* doc/lispref/os.texi (Time Calculations):
Document time-equal, and the behavior on NaNs and infinities of
time-less-p, time-add, time-subtract.
* etc/NEWS: Mention the change.
* src/editfns.c (time_arith): Change last arg from function
to bool. All callers changed. Do the right thing with
infinities and NaNs.
(time_cmp): New function, which handlesx infinities and NaNs.
(Ftime_less_p): Use it.
(Ftime_equal): New function.
* test/lisp/emacs-lisp/timer-tests.el (timer-test-multiple-of-time):
Use it.
* lisp/progmodes/cc-engine.el (c-forward-over-token): New function, extracted
from ...
(c-forward-over-token-and-ws): Refactor to use the above.
(c-forward-type): Use c-check-qualified-type in place of c-check-type
(twice).
(c-forward-over-compound-identifier): New function.
(c-check-qualified-type): New function.
These changes remove some assumptions of Lisp code on timestamp
format. Although we’re not going to change the default format any
time soon, I went looking for code that was too intimate about
details of timestamp format and removed assumptions where this was
easy to do with current Emacs primitives.
* lisp/ido.el (ido-wash-history):
Fix test for zero timestamp.
* lisp/time.el (display-time-event-handler):
Use time-less-p rather than doing it by hand.
(display-time-update): Simplify by using float-time
instead of doing the equivalent by hand.
* lisp/url/url-auth.el (url-digest-auth-make-cnonce):
* test/lisp/calendar/parse-time-tests.el (parse-time-tests):
* test/lisp/emacs-lisp/timer-tests.el (timer-test-multiple-of-time):
* test/lisp/net/tramp-tests.el:
(tramp-test19-directory-files-and-attributes)
(tramp-test22-file-times, tramp-test23-visited-file-modtime):
Don’t assume detailed format of returned Lisp timestamps.
* src/bignum.c (mpz_bufsize): New function.
(bignum_bufsize): Use it.
(mpz_get_d_rounded): New function.
(bignum_to_double): Use it.
* src/bignum.c (bignum_to_double):
* src/data.c (bignum_arith_driver):
When converting bignums to double, round instead of
truncating, to be consistent with what happens with fixnums.
* test/src/floatfns-tests.el (bignum-to-float): Test rounding.
The main motivation for this change was the introduction of
project-query-replace. dired's multi-file query&replace was implemented
on top of etags.el even though it did not use TAGS in any way, so I moved
this generic multifile code into its own package, with a nicer interface,
and then used that in project.el.
* lisp/progmodes/project.el (project-files): New generic function.
(project-search, project-query-replace): New commands.
* lisp/dired-aux.el (dired-do-search, dired-do-query-replace-regexp):
Use multifile.el instead of etags.el.
* lisp/progmodes/etags.el: Remove redundant :groups.
(next-file-list): Remove var.
(tags-loop-revert-buffers): Make it an obsolete alias.
(next-file): Don't autoload (it can't do anything useful before some
other etags.el function setup the multifile operation).
(tags--all-files): New function, extracted from next-file.
(tags-next-file): Rename from next-file.
Rewrite using tags--all-files and multifile-next-file.
(next-file): Keep it as an obsolete alias.
(tags-loop-operate, tags-loop-scan): Mark as obsolete.
(tags--compat-files, tags--compat-initialize): New function.
(tags-loop-continue): Rewrite using multifile-continue. Mark as obsolete.
(tags--last-search-operate-function): New var.
(tags-search, tags-query-replace): Rewrite using multifile.el.
* lisp/emacs-lisp/generator.el (iter-end-of-sequence): Use 'define-error'.
(iter-make): New macro.
(iter-empty): New iterator.
* lisp/menu-bar.el (menu-bar-search-menu, menu-bar-replace-menu):
tags-loop-continue -> multifile-continue.
Formerly nil meant both that DST was not in effect and that
the DST flag was unknown, and different functions interpreted
the flag differently. Now the meaning is consistently nil for
DST not in effect, and -1 for DST flag not known.
* doc/lispref/os.texi (Time Conversion): The DST slot is
now three-valued, not two-.
* doc/misc/emacs-mime.texi (time-date): Adjust to new behavior.
* etc/NEWS: Mention this.
* lisp/calendar/parse-time.el (parse-time-string):
* src/editfns.c (Fdecode_time):
Return -1 for unknown DST flag.
* test/lisp/calendar/parse-time-tests.el (parse-time-tests):
Adjust tests to match new behavior, and add a new
test for nil vs -1.
* doc/lispref/processes.texi (Bindat Examples): Remove, fixing
a FIXME in the manual. The long example had bitrotted to some
extent, compared to the more-up-to-date example in bindat.el
commentary, which apparently what people were referring to
anyway. The short example was confusing and not that useful
and will be obsolescent anyway if we change timestamp format.
Suggested by Stefan Monnier here:
https://lists.gnu.org/archive/html/emacs-devel/2018-09/msg00783.html
* lisp/replace.el (occur--parse-occur-buffer): Since point is at the
beginning of the buffer, use `point'.
(occur-revert-function): Prefer `pcase-let' and `point-min'.
Check whether `region-start' or `region-end' are non-nil.
d28d54c (origin/emacs-26) More accurate docs for 'text-char-description'
b3baf99 Document synchronous behavior of eshell/make (Bug#32513)
98544ea Fix bs-show with wide characters (Bug#17822)
85af51b Improve Custom menu labels for 2 options
72a2a36 Improve wording of last change in dired-x.texi
d4fa83b Fix GnuTLS test suite with GnuTLS versions 3.4.x
b5bee6b Fix build with gnutls versions 3.0 to 3.2 (Bug#32446)
67eb80e ; * etc/enriched.txt (hanging-indents): Remove extra indent.
c71cfb7 Fix the Bubbles game on TTY frames
3bbf21b Add choice to reshow certificate information (Bug#31877)
6f2c471 * src/alloc.c (Fbool_vector, Flist, Fvector): Doc tweak.
39eecb3 * src/alloc.c (vector): Fix grammatical error in doc string: ...
* src/keymap.c (Ftext_char_description):
* doc/lispref/help.texi (Describing Characters): More accurate
description of 'text-char-description'. Remove incorrect
examples from the ELisp manual. (Bug#32743)
* doc/misc/eshell.texi (Built-ins):
* lisp/eshell/em-unix.el (eshell/make): Mention that it falls back to
the external 'make' command when called synchronously.
Store the region and orig line into the *Occur* header line.
Retrieve this information in `occur-revert-function'.
* lisp/replace.el (occur--parse-occur-buffer): New defun.
(occur-revert-function): Use it.
(occur-engine): Store region and original position as text properties
into the *Occur* header line.
* lisp/replace.el (occur-engine): Add sensible default values for
(occur--orig-line and nlines.
* src/gnutls.c (gnutls_cipher_get_tag_size): Make it return
zero only for versions of GnuTLS < 3.2.2, where
gnutls_cipher_get_tag_size was introduced. This fixes the
GnuTLS test suite, which assumes that any cipher whose tag
size is non-zero is AEAD-capable, and doesn't test such ciphers
if AEAD is not available, i.e. for GnuTLS < 3.5.1. (Bug#32446)
We previously used functions available only in 3.2+ for all 3.x
versions.
* src/gnutls.c [GNUTLS_VERSION_NUMBER < 0x030501]: Replace calls to
gnutls_cipher_get_tag_size with 0.
[GNUTLS_VERSION_NUMBER < 0x030200]: Alias gnutls_cipher_get_iv_size
to gnutls_cipher_get_block_size, gnutls_digest_list to
gnutls_mac_list, and gnutls_digest_get_name to gnutls_mac_get_name.
[WINDOWSNT]: Adjust DLL function definitions and declarations
accordingly.
(ad-get-advice-info): Mark it inlinable.
(ad-get-advice-info-macro): Make it an obsolete alias.
(ad-copy-advice-info, ad-is-advised, ad-get-advice-info-field)
(ad-find-advice, ad-macrofy, ad-lambdafy, ad-lambda-p, ad-advice-p)
(ad-compiled-p, ad-compiled-code, ad-get-cache-definition)
(ad-get-cache-id, ad-set-cache): Turn macros into defsubsts.
(ad-defadvice-flags): Make it into a plain list.
(ad-set-advice-info-field): Apply a bit of CSE.
(ad-read-advised-function, ad-do-advised-functions): Adjust accordingly.
(ad-pushnew-advised-function, ad-pop-advised-function): Also make them
into functions.
* lisp/play/bubbles.el (bubbles--col-offset)
(bubbles--row-offset): Doc fixes.
(bubbles--compute-offsets): Conflate the GUI and TTY code into
a single common version. Set the offsets to simple numbers,
not to lists.
(bubbles--initialize, bubbles--show-scores): Wrap offset
values in a list, so that they are interpreted as pixel
values, not as units of character width. This fixes the game
on TTY frames. (Bug#32744)
This change was motivated by the desire to remove the weird
dependency of lib-src/profile.o on src/systime.h. profile.c
included systime.h only for current_timespec, and this
inclusion required systime.h to have #ifdef emacs in multiple
places and complicated further changes I have in mind.
The current_timespec decl belongs in timespec.h anyway,
and the main effect of this change is to move it there.
* lib-src/profile.c (INLINE): Remove.
Include timespec.h, not systime.h.
* lib/gettime.c (gettime): Prefer clock_gettime to nanotime,
and don’t worry about it failing on a CLOCK_REALTIME arg.
POSIX requires it to succeed and I don’t know of any
counterexamples where the fallbacks would work.
(current_timespec): Move here from src/systime.h.
Nowadays it seems to be better to not have this function
be inline.
* lib/timespec.h: Include arg-nonnull.h.
(current_timespec): New declaration.
(gettime, settime): Declare args to be nonnull.
* lib/gettime.c, lib/timespec.h: Copy from Gnulib.
* src/systime.h: Simplify by assuming ‘emacs’ is defined,
which it always is now.
(current_timespec): Move to lib/timespec.h.
In various situations, the window displaying the certificate
information can be hidden (such as if the user accidentally presses ?,
which causes the read-multiple-choice help window to replace it).
Instead of leaving the user to make a choice blindly, add a choice to
reshow the certification information.
* lisp/net/nsm.el (nsm-query-user): Add reshow choice.
* lisp/cedet/ede/pconf.el (ede-proj-configure-synchronize): Check the
first list element of mode-line-process (Bug#32564).
Copyright-paperwork-exempt: yes
* src/atimer.c (start_atimer, debug_timer_callback):
Don’t assume support for setitimer merely because struct
itimerspec works. POSIX no longer requires support for the
obsolescent setitimer function.
* test/lisp/calendar/icalendar-tests.el (icalendar--create-uid):
Do not intrude into or rely upon undocumented internal
implementation details of icalendar--create-uid.
Problem reported by Glenn Morris in:
https://lists.gnu.org/r/emacs-devel/2018-09/msg00660.html
This is a mess. This function was added in master in afba4ccb8b,
undocumented. It was then tweaked and documented in 9824885fab.
Much later in cc233365a9 the original untweaked and undocumented
version was backported to emacs-26. This was independently
redocumented in emacs-26 in cc8f334d2d, which conflicted with the
interface changes that had been made on master.
