Re: Issue with Mozilla/Firefox & named anchor with Japanese/Korean character

Mark Tyndall wrote:
> kgs wrote:
>> The following link has Korean char & doesn't work.
>>
>> http://docs.hp.com/ko/allproducts.html#bh7800%20%BC%AD%B9%F6
> 
> It also has a space character (%20).

It does not have a space character.

The content of the name attribute is not an url, so it can not be 
urlencoded, so this should be interpreted as really %20, and not as the 
url encoding of the space character.

The same for the rest of the content of that name attribute.

>  I don't think that spaces are 
> valid in id=""s or name=""s, even if encoded.

It /seems/ that they are valid in 'name', even if not in 'id'.
Despite the fact 'name' and 'id' share the same namespace, they are not 
of the same type.

The type of the 'id' attribute is 'ID' and this will not allow a space, 
but the 'name' attribute is of type 'CDATA', so can have a space.

Back to the point, a little editing of the document shows that if the 
content of name is directly the value intended, without the url 
encoding, the link to it works.

The editor in the suite doesn't like at all the urlencoded form (when 
you go to source editing, at first it destroys the content, and the 
second time, it supresses the tag as the content is empty).

The really surprising part is that this page is generated by an XSL 
processor. Usually, you simply can not get generated invalid html with them.

RESOLVED/INVALID.