Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Inconsistent Interpretation of JSClass.name
15 views
Skip to first unread message
Masquerade
May 13, 2012, 8:20:09 AM5/13/12
to
This looks like a bug but I am not able to confirm it. I describe
what I saw here and let other experts to comment on it:
Currently, after defining JS_C_STRINGS_ARE_UTF8 or calling
JS_SetCStringsAreUTF8 (), strings (char* ) are considered as UTF8
encoded values. The JSClass.name is const char* type, and the way to
get a Unicode class name is probably by defining
JS_C_STRINGS_ARE_UTF8/ calling JS_SetCStringsAreUTF8 () and supply a
UTF8 enclosed value for the field. I've done some tests. This seems
to work and I am able to create my own classes having non-ASCII
Unicode names.
However, there are some standard functions interpreting this field as
plain ASCII only. and does not decode it as UTF8 string. For example,
standard Object.prototype.toString() implementation (obj_toString()
method in jsobj.cpp) interpret this Class.name field as plain ASCII
value and return it directly.
What should be the proper way to create a non-ASCII Unicode class
name? Should we make sure all standard functions interpret the
JSClass.name as UTF-8 string when JS_C_STRINGS_ARE_UTF8 is defined?
what I saw here and let other experts to comment on it:
Currently, after defining JS_C_STRINGS_ARE_UTF8 or calling
JS_SetCStringsAreUTF8 (), strings (char* ) are considered as UTF8
encoded values. The JSClass.name is const char* type, and the way to
get a Unicode class name is probably by defining
JS_C_STRINGS_ARE_UTF8/ calling JS_SetCStringsAreUTF8 () and supply a
UTF8 enclosed value for the field. I've done some tests. This seems
to work and I am able to create my own classes having non-ASCII
Unicode names.
However, there are some standard functions interpreting this field as
plain ASCII only. and does not decode it as UTF8 string. For example,
standard Object.prototype.toString() implementation (obj_toString()
method in jsobj.cpp) interpret this Class.name field as plain ASCII
value and return it directly.
What should be the proper way to create a non-ASCII Unicode class
name? Should we make sure all standard functions interpret the
JSClass.name as UTF-8 string when JS_C_STRINGS_ARE_UTF8 is defined?
Masquerade
May 13, 2012, 8:20:14 AM5/13/12
to
Dave Mandelin
May 21, 2012, 3:42:54 PM5/21/12
to
Dave
Masquerade
May 27, 2012, 3:36:22 AM5/27/12
to
>
> - Show quoted text -
I am using the Spidermonkey source code from FF10 release.
Tom Schuster
May 27, 2012, 6:13:12 AM5/27/12
to Masquerade, dev-tech-js-en...@lists.mozilla.org
Yes I changed the code for that in
https://bugzilla.mozilla.org/show_bug.cgi?id=700169, which is tagged
mozilla-11, so only slightly missed the esr train. :(
But hey it's going to be fixed in the next esr version, for now you
could easily patch this manually. Just look at the patch in bug
700169.
> _______________________________________________
> dev-tech-js-engine-internals mailing list
> dev-tech-js-en...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-tech-js-engine-internals
https://bugzilla.mozilla.org/show_bug.cgi?id=700169, which is tagged
mozilla-11, so only slightly missed the esr train. :(
But hey it's going to be fixed in the next esr version, for now you
could easily patch this manually. Just look at the patch in bug
700169.
> dev-tech-js-engine-internals mailing list
> dev-tech-js-en...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-tech-js-engine-internals
0 new messages