[
Lists Home |
Date Index |
Thread Index
]
At 10:19 AM -0800 1/24/04, Tim Bray wrote:
>It dawned on me that with all the checking and so on, genx might not
>be as fast as it could possibly be. The way to fix this is obvious,
>but it might amount to premature optimization. The idea is that you
>predeclare your elements and attributes and get handles to them, so
>that they only need to be namechecked and sorted once.
Definitely premature. It's possible to optimize the hell out of
verification to the point where the I/O cost is much larger. Xerces
does a very nice job of this, mostly by using table lookup, and XOM
has further improved on those techniques by short circuiting some
common ASCII cases and generating the tables at compile time rather
than runtime. I'm not convinced this is as fast as verification can
get yet either. Don't worry abut the cost of verification until
profiling proves it's an issue, you've optimized the verification
code, and it still prove to be the bottleneck.
--
Elliotte Rusty Harold
elharo@metalab.unc.edu
Effective XML (Addison-Wesley, 2003)
http://www.cafeconleche.org/books/effectivexml
http://www.amazon.com/exec/obidos/ISBN%3D0321150406/ref%3Dnosim/cafeaulaitA
|
