Character classification

[Istvan Cseri]

For better speed I would suggest an alternative solution: use a quick
array lookup for characters below 256 and go to the more expensive
method above... It will do wonders with your parser.

Istvan

> ----------
> From: 	Tim Bray[SMTP:tbray at textuality.com]
> Reply To: 	Tim Bray
> Sent: 	Wednesday, September 03, 1997 12:51 PM
> To: 	xml-dev at ic.ac.uk
> Subject: 	Character classification
> 
> <<File: CharClasses.java.txt>>
> I've been working on making Lark really do Unicode.  JDK 1.1 is
> supposed
> to have, unlike 1.0, a usable input method; thus the problem is to
> check,
> when you're reading a GI or Attribute name, whether the characters are
> legal namestart/name characters.
> 
> It turns out to be quite a lot of work, so this is an offer to share.
> I wrote a program (based on Lark) that pulls the relevant character
> classes out of the XML spec, picks apart the markup, and writes
> another
> Java class that has some static arrays and offers two methods:
> 
> package textuality.lark;
> public class CharClasses
> {
>  public static boolean isNameC(char c)
>  public static boolean isNameStart(char c)
> }
> 
> It needs about 4k of tables (which it binary-searches); it might be
> faster
> with 128k of byte-addressable tables or 16K of bitmaps, neither of
> which
> would be hard to implement.
> 
> (a) is this a waste of time, i.e. are there Unicode library calls that
>     do it?
> (b) if not, has everyone else already done this?
> (c) if not, if I'm going to publish this, is the API above OK?
> 
> I've attached the current Java source file for those who find the 
> explanation above insufficiently clear.
> 
> Cheers, Tim Bray
> tbray at textuality.com http://www.textuality.com/ +1-604-708-9592
> 

xml-dev: A list for W3C XML Developers
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To unsubscribe, send to majordomo at ic.ac.uk the following message;
unsubscribe xml-dev
List coordinator, Henry Rzepa (rzepa at ic.ac.uk)