Improve the handling of `or` patterns where not all sub-patterns bind the
same set of variables. This used to be "unsupported" and behaved in
somewhat unpredictable ways.
(pcase--expand): Rewrite.
(pcase-codegen): Delete.
* doc/lispref/control.texi (pcase Macro): Adjust accordingly.
Also remove the warning about "at least two" sub patterns.
These work fine, AFAICT, and if not we should fix it.
* test/lisp/emacs-lisp/pcase-tests.el (pcase-tests-or-vars): New test.
Revert commit a218c98615, but in order
to avoid the spurious warnings that this commit tried to squash,
keep track of the vars used during the match so as to add
corresponding annotations to explicitly silence the spurious warnings.
To do this, we change the VARS used in `pcase-u` (and throughout
the pcase code): they used to hold elements of the form (NAME . VAL)
and now they hold elements of the form (NAME VAL . USED).
(pcase--expand): Bind all vars instead of only those found via fgrep.
(pcase-codegen): Silence "unused var" warnings for those vars that have
already been referenced during the match itself.
(pcase--funcall, pcase--eval): Record the vars that are used.
(pcase--u1): Record the vars that are used via non-linear patterns.
* lisp/textmodes/mhtml-mode.el (mhtml-forward):
* lisp/vc/diff-mode.el (diff-goto-source): Silence newly
discovered warnings.
* test/lisp/emacs-lisp/pcase-tests.el (pcase-tests-bug46786): New test.
Fix incorrect variable scoping in `let*`, `for` and `foreach`.
Fix loop variable value in `foreach` (should be element, not tail).
Fix function quoting, as in ('cons x y) -- didn't work at all.
Reported by Stephan Neuhaus.
* lisp/calc/calc-prog.el (math-define-exp, math-handle-foreach):
* test/lisp/calc/calc-tests.el: (var-g, test1, test2, test3, test4)
(test5, test6, test7, calc-defmath): Test various defmath forms.
* doc/misc/calc.texi: Remove references to the old Calc representation
of big integers, outdated references to fixnums, an any text and
examples that only made sense at the time.
* etc/NEWS: Document new function.
* lisp/custom.el (require-theme): Add function.
This follows from the discussion on bug#45068 where it became apparent
that there was no equivalent mechanism to 'require' that read through
the 'custom-theme-load-path'.
* lisp/progmodes/cc-cmds.el (c-where-wrt-brace-construct): Reformulate latter
part such that the least enclosing braces and parentheses are used when
determining containment in such. c-beginning-of-decl-1 has been superseded
by list movement and syntactic whitespace movement.
(c-backward-to-nth-BOF-{): Work on least enclosing parens rather than parens
at any level when moving back to an opening brace.
(c-forward-to-nth-EOF-\;-or-}): Work on least enclosing parens, as above.
Move the correction of point when in a "function trailer" to after the main
loop, correcting a minor bug.
These seem to be left overs from Emacs<24 when `macroexpand-all` was
implemented in the CL library and hence the macros's evaluation
environment could come from different places depending on the
circumstance (either `byte-compile-macro-environment`, or
`cl-macro-environment`, or ...).
`byte-compile-macro-environment` contains definitions which expand to
code that is only understood by the rest of the byte-compiler,
so using it for code which isn't being byte-compiled leads to errors
such as references to non-existing function
`internal--with-suppressed-warnings`.
* lisp/emacs-lisp/cl-extra.el (cl-prettyexpand): Remove left-over
binding from when `macroexpand-all` was implemented in the CL library.
* lisp/emacs-lisp/ert.el (ert--expand-should-1):
* lisp/emacs-lisp/cl-macs.el (cl--compile-time-too): Properly preserve the
macroexpand-all-environment.
(cl--macroexp-fboundp): Pay attention to `cl-macrolet` macros as well.
Extend assumes allowing the following form:
(assume dst (and-nhc src1 src2))
`and-nhc' assume operator allow for constraining correctly
intersections where non hash consed values are not propagated as
values but rather promoted to their types.
* lisp/emacs-lisp/comp-cstr.el
(comp-cstr-intersection-no-hashcons): New function.
* lisp/emacs-lisp/comp.el (comp-emit-assume): Logic update to emit
`and-nhc' operator (implemented in fwprop by
`comp-cstr-intersection-no-hashcons').
(comp-add-cond-cstrs): Map `eq' to `and' assume operator and
`equal' `eql' into `and-nhc'.
(comp-fwprop-insn): Update to handle `and-nhc'.
* test/src/comp-tests.el (comp-tests-type-spec-tests): Add two
tests covering `eql' and `equal' propagation of non hash consed
values.
Further testing seems to confirm my suspicion that the increase in the
specpdl comes from the recent change to `pcase--if`.
* lisp/international/mule-cmds.el (update-leim-list-file): Revert workaround.
* src/marker.c (unchain_marker): Clean up obsolete comment
(bug#46836).
Commit cf3164523b (* src/alloc.c: Avoid O(N²) complexity when
unchaining markers (bug#24548)., 2018-03-23) removed the call from GC
into unchain_marker, so there is no need to warn about it.
* lisp/net/puny.el: change puny-skew to match value given in RFC3492.
* test/lisp/net/puny-tests.el (puny-test-encode-domain)
(puny-test-decode-domain): add regression case for popular
domain. (bug#46838).
* lisp/emacs-lisp/lisp-mode.el (lisp-mode-map): Move menu from
here...
(lisp-mode-menu): ...to here, and convert to easymenu.
* lisp/progmodes/elisp-mode.el (lisp-interaction-mode-map): Move
menu definition from here...
(lisp-interaction-mode-menu): ...to here, and convert to easymenu.
* lisp/replace.el (occur-menu-map): Convert to easymenu.
Extend assumes allowing the following form
(assume dst (= src1 src2))
to caputure '=' semanting during fwprop handling float integer
conversions.
* lisp/emacs-lisp/comp.el (comp-equality-fun-p): Don't treat '=' as
simple equality.
(comp-arithm-cmp-fun-p, comp-negate-arithm-cmp-fun)
(comp-reverse-arithm-fun): Rename and add '=' '!='.
(comp-emit-assume, comp-add-cond-cstrs, comp-fwprop-insn): Update
for new function nameing and to handle '='.
* lisp/emacs-lisp/comp-cstr.el (comp-cstr-=): New function.
* test/src/comp-tests.el (comp-tests-type-spec-tests): Add a bunch
of '=' specific tests.
* etc/NEWS: Amend.
* doc/lispref/minibuf.texi (Minibuffer History): Amend the entry about
goto-line-history, which is now buffer local only after being so customized.
* lisp/emacs-lisp/byte-opt.el (pure-fns): Treat string>,
string-greaterp, string-empty-p, string-blank-p, string-prefix-p and
string-suffix-p as pure functions in the compiler.
* lisp/ibuffer.el (ibuffer-mode-map): Move menu from here...
(ibuffer-mode-mark-menu, ibuffer-mode-view-menu): ...to here.
Convert to easymenu.
(ibuffer-mode--groups-menu-definition): New function. Fix bug where
"Kill filter by group named" and "Yank last killed filter group
before" was overwritten and never shown in the menu.
(ibuffer-mode-groups-popup): Convert to easymenu.
(ibuffer-mode-operate-menu): Rename from 'ibuffer-mode-operate-map'
and update uses. Convert to easymenu.
(ibuffer-mode-operate-map): Make into obsolete alias for above
renamed variable.
The pcase 'rx' pattern would in some cases allow the match data to be
clobbered before it is read. For example:
(pcase "PQR"
((and (rx (let a nonl)) (rx ?z)) (list 'one a))
((rx (let b ?Q)) (list 'two b)))
The above returned (two "P") instead of the correct (two "Q").
This occurred because the calls to string-match and match-string were
presented as separate patterns to pcase, which would interleave them
with other patterns.
As a remedy, combine string matching and match-data extraction into a
single pcase pattern. This introduces a slight inefficiency for two
or more submatches as they are grouped into a list structure which
then has to be destructured.
Found by Stefan Monnier. See discussion at
https://lists.gnu.org/archive/html/emacs-devel/2021-02/msg02010.html
* lisp/emacs-lisp/rx.el (rx--reduce-right): New helper.
(rx [pcase macro]): Combine string-match and match-string calls into a
single pcase pattern.
* test/lisp/emacs-lisp/rx-tests.el (rx-pcase): Add test cases.
(macroexp--expand-all): Use `pcase--dontcare` so pcase generates
slightly better code. Don't hardcode which functions takes function
arguments; rely on a new `funarg-positions` symbol property instead.
* lisp/isearch.el (isearch-menu-bar-yank-map)
(isearch-menu-bar-map, isearch-mode-map): Move menu definition
from here...
(isearch-menu-bar-map): ...to here, and convert to easymenu.
* lisp/loadup.el ("emacs-lisp/easymenu"): Move before isearch.el.
* lisp/tab-bar.el (tab-bar--define-keys): Update global-mode-string
in mode-line-misc-info with condition to disable global-mode-string
in the mode line.
(tab-bar-format): New variable.
(tab-bar-format-history, tab-bar-format-add-tab)
(tab-bar-format-tabs): New functions with body from
'tab-bar-make-keymap-1'.
(tab-bar-format-align-right, tab-bar-format-global): New functions for
'tab-bar-format' list.
(tab-bar-format-list): New utility function.
(tab-bar-make-keymap-1): Just call 'tab-bar-format-list'.
https://lists.gnu.org/archive/html/emacs-devel/2020-11/msg01210.html
* lisp/subr.el (read-char-choice): Move most of the function body to
'read-char-choice-with-read-key'.
(read-char-choice-with-read-key): New function with body from
'read-char-choice'.
