Mercurial > hg > xemacs-beta
changeset 4619:75e7ab37b6c8
Fix query-coding-tests.el failures, non-mule ELCs, mule build.
lisp/ChangeLog addition:
2009-02-17 Aidan Kehoe <kehoea@parhasard.net>
* unicode.el (unicode-query-coding-region):
Avoid eval-when-compile where that gives incorrect results when
compiled on a non-Mule XEmacs but run on a Mule XEmacs. Fixes
problems seen by
20a807210902131835o2f25930apaffabb6b62a95e5f@mail.gmail.com ,
thank you Vin!
(An equivalent way to get the efficiency of the eval-when-compile
would be to provide two defvars, evaluated at dump time, but this
code will be in C soon enough, and we can use the relevant
preprocessor constants there.)
| author | Aidan Kehoe <kehoea@parhasard.net> |
|---|---|
| date | Tue, 17 Feb 2009 08:35:13 +0000 |
| parents | 69cd8e2d2a88 |
| children | 4dc42d1fe684 |
| files | lisp/ChangeLog lisp/unicode.el |
| diffstat | 2 files changed, 23 insertions(+), 8 deletions(-) [+] |
line wrap: on
line diff
--- a/lisp/ChangeLog Sun Feb 15 12:32:48 2009 +0000 +++ b/lisp/ChangeLog Tue Feb 17 08:35:13 2009 +0000 @@ -1,3 +1,16 @@ +2009-02-17 Aidan Kehoe <kehoea@parhasard.net> + + * unicode.el (unicode-query-coding-region): + Avoid eval-when-compile where that gives incorrect results when + compiled on a non-Mule XEmacs but run on a Mule XEmacs. Fixes + problems seen by + 20a807210902131835o2f25930apaffabb6b62a95e5f@mail.gmail.com , + thank you Vin! + (An equivalent way to get the efficiency of the eval-when-compile + would be to provide two defvars, evaluated at dump time, but this + code will be in C soon enough, and we can use the relevant + preprocessor constants there.) + 2009-02-02 Aidan Kehoe <kehoea@parhasard.net> * descr-text.el (describe-char-unicodedata-file):
--- a/lisp/unicode.el Sun Feb 15 12:32:48 2009 +0000 +++ b/lisp/unicode.el Tue Feb 17 08:35:13 2009 +0000 @@ -640,6 +640,14 @@ (ranges (make-range-table)) (looking-at-arg (concat "[" skip-chars-arg "]")) (case-fold-search nil) + (invalid-sequence-lower-unicode-bound + (char-to-unicode + (aref (decode-coding-string "\xd8\x00\x00\x00" + 'utf-16-be) 3))) + (invalid-sequence-upper-unicode-bound + (char-to-unicode + (aref (decode-coding-string "\xd8\x00\x00\xFF" + 'utf-16-be) 3))) fail-range-start fail-range-end char-after failed extent char-unicode failed-reason previous-failed-reason) (save-excursion @@ -662,16 +670,10 @@ ;; If the character is in the Unicode range that ;; corresponds to an invalid octet, we want to ;; treat it as unencodable. - (<= (eval-when-compile - (char-to-unicode - (aref (decode-coding-string "\xd8\x00\x00\x00" - 'utf-16-be) 3))) + (<= invalid-sequence-lower-unicode-bound char-unicode) (<= char-unicode - (eval-when-compile - (char-to-unicode - (aref (decode-coding-string "\xd8\x00\x00\xFF" - 'utf-16-be) 3)))) + invalid-sequence-upper-unicode-bound) (setq failed-reason 'invalid-sequence))) (or (null previous-failed-reason) (eq previous-failed-reason failed-reason)))