* lisp/progmodes/project.el (project-switch-commands):
Remove the ###autoload instruction. It's unnecessary and can
cause surprises in some circumstances (bug#46986).
* lisp/progmodes/xref.el (xref--insert-xrefs): Cut up the current
line into pieces here for multiple matches's summaries, so that
xref--insert-xrefs can do less work (bug#46859).
(xref--insert-xrefs): Do less work.
(xref--outdated-p):
Update accordingly to how the summary creation logic changed.
(xref--buf-pairs-iterator): Update to the new calling convention.
(xref-location-column): Effectively rename back to
xref-file-location-column since the generic version is now unused.
* test/lisp/progmodes/xref-tests.el
(xref-matches-in-directory-finds-two-matches-on-the-same-line)
(xref-matches-in-directory-finds-an-empty-line-regexp-match):
Adjust to the xref-location-column change.
(xref-matches-in-files-trims-summary-for-matches-on-same-line):
New test.
* test/lisp/progmodes/xref-resources/file1.txt:
Change contents slightly to test the new xref--outdated-p code.
This replaces most json-encode-* functions with similar
json--print-* counterparts that insert into the current buffer
instead of returning a string (bug#46761).
Some unused but useful json-encode-* functions are kept for backward
compatibility and as a public API, and the rest are deprecated.
* etc/NEWS: Announce obsoletions.
* lisp/json.el: Document organization of library. Make subsection
headings more consistent.
(json--encoding-current-indentation): Rename...
(json--print-indentation-prefix): ...to this, to reflect new use.
(json--encode-stringlike, json--encode-alist): Rename...
(json--print-stringlike, json--print-alist): ...to these,
respectively, and encode argument into current buffer instead. All
callers changed.
(json--print-string, json--print-unordered-map, json--print-array)
(json--print): New functions.
(json-encode-string, json-encode-plist, json-encode-array)
(json-encode): Use them, respectively.
(json-encode-number, json-encode-hash-table): Mark as obsolete
aliases of json-encode.
(json-encode-key, json-encode-list): Mark as obsolete in preference
for json-encode.
(json--print-indentation-depth, json--print-keyval-separator): New
variables.
(json--with-output-to-string): New macro.
(json--print-indentation, json--print-keyword, json--print-key)
(json--print-pair, json--print-map, json--print-list): New
functions.
(json--with-indentation): Use json--print-indentation-depth to avoid
unnecessary string allocation.
(json-encoding-default-indentation, json-pretty-print-max-secs):
Clarify docstrings.
(json--escape, json--long-string-threshold, json--string-buffer):
Remove; no longer used.
* lisp/progmodes/js.el (js--js-encode-value): Replace
json-encode-string and json-encode-number with json-encode.
(js-eval-defun): Use json--print-list to avoid
json-encode-list->insert roundtrip.
* test/lisp/json-tests.el (test-json-encode-number)
(test-json-encode-hash-table, test-json-encode-hash-table-pretty)
(test-json-encode-hash-table-lisp-style)
(test-json-encode-hash-table-sort, test-json-encode-list): Replace
uses of obsolete functions with the equivalent use of json-encode.
(test-json-encode-key): Suppress obsoletion warnings.
(test-json-encode-string): Check that text properties are stripped.
This fixes bug #46951.
* lisp/progmodes/cc-engine.el (c-append-lower-brace-pair-to-state-cache):
Ensure the starting point for backward scanning is not within a macro which
doesn't contain HERE.
Fix some alignment functionality in cc-align.el.
* lisp/progmodes/cc-align.el (c-lineup-arglist-intro-after-paren): Align the
next line under the previous entry rather than one to the right of the paren.
(c-lineup-2nd-brace-entry-in-arglist): Take the anchor point from the
brace-list-entry element, not the brace-list-intro one.
* lisp/progmodes/cc-engine.el (c-looking-at-decl-block): Use
c-looking-at-statement-block to test whether "struct A {" begins a brace list
or a struct declaration.
(c-looking-at-or-maybe-in-bracelist): Several detailed amendments, correctly
to recognize brace lists.
(c-looking-at-statement-block): No longer search for commas, as they are not
reliable indicators of a brace list. Search now for a restricted set of
keywords, since some can appear in brace lists in C++ mode.
* lisp/progmodes/cc-langs.el (c-stmt-block-only-keywords)
(c-stmt-block-only-keywords-regexp): New lang consts/vars.
(c-pre-id-bracelist-kwds): New lang const.
(c-pre-id-bracelist-key): Derive now from the above.
(c-pre-brace-non-bracelist-key): New lang const/var.
* 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.
* 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.
Instead of warning about unused vars during the analysis phase of
closure conversion, do it in the actual closure conversion by
annotating the code with "unused" warnings, so that the warnings
get emitted later by the bytecomp phase, like all other warnings,
at which point the line-number info is a bit less imprecise.
Take advantage of this change to wrap the expressions of unused
let-bound vars inside (ignore ...) so the byte-compiler can better
optimize them away.
Finally, promote `macroexp--warn-and-return` to "official" status
by removing its "--" marker.
(cconv-captured+mutated, cconv-lambda-candidates): Remove vars.
(cconv-var-classification): New var to replace them.
(cconv-warnings-only): Delete function.
(cconv--warn-unused-msg, cconv--var-classification): New functions.
(cconv--convert-funcbody): Add warnings for unused args.
(cconv-convert): Add warnings for unused vars in `let` and `condition-case`.
(cconv--analyze-use): Don't emit an "unused var" warning any more,
but instead remember the fact in `cconv-var-classification`.
* lisp/emacs-lisp/bytecomp.el (byte-compile-force-lexical-warnings):
Remove variable.
(byte-compile-preprocess): Remove corresponding case.
* lisp/emacs-lisp/pcase.el (pcase--if): Don't throw away `test` effects.
(\`):
* lisp/emacs-lisp/cl-macs.el (cl--do-arglist): Use `car-safe` instead
of `car`, so it can more easily be removed by the optimizer if the
result is not used.
* lisp/emacs-lisp/macroexp.el (macroexp--warn-wrap): New function.
(macroexp-warn-and-return): Rename from `macroexp--warn-and-return`.
* lisp/progmodes/simula.el (simula-mode-map): Move menu definition
from here...
(simula-mode-menu): ...to here, and use easy-menu-define.
(simula-popup-menu): Declare unused function obsolete.
Yes, finally: a function that tells you the name of the file where
the code is located. Finding this name is non-trivial in practice,
as evidenced by the "4 shift/reduce conflicts" warning when compiling
CEDET's python.el, because its `wisent-source` got it wrong in that
case, thinking the grammar came from `python.el` instead of
`python-wy.el`.
While at it, also made `macroexp-compiling-p` public, since it's
useful at various places.
(macroexp-compiling-p): Rename from `macroexp--compiling-p`.
* lisp/emacs-lisp/bytecomp.el (byte-compile-close-variables):
Bind `load-file-name` to nil so we can distinguish a load that calls
the byte compiler from a byte compilation which causes a load.
* lisp/cedet/semantic/wisent/python.el (wisent-python--expected-conflicts):
Remove; it was just a workaround.
* lisp/subr.el (do-after-load-evaluation): Avoid `byte-compile--` vars.
* lisp/cedet/semantic/fw.el (semantic-alias-obsolete):
Use `macroexp-compiling-p` and `macroexp-file-name`.
* lisp/cedet/semantic/wisent/comp.el (wisent-source): Use `macroexp-file-name`
(wisent-total-conflicts): Tighten regexp.
* lisp/emacs-lisp/cl-lib.el (cl--compiling-file): Delete function
and variable. Use `macroexp-compiling-p` instead.
* lisp/progmodes/flymake.el (flymake-log):
* lisp/emacs-lisp/package.el (package-get-version):
* lisp/emacs-lisp/ert-x.el (ert-resource-directory):
Use `macroexp-file-name`.
* lisp/progmodes/cc-engine.el (c-invalidate-state-cache-1): Rewrite part of
it, following the code in c-parse-state-1, to get a proper setting of
c-state-cache-good-pos.
(antlr-mode): Remove compatibility code with older CC-mode.
This somehow appears to fix the error:
In antlr-mode:
progmodes/antlr-mode.el:2426:20: Error: `c-init-language-vars' defined
after use in (c-init-language-vars) (missing `require' of a library
file?)
progmodes/antlr-mode.el:2427:26: Warning: c-init-language-vars called
with 0 arguments, but requires 1
No idea what caused the error to appear after the previous patch either.
(cond-emacs-xemacs, cond-emacs-xemacs-macfn, defunx, ignore-errors-x):
Remove those functions and macros. Replace every use with the result
of their use.
(antlr-default-directory): Remove function, use the `default-directory`
variable instead.
(antlr-read-shell-command): Remove function, use
`read-shell-command` instead.
(antlr-with-displaying-help-buffer): Remove function, by inlining it at
its only call site.
(antlr-end-of-rule, antlr-beginning-of-rule, antlr-end-of-body)
(antlr-beginning-of-body): Mark them as movement commands.
* lisp/progmodes/make-mode.el (makefile-mode-map): Move menu
definition from here...
(makefile-mode-menu): ...to here, and rewrite using easy-menu-define.
* lisp/progmodes/cperl-mode.el (cperl--package-regexp):
Avoid double repetition; cperl--ws-or-comment-regexp is already
repeated with 1+.
* test/lisp/textmodes/dns-mode-tests.el
(dns-mode-tests-dns-mode-soa-increment-serial): Escape literal '$'.
* test/lisp/emacs-lisp/rx-tests.el (rx-regexp): Modify test to not
trigger a linting warning while retaining its testing power.
(edebug-all-defs, edebug-all-forms): Don't autoload since the problem
it was working around has been fixed a while back.
(edebug--eval-defun): Rename from `edebug-eval-defun` and simplify by
making it an `:around` advice.
(edebug-install-read-eval-functions)
(edebug-uninstall-read-eval-functions): Adjust accordingly.
(edebug-eval-defun): Redefine as an obsolete wrapper.
* lisp/progmodes/elisp-mode.el (elisp--eval-defun):
Use `load-read-function` so it obeys `edebug-all-(defs|forms)`.
(elisp--eval-defun): Fix recent regression introduced with
`elisp--eval-defun-result`.
* lisp/progmodes/cperl-mode.el
(cperl-imenu-addback): Customization variable deleted. This
variable has been declared obsolete in 1998.
(cperl--basic-identifier-regexp) and many other variables:
defining regular expressions for basic Perl constructs.
(cperl-imenu--create-perl-index): This function has been
completely rewritten, keeping only some parts of the output
formatting. It now recognizes a lot more package and
subroutine declarations which came since Perl 5.14: Packages
with a version and/or a block attached, lexical subroutines,
declarations with a newline between the keyword "package" and
the package name, and several more. This version also
correctly separates subroutine names from attributes, does no
longer support "unnamed" packages (which don't exist in Perl),
and doesn't fall for false positives like stuff that looks
like a declaration in a multiline string.
(cperl-tags-hier-init): Eliminate call to
`cperl-imenu-addback` (which actually was commented out in
1997)
* test/lisp/progmodes/cperl-mode-tests.el
(cperl-test--validate-regexp) and six other new tests for the
new regular expressions and the index creation.
* test/lisp/progmodes/cperl-mode-resources/grammar.pl: New
file showcasing different syntax variations for package and
sub declarations (bug#46574).
