slate unicode support

Brian T. Rice water at tunes.org
Fri Jul 30 11:15:36 PDT 2004


� wrote:
> On Thu, 29 Jul 2004, Brian T. Rice wrote:
>> Re: Lifting String into the image: agreed. The lexer and parser only 
>> depend on being able to grab characters from a ReadStream, so there 
>> should not be a problem. The lexer's behavior should be as it is, 
>> since Slate perspectives on character use is particular to the syntax, 
>> and not general.

> What do you mean by the last sentence? That one shouldn't use any other
> letters in slot/method/whatever names than [A-z0-9]? How would using for
> example chinese or tamil letters break the syntax?

Aside from Attila's interesting remarks, mostly that I mean that the 
lexer decides this and not Strings /on their own/. If you're really 
interested in how to do it for multiple languages, you'd want to 
parametrize the lexer on something else, like making a Latin1 Lexer vs. 
a Hebrew Lexer.

I really didn't mean much specific on it, and when the syntax is truly 
abstract and object-based, I won't really give a damn, since we 
presumably would have a much different and looser system anyway, 
directly acting on a parse tree. Note, however, that structured editors 
are too restrictive in editing per se; there will still need to be 
looseness if/when we do this.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: water.vcf
Type: text/x-vcard
Size: 218 bytes
Desc: not available
Url : /archives/slate/attachments/20040730/e76ee2e4/water.vcf


More information about the Slate mailing list