option has been
initialized with a value. Otherwise the user option may be
initialized to nil.
* test/lisp/custom-tests.el (custom--test-local-option)
(custom--test-permanent-option): New :local user options.
(custom-test-local-option): New test for defcustom :local keyword.
A choice widget should be able to match either no inline values or
inline values, upon request. (Bug#44579)
* lisp/wid-edit.el (choice): New property, :inline-bubbles-p. A
predicate that returns non-nil if the choice widget can act as an
inline widget. Document it.
(widget-choice-inline-bubbles-p): New function, for the
:inline-bubbles-p property of the choice widget.
(widget-inline-p): New function. Use the :inline-bubbles-p property
of the widget, if any.
(widget-match-inline): Use the above to see if the widget can act like
an inline widget. Document it.
(widget-choice-value-create): Account for the case of a choice widget
that has inline members.
(widget-checklist-add-item, widget-editable-list-value-create)
(widget-group-value-create): Use widget-inline-p rather than just
checking for a non-nil :inline property, allowing these functions to
pass the complete information to widgets like the choice widget to
create their values.
* test/lisp/wid-edit-tests.el (widget-test-choice-match-no-inline)
(widget-test-choice-match-all-inline)
widget-test-choice-match-some-inline): New tests, to check that choice
widgets can match its choices, inline or not.
(widget-test-inline-p): New test, for the new function
widget-inline-p.
(widget-test-repeat-can-handle-choice)
(widget-test-repeat-can-handle-inlinable-choice)
(widget-test-list-can-handle-choice)
(widget-test-list-can-handle-inlinable-choice)
(widget-test-option-can-handle-choice)
(widget-test-option-can-handle-inlinable-choice): New tests. This
grouping widgets need to be able to create a choice widget regardless
if it has inline choices or not.
* lisp/filesets.el (filesets-files-under): New function, used to get
all files for a :tree fileset.
(filesets-get-filelist): Use it. Look for the directory and the
pattern in the right place inside entry.
* doc/misc/gnus.texi (Summary Maneuvering): Document it.
* lisp/gnus/gnus-sum.el (gnus-paging-select-next): New variable.
(gnus-summary-prev-page, gnus-summary-next-page): Use it.
* src/Makefile.in (temacs$(EXEEXT)): Codesign the executable on
recent (ARM) MacOS systems (bug#43878). Without this, building
Emacs fails.
Copyright-paperwork-exempt: yes
* lisp/server.el (server-start): Delete the server directory upon
Emacs exit (bug#44644). This fixes the problem of /tmp/emacs0
directories being left behind when running an Emacs server as root.
* lisp/progmodes/flymake.el (flymake--mode-line-format): Don't put
a face on the the "?" unknown backend indicator, because that
looks odd in inactive windows (bug#44689).
* lisp/outline.el (outline--cycle-state): Only consider outline
overlays that are on outline headings; when subtree end is
point-max, return overlay-end +1 because final subtree overlay
only reaches point-max -1 (bug#41198).
(outline-cycle-buffer): Check that buffer has top-level headings
before calling outline-hide-sublevels 1 thus preventing
disconcerting buffer state of content reduced to single "..."
* lisp/emacs-lisp/comp.el (native-compile-async-skip-p): New function
ripping out logic from `native--compile-async' and accounting for
SELECTOR.
(native--compile-async): Add SELECTOR parameter, make use of
`native-compile-async-skip-p' and move it with other private
functions.
(native-compile-async): Add SELECTOR parameter.
