This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
Validator can not handle multi-byte font name. Also we can not write multi byte character in comment also.
I do see a bug in the display of the "validated" stylesheet but nothing along the lines of "can not handle" multi-byte font name. Could you give more details?
I validate following web page by latest CSS-Validator. http://hobbit.ddo.jp/ie.css.utf8 Result is as follows. Value Error : font-family Parse Error "��³ �°ææ", serif} H2 {�³ �°ææ", serif} DT {�³ �°ææ", serif} body {�³ �°ææ", serif} .min {�³ �°ææ", serif} .cjk {�³ ææ", serif} :link:hover, :visited:hover { background: #4040FA; color: #F0F8FF} .shdw {width: 100%; filter: Shadow(color=#C0C0C0, direction=135)} Contents of validated CSS file is as follows. @charset "utf-8"; H1 { font-family: "Times New Roman", Times, "lr o¾©", serif} H2 { font-family: "Times New Roman", Times, "lr o¾©", serif} DT { font-family: "Times New Roman", Times, "lr o¾©", serif} body { font-family: "Times New Roman", Times, "lr o¾©", serif} .min { font-family: "lr o¾©", serif} .cjk { font-family: "lr ¾©", serif} :link:hover, :visited:hover { background: #4040FA; color: #F0F8FF} .shdw {width: 100%; filter: Shadow(color=#C0C0C0, direction=135)} If CSS Validator read the name of font that contains multi-byte characters, CSS Validator could not accept it, and abort.
This leaves me confused. Neither the last stable release of the validator nor the development instance (using the cvs HEAD) behave in the way you describe. http://qa-dev.w3.org:8001/css-validator/validator?uri=http%3A%2F%2Fhobbit.ddo.jp%2Fie.css.utf8 http://jigsaw.w3.org/css-validator/validator?uri=http%3A%2F%2Fhobbit.ddo.jp%2Fie.css.utf8
(In reply to comment #3) > This leaves me confused. Neither the last stable release of the validator nor > the development instance (using the cvs HEAD) behave in the way you describe. Yves suggests that the difference in behavior might come from locale setup differences between your system and the one on w3.org. AFAICT on the w3 servers the servlet is running on a system with the locale LANG=en_US.UTF-8 is enabled.
I run CSS validator on Windows XP of Japanese version. Do you mean that CSS Validator only run in English language environment?
(In reply to comment #5) > I run CSS validator on Windows XP of Japanese version. > Do you mean that CSS Validator only run in English language environment? Obviously, no. It should work nicely whatever your system language is, but apparently your system (java install?) does not decode utf-8 encoded stylesheets in the same way as our (linux or MacOSX) test systems do. Admittedly this is way beyond my knowledge of how java and environment/system interact. I am copying Yves in case he has more to say on the issue.
Created attachment 626 [details] Japnese version read CP932 page properly As screenshot attached, Japanese version of Java read CP932 page properly and convert that to utf-8 properly. Maybe CSS Validator do not specify input characterset explicitly.
Created attachment 627 [details] Sample file of CSS of CP932(Shift_jis) This is sample file of CSS of CP932(Shift_jis). Java of Japanese environment read this properly.
Created attachment 628 [details] Sample of CSS of UTF-8 This is sample of CSS of UTF-8. Both English and Japnese JAVA cannot read this properly.
The behaviour difference between two installation were almost probably due to the default charset of the java installation. The recent patches made should work for charset definition sent with the Content-Type header, or the BOM if present in UTF-8 streams, next step is obviously to switch charsets based on the @charset rule. http://qa-dev.w3.org:8001/css-validator/validator?uri=http%3A%2F%2Fhobbit.ddo.jp%2Fie.css.utf8 should be displayed properly now. (bug not closed, as it is pending @charset full support)
Should work now with @charset even if no charset was defined using in the Content-Type HTTP header. See http://qa-dev.w3.org:8001/css-validator/validator?uri=http%3A%2F%2Fwww.w3.org%2FBugs%2FPublic%2Fattachment.cgi%3Fid%3D627&profile=css21&usermedium=all&warning=1
Created attachment 629 [details] Sample of CSS file of UTF-8 with BOM "By Upload", attached file is not read properly.
Created attachment 630 [details] Sample of CSS file of UTF-8 without BOM "By Upload", attached file could not be read properly.
It should work properly now with text input and file upload.
Confirmed in Japanese language environment also.
last fix done for no charset declaration and BOM in uploaded file. Closing now, thank you for all the tests !