This allows us to reuse much of the lisp build and installation machinery,
rather than duplicating it.
* Makefile.in (abs_builddir, leimdir): Remove.
(buildlisppath, SUBDIR, COPYDIR, COPYDESTS): No more leim directory.
(epaths-force-w32): No longer set BLD.
(leim): Remove.
(install-arch-indep): No longer run or install leim.
(mostlyclean, clean): No longer run leim rule.
(bootstrap-clean): Change leim target.
(maintainer-clean): Add leim.
(check-declare): Remove leim.
* README: Update for leim changes.
* configure.ac (leimdir): Remove.
(standardlisppath): No more leimdir.
* make-dist: Update for files from leim/ now being in lisp/leim/.
* doc/lispref/loading.texi (Library Search):
* doc/lispref/os.texi (Startup Summary): No more leim directory.
* leim/Makefile.in (leimdir): New variable.
(TIT_GB, TIT_BIG5, MISC, changed.tit, changed.misc)
(${leimdir}/leim-list.el, ${leimdir}/ja-dic/ja-dic.el):
Generate in $leimdir.
(all): Remove compilation, add ja-dic.
(leim-list.el): Now PHONY.
(setwins, compile-targets, compile-main, clean, mostlyclean)
(extraclean): Remove.
(bootstrap-clean): Delete all generated files.
* leim/README: Update for moved leim/ directory.
* leim/leim-ext.el (ucs-input-activate, hangul-input-method-activate):
Remove manual autoloads; now in loaddefs.el.
Disable byte-compile, version-control, autoloads in the output.
* lisp/Makefile.in (setwins_for_subdirs): Skip leim/ directory.
(compile-main): Depend on lisp/leim rule.
(leim): New rule.
* lisp/loadup.el: Move leim-list.el to leim/ subdirectory.
* lisp/startup.el (normal-top-level): No more leim directory.
* lisp/international/ja-dic-cnv.el (skkdic-convert):
Disable version-control and autoloads in output files.
* lisp/international/titdic-cnv.el (titdic-convert, miscdic-convert):
Disable version-control and autoloads in output files.
* lisp/leim/quail: Move here from ../leim.
* lisp/leim/quail/hangul.el (hangul-input-method-activate):
Add autoload cookie.
(generated-autoload-load-name): Set file-local value.
* lisp/leim/quail/uni-input.el (ucs-input-activate): Add autoload cookie.
(generated-autoload-load-name): Set file-local value.
* nt/README.W32:
* nt/addpm.c (env_vars):
* nt/epaths.nt (PATH_LOADSEARCH, PATH_DUMPLOADSEARCH):
* nt/paths.h (PATH_LOADSEARCH): No more leim directory.
* src/Makefile.in (leimdir): Now in lisp source directory.
($(leimdir)/leim-list.el): Just use ../leim .
* src/epaths.in (PATH_DUMPLOADSEARCH):
* src/lread.c (load_path_default):
* src/nsterm.m (ns_load_path): No more leim directory.
* .bzrignore: Update for relocated leim files.
Note: I used Emacs 24.3 to generate the uni-*.el files, owing to
unexplained (possibly harmless) differences that the current trunk's
use of lexical-binding for subr.el causes. Ref:
http://debbugs.gnu.org/15825#29
* admin/unidata/BidiMirroring.txt, admin/unidata/UnicodeData.txt:
Update to 6.3.0.
* lisp/international/uni-bidi.el, lisp/international/uni-category.el:
* lisp/international/uni-name.el, lisp/international/uni-numeric.el:
Regenerate.
* test/automated/mule-util.el: New file, with tests extracted from
lisp/international/mule-util.el.
Copyright years based on when the original test-suite comment was
added to lisp/international/mule-util.el (in 2002).
* lisp/international/mule-util.el: Extract comment test-suite out to
separate file.
lisp/international/mule.el (define-coding-system): Doc fix.
etc/NEWS: Document prefer-utf-8 coding-system and the new attributes
:inhibit-null-byte-detection, :inhibit-iso-escape-detection, and
:prefer-utf-8.
properties :inhibit-null-byte-detection,
:inhibit-iso-escape-detection, and :prefer-utf-8.
(set-buffer-file-coding-system): If :charset-list property of
CODING-SYSTEM is `emacs', do not check if CODING-SYSTEM is
appropriate for setting.
* calculator.el (electric, ehelp): No need to load when compiling.
(Electric-command-loop, electric-describe-mode): Declare.
* doc-view.el (doc-view-current-converter-processes): Move before use.
* files.el (dired-unmark):
* progmodes/gud.el (gdb-input): Update declarations.
* emacs-lisp/easy-mmode.el (define-globalized-minor-mode):
Move MODE-set-explicitly definition before use.
* erc/erc-log.el (erc-network-name): Declare.
* erc/erc-notify.el (pcomplete--here): Declare.
(pcomplete/erc-mode/NOTIFY): Require pcomplete.
* international/mule-diag.el (mule-diag):
Don't use obsolete window-system-version.
* mail/feedmail.el (smtpmail): No need to load when compiling.
(smtpmail-via-smtp, smtpmail-smtp-server): Declare.
* mail/mail-utils.el (rfc822): No need to load when compiling.
(rfc822-addresses): Autoload it.
(mail-strip-quoted-names): Trivial simplification.
* mail/rmail.el (rmail-mime-message-p, rmail-mime-toggle-raw): Declare.
(rmail-retry-failure): Don't assume that rmail-mime-feature == rmailmm.
* net/snmp-mode.el (tempo): Don't duplicate requires.
* progmodes/prolog.el (info): No need to load when compiling.
(comint): Require before shell requires it.
(Info-goto-node): Autoload it.
(Info-follow-nearest-node): Declare.
(prolog-help-info, prolog-goto-predicate-info): No need to require info.
* textmodes/artist.el (picture-mode-exit): Declare.
* textmodes/reftex-parse.el (reftex-parse-from-file):
Trivial rewrite so the compiler can parse it better.
* url/url-dav.el (url-http): Require it.
(url-http-head-file-attributes): Don't autoload it.
* url/url-proxy.el (url-http): Autoload it.
* international/ja-dic-cnv.el (skkdic-collect-okuri-nasi)
(batch-skkdic-convert): Suppress most of the chatter.
It's not needed so much now that machines are faster,
and its non-ASCII component was confusing; see Dmitry Gutov in
<http://lists.gnu.org/archive/html/emacs-devel/2013-03/msg00508.html>.
ja-dic.el no longer needs to be in the repository: it's now
generated as part of the build from bzr. Also, update SKK-JISYO.L to
match the upstream source exactly.
* .bzrignore: Add leim/ja-dic/.
* leim/ja-dic/ja-dic.el: Remove from repository. It is still distributed
as part of the Emacs tarball.
* leim/Makefile.in ($(srcdir)/ja-dic/ja-dic.el): New rule.
(compile-main): Depend on it.
* leim/SKK-DIC/README: Update to reflect new build procedure.
* leim/SKK-DIC/SKK-JISYO.L: Update to match source exactly.
This is now the annotated version, to match the upstream file name;
the unannotated one is built from it automatically.
* lisp/international/ja-dic-cnv.el (skkdic-convert): Remove the annotations
from the input, rather than assume that it's been done for us by the
SKK script unannotate.awk. Switch ja-dic.el to UTF-8. Don't put
the current date into a ja-dic.el comment, as that complicates
regression testing.
Fixes: debbugs:13984
of completion-ignore-case around completing-read to fix regression
exhibited by the test case `C-x 8 RET *acc TAB' and caused by
`string-match-p' using the nil value of `case-fold-search' and
`completion-ignore-case' in `completion-pcm--all-completions'..
Fixes: debbugs:12615
* international/mule-cmds.el (read-char-by-name): Signal an error
if the user does not supply a valid character.
* editfns.c (Finsert_char): Since read-char-by-name now signals an
error for invalid chars, don't check for a nil return value.
Fixes: debbugs:13177
`isearch-insert-char-by-name'.
(with-isearch-suspended): New defmacro with body mostly from
`isearch-edit-string' except the part that sets
`isearch-new-string' and `isearch-new-message'.
(isearch-edit-string): Use new macro `with-isearch-suspended' with
body that sets `isearch-new-string' and `isearch-new-message'.
(isearch-insert-char-by-name): New command.
* lisp/international/mule-cmds.el (read-char-by-name): Let-bind
`enable-recursive-minibuffers' to t.
http://lists.gnu.org/archive/html/emacs-devel/2012-12/msg00234.html
