Re: Question on parsing of "font" property

Ian Hickson wrote:
>>Note that the whitespace issue is a non-issue here, as you pointed out; I should
>>have researched a bit more and not included it at all.  The processing of the
>>token stream is the only question.
> 
> If the token stream is the only question, then why is there a question?

The token stream is not a question.  The _processing_ of the token 
stream is.  That is, there is no tokenizer issue; the issue is what the 
parser should do with the tokens.

> Or are you asking "if there is a '/' token after the font-size has been
> parsed, how do you know that it is delimiting a line-height and is not
> simply part of the font-family"?

Precisely.  Was my first email (with parsing examples) really that 
unclear?  :(

> Assume <generic-family> cannot start with a slash.

You mean <family-name>?  <generic-family> comes from a fixed list of 
keywords, none of which start with a slash.

> The working group will probably discuss this issue in more detail
> next week, but that is the solution I will be suggesting.

Sounds perfect to me.  That's exactly the sort of action I was hoping 
for -- restricting the syntax to disambiguate the parsing.

-Boris

Received on Wednesday, 19 May 2004 13:29:55 UTC