Re: lex, yacc, and xml

Tim Bray (tbray@textuality.com)
Mon, 22 Dec 1997 12:18:31 -0800


At 02:04 PM 22/12/97 -0600, Ward Harold wrote:
><question name="why hand code parsers" class="potentially stupid">
>Why is it that all of the XML parsers/processors I've seen appear to be
>hand coded rather than generated via lex/yacc or flex/bison?

It could be done. At least one person I know is working on it. And
in fact Norbert Mikula's NXP uses, I believe, a lex/yacc-like-thingie
for Java. On the other hand, this is probably going to get you quite
large code size; also there will likely be problems handling encodings;
also these generators tend to make it hard to generate high-quality
error messages.

On the other hand, if you want a quick one-off in C, lex/yacc are proabably
a very reasonable strategy. -Tim