* lisp/wid-edit.el (widget-specify-field): Store the end overlay
that we sometimes create for an editable-field widget.
(widget-field-value-delete): Make sure we delete all overlays
that belong to the widget. (Bug#75646)
* test/lisp/wid-edit-tests.el (widget-test-delete-field-overlays): New
test.
Starting with Emacs 28, filling strings now happens in a
narrowed scope, and looses the leading indentation and can cause
the string to extend past the fill-column value. Introduce
`lisp-fill-paragraph-as-displayed' as a new variable allowing
opting out of this new behavior in specific scenarios (such as
when using the Scheme major mode, say).
* lisp/emacs-lisp/lisp-mode.el
(lisp-fill-paragraph-as-displayed): New variable.
(lisp-fill-paragraph): Honor it, by avoiding the logic narrow to
strings before applying fill-paragraph.
* test/lisp/emacs-lisp/lisp-mode-tests.el
(lisp-fill-paragraph-respects-fill-column): Test it.
(lisp-fill-paragraph-docstring-boundaries): New test, as a
safeguard to avoid regressions.
Fixes: bug#56197
* lisp/time-stamp.el: (time-stamp-string-preprocess): Rename 'alt-form'
to 'colon-cnt' because it is now an integer.
* test/lisp/time-stamp-tests.el: Rename 'formatz-generate-tests' to
'define-formatz-tests'.
* test/lisp/progmodes/eglot-tests.el
(eglot-test-snippet-completions-with-company): Don't expect fooba to be
there, do expect two completions to be candidates.
* test/src/emacs-tests.el (emacs-tests/bwrap/allows-stdout):
Skip test if bwrap prohibits even "echo Hi",
which it does on my Ubuntu 24.10 platform.
This extra check means we no longer need to test
separately whether the filter is readable.
Emacs partial completion simply doesn't make sense in LSP.
Attempts to make it make some sense end up failing for one
reason or another.
This commit restores the original intention of the
eglot--dumb-allc and eglot--dumb-tryc functions, which was to be
dumb and pop up a *Completions* buffer (or a company tooltip),
as soon as there is doubt over what the user wants to do.
It also fixes tests, including an expensive Rust test that had
been broken for a long time.
* lisp/progmodes/eglot.el (eglot--dumb-allc): Consider point.
(eglot--dumb-tryc): Make it suitably dumb again.
* test/lisp/progmodes/eglot-tests.el
(eglot--wait-for-rust-analyzer): Wait longer.
(eglot-test-common-prefix-completion): Delete this test. It's not
the intended behavior.
(eglot--kill-completions-buffer): New helper.
(eglot-test-try-completion-nomatch): Rework test.
(eglot-test-try-completion-inside-symbol): Pops *Completions*
buffer because mustn't partial complete.
(eglot-test-try-completion-inside-symbol-2): Does complete.
(eglot-test-rust-completion-exit-function): Fix long-broken test.
853719c4c2 ; * lisp/net/eww.el (eww-download): Doc fix (bug#75585)
45ec5865aa ; * etc/NEWS: Tweak wording of NSM items.
b981889e9e ; cperl-mode-tests.el: Don't run the newest test in perl-...
6a864ef8f3 Add smtpmail cross-reference to 'auth-sources'.
1fd7957bc7 ; cperl-mode.el: Add a test for Bug#74245
8d289670d6 Merge branch 'emacs-30' of git.savannah.gnu.org:/srv/git/...
b74ac4af94 ; cperl-mode.el: Allow bare $ in a signature (Bug#74245)
30e84fc653 Emphasize the use of :tag for new customization types
# Conflicts:
# etc/NEWS
* lisp/treesit.el (treesit--imenu-merge-entries):
(Ftreesit_parser_set_included_ranges): Warn in the docstring
that the function is destructive/owns the argument.
* test/src/treesit-tests.el (treesit-range-fixup-after-edit):
(treesit-imenu): Fix tests.
* lisp/repeat.el (repeat-get-map-sym): New function
refactored from 'repeat-get-map'.
(repeat-get-map): Move continue-related code to 'repeat-get-map-sym'.
(repeat-pre-hook): Use 'repeat-get-map-sym' and 'repeat-get-map'.
Set 'repeat-map' to 'map-sym'.
(repeat-post-hook): Use 'repeat-get-map-sym'.
* test/lisp/repeat-tests.el (repeat-tests-continue-another):
Improve to invoke double key 'C-M-o C-M-o' that should not switch
between maps (bug#74140).
Further improve flymake-mode idempotency by not nuke existing overlays.
This means multiple flymake-mode invocations do the same as just one
one, with minimal or no additional side effects. This is good for
people with lots of 'flymake-mode' in hooks.
The foreign diagnostic importation has been refactored into a separate
function and moved to the "really start" section of 'flymake-start'.
The duplication problem appears to be avoided by some heuristics in
flymake-highlight-line.
A new test has been added.
* lisp/progmodes/flymake.el (flymake--import-foreign-diagnostics): New helper
(flymake-start): Use it.
(flymake-mode): Don't nuke overlays here.
* test/lisp/progmodes/flymake-tests.el (foreign-diagnostics): New
test.
* test/lisp/progmodes/cperl-mode-resources/cperl-bug-74245.pl: New test data.
* test/lisp/progmodes/cperl-mode-tests.el (cperl-test-bug-74245): Verify
that a bare \"$\" can appear at the end of a subroutine signature.
* lisp/progmodes/cperl-mode.el (cperl--signature-rx): Allow bare
sigils for unused parameters in signatures.
(cperl-find-pods-heres): Avoid $) at the end of a signature being
treated as the punctuation variable $) by treating this dollar as
punctuation
* test/lisp/progmodes/cperl-mode-tests.el
(cperl-test-signature-rx): Add ($first,$) as a valid signature,
remove ($) from the list of invalid signatures.
* lisp/progmodes/c-ts-mode.el (c-ts-mode--simple-indent-rules):
Use standalone-parent instead of parent.
* test/lisp/progmodes/c-ts-mode-resources/indent.erts:
New test.
* test/lisp/erc/erc-scenarios-keep-place-indicator.el
(erc-scenarios-keep-place-indicator--follow): Don't run on EMBA pending
investigation of test timing out.
* etc/ERC-NEWS: Announce deprecation of `erc-spelling-flyspell-verify'
and `erc-spelling-unhighlight-word'. A slight behavioral change not
worth mentioning is that, previously, ERC arranged for Flyspell to
ignore any word immediately following a forward slash anywhere in the
prompt input, even those for which the slash served as mere
punctuation (a "stroke"), as in "something/misspelt." As of this
change, Flyspell only unconditionally exempts an initial slash-prepended
word, like "tableflip" in "ERC> /tableflip", and checks all others that
follow against known slash commands.
* lisp/erc/erc-spelling.el: Change top-level assignment of `erc-mode'
symbol-property `flyspell-mode-predicate' from
`erc-spelling-flyspell-verify' to `erc-spelling--flyspell-input-p'.
(erc-spelling-mode, erc-spelling-disable): Remove local member from
`flyspell-incorrect-hook'.
(erc-spelling-init): Add `erc-spelling--flyspell-check' to
`flyspell-incorrect-hook' locally. Don't bother explicitly setting
`flyspell-generic-check-word-predicate' because Flyspell already does
that for clients using the `flyspell-mode-predicte' interface.
(erc-spelling-flyspell-verify, erc-spelling-unhighlight-word): Mark
obsolete.
(erc-spelling--flyspell-check, erc-spelling--flyspell-input-p): New
functions, essentially the two halves of a reworked and bifurcated
`erc-spelling-flyspell-verify'. Though used as a predicate, the first
is not named as such because it performs side effects.
* test/lisp/erc/erc-scenarios-spelling.el: New file.
* test/lisp/erc/resources/spelling/auto-correct.eld: New file.
(Bug#75327)
* lisp/erc/erc-nicks.el (erc-nicks-colors): Tweak doc.
(erc-nicks--create-pool-function): On graphic displays, set value to
`erc-nicks--create-culled-pool', the original default. This fixes a bug
in which ERC mapped all pool members to a primary color, thus excluding
the bulk of them. Thanks to Trevor Arjeski for discovering it.
(erc-nicks--create-coerced-pool): Mention the text-terminal-only
requirement in doc.
(erc-nicks-refresh): Improve doc.
* test/lisp/erc/erc-nicks-tests.el (erc-nicks-tests--track-faces):
Enable `erc-track-mode' prior to running body.
The assumption here is that if there is a C compiler, there is also a
math.h header somewhere. The test should fail if we can't find that
file, and hopefully that will provoke users to create bug reports.
Let's see how far we can take this idea; we might have to give up and
disable the test in some configurations. But doing that now seems
premature, even if we had a list of affected systems (which we don't).
* lisp/man.el (man--find-header-file): Factor out new function...
(Man-view-header-file): ...from here.
* test/lisp/man-tests.el (man-tests-find-header-file): New test.
This fixes following "#include" references in 'M-x man' buffers, and
using ffap, both on macOS machines, and on systems where for some reason
clang is available but gcc is not.
* lisp/subr.el (internal--c-header-file-path): Fix finding C
headers with clang.
(internal--gcc-is-clang-p): New function factored out from...
* lisp/emacs-lisp/ert-x.el (ert-gcc-is-clang-p): ...here.
* lisp/man.el (Man-header-file-path): Bump :version tag.
* test/lisp/subr-tests.el
(subr-tests-internal--c-header-file-path/clang-mocked): New test.
It is not clear to me where this function properly belongs, so let's put
it in subr.el for now. This avoids code duplication without introducing
a dependency between man and ffap. It can always be moved later.
* lisp/subr.el (internal--c-header-file-path): New function.
* lisp/man.el (Man-header-file-path):
* lisp/ffap.el (ffap-c-path): Use above new function.
* test/lisp/subr-tests.el (ert-x): Require.
(subr-tests-internal--c-header-file-path)
(subr-tests-internal--c-header-file-path/gcc-mocked): New tests.
* lisp/time-stamp.el (time-stamp-string-preprocess): new formats for
lowercase weekday name, lowercase month name, and system name
* test/lisp/time-stamp.el: new tests
(formatz-generate-tests): Better doc strings for generated test fns.
* test/lisp/dired-tests.el (dired-test-bug27243-02): Don't rely on the
expected point, since columns will move if the size of the parent directory
changes order of magnitude. (Bug#72120)
On the latest update of the grammar expressions were added and won't be
seen as directives anymore.
* lisp/progmodes/heex-ts-mode.el
(heex-ts--sexp-regexp): Match on expression as well.
(heex-ts--indent-rules): Indent on expression end. (Bug#75191)
* test/lisp/progmodes/heex-ts-mode-resources/indent.erts: Add expression
indent test.
* lisp/net/tramp-sh.el (tramp-sh-handle-process-file): Do proper quoting.
* lisp/net/tramp.el (tramp-string-empty-or-nil-p):
Declare `tramp-suppress-trace' property.
(tramp-skeleton-make-symbolic-link): Drop volume letter when flushing.
* test/lisp/net/tramp-tests.el (tramp-test39-make-lock-file-name):
Adapt test.
(tramp-test42-utf8): Do not run expensive tests on MS Windows.
* doc/misc/erc.texi (Integrations): Set `erc-server-reconnect-function'
to `erc-server-delayed-check-reconnect' in SOCKS example, and add
definition for `erc-open-socks-tls-stream'. Mention possible
inaccuracies related to error detection with certain reconnect
strategies.
* lisp/erc/erc-backend.el (erc-server--reconnect-opened): New function.
(erc-server-delayed-check-reconnect): Attempt to reuse process if server
sends a complete PONG, and attempt to accommodate connectors that set
:nowait to nil.
(erc--server-delayed-check-connectors): Remove variable.
(erc-server-prefer-check-reconnect): Inline what was the internal
variable `erc--server-delayed-check-connectors' because it's no longer
used in unit tests. Add `erc-open-socks-tls-stream' to the set of
connector functions thought to be compatible with the "check" reconnect
strategy.
* test/lisp/erc/erc-scenarios-base-auto-recon.el
(erc-scenarios-base-auto-recon-no-proto): Adapt to expect "reuse" behavior.
* test/lisp/erc/resources/base/reconnect/ping-pong.eld: Delete unused file.
* test/lisp/erc/resources/base/reconnect/unexpected-disconnect.eld:
Capture PING cookie to send back to client.
* test/lisp/erc/resources/erc-d/resources/proxy-solo.eld: Delete unused
file. (Bug#62044)
* lisp/uniquify.el (uniquify-after-kill-buffer-flag)
(uniquify-trailing-separator-flag): Rename from
'uniquify-after-kill-buffer-p' and 'uniquify-trailing-separator-p', to
better adhere to our conventions. Update all users and leave old names
as obsolete variable aliases.
Now we support setting up Imenu for multiple languages
* doc/lispref/modes.texi: Update manual.
* lisp/treesit.el:
(treesit-aggregated-simple-imenu-settings): New variable.
(treesit--imenu-merge-entries): New function.
(treesit--generate-simple-imenu): This was previously
treesit-simple-imenu.
(treesit-simple-imenu): Support
treesit-aggregated-simple-imenu-settings.
(treesit-major-mode-setup): Recognize
treesit-aggregated-simple-imenu-settings.
* test/src/treesit-tests.el (treesit-imenu): New test.
* lisp/files.el (auto-mode-alist): Match more gdbinit files,
including XDG, and MS-Windows. Avoid false positives.
(set-auto-mode--find-matching-alist-entry): Break out function...
(set-auto-mode--apply-alist): ...from here. (Bug#74946)
* test/lisp/files-tests.el (files-tests--check-mode): New function.
(files-tests-auto-mode-alist): New test.
* lisp/emacs-lisp/cl-macs.el (cl-flet, cl-labels): Wrap function
bodies in `cl-block`.
* test/lisp/emacs-lisp/cl-macs-tests.el (cl-macs--test-flet-block): New test.
* doc/misc/erc.texi (Sample Configuration): Remove customization in
`use-package' declaration for `erc-server-reconnect-function' as well as
related language in the customization walk-through. Do this because the
new default incorporates `erc-server-delayed-check-reconnect' behavior
for compatible connect functions.
* etc/ERC-NEWS: Announce new default for `erc-server-reconnect-function'.
* lisp/erc/erc-backend.el (erc-server-reconnect-function): Change
default to `erc-server-prefer-check-reconnect'.
(erc-server-delayed-check-reconnect): Use `process-send-string' instead
of `send-string'.
(erc--server-delayed-check-connectors): New variable.
(erc-server-prefer-check-reconnect): New function.
* test/lisp/erc/erc-scenarios-base-auto-recon.el
(erc-scenarios-base-auto-recon-unavailable)
(erc-scenarios-base-auto-recon-no-proto): Remove unnecessary
`erc-server-reconnect-function' binding because the new default
incorporates the behavior being tested for.
* test/lisp/erc/erc-scenarios-base-buffer-display.el
(erc-scenarios-base-buffer-display--reconnect-common):
* test/lisp/erc/erc-scenarios-base-compat-rename-bouncer.el
(erc-scenarios-common--base-compat-no-rename-bouncer):
* test/lisp/erc/erc-scenarios-base-netid-bouncer-recon-base.el
(erc-scenarios-base-netid-bouncer--recon-base):
* test/lisp/erc/erc-scenarios-base-netid-bouncer-recon-both.el
(erc-scenarios-base-netid-bouncer--recon-both):
* test/lisp/erc/erc-scenarios-base-netid-bouncer-recon-id.el
(erc-scenarios-base-netid-bouncer--reconnect-id-foo)
(erc-scenarios-base-netid-bouncer--reconnect-id-bar):
* test/lisp/erc/erc-scenarios-base-reconnect.el
(erc-scenarios-base-reconnect-timer)
(erc-scenarios-base-cancel-reconnect):
* test/lisp/erc/erc-scenarios-services-misc.el
(erc-scenarios-services-misc--reconnect-retry-nick):
* test/lisp/erc/erc-scenarios-stamp.el
(erc-scenarios-stamp--date-mode/reconnect): Explicitly bind
`erc-server-reconnect-function' to `erc-server-delayed-reconnect', the
former default, which does not do any probing. (Bug#62044)
* lisp/erc/erc-backend.el (define-erc-response-handler): Add
`doc-string' to `declare' specification.
* test/lisp/erc/resources/erc-tests-common.el
(erc-tests-common-add-imenu-expressions): New function for defining and
undefining Imenu patterns useful to ERC developers.
When using certain VPN clients, the interface info returned by
'getifaddrs' does not have the address family filled in for the
ifa_netmask component (this is allowed by the standard, since
the value is not specified). The resulting address info causes
network tests suite failures. Fix by copying the address family
from the returned interface address.
* src/process.c (network_interface_list): Copy the interface address
sa_family to the netmask address.
(Fnetwork_lookup_address_info): Fix test for non-IP addresses to
properly account for IPv6.
* test/src/process-tests.el (process-tests-check-bug-74907): New
test, checks that 'network-interface-list' output is as
expected.
(Bug#74907)
