[XOM-interest] Canonicalizer API
Elliotte Harold
elharo at metalab.unc.edu
Wed Feb 16 07:13:41 EST 2005
I've decided to make at least one more change to the Canonicalizer API.
I'm going to remove the write(Nodes, String inclusiveNamespacePrefix)
method. Instead, I'm going to add a setInclusiveNamespacePrefixes
method. The reasons are two-fold:
1. The inclusiveNamespacePrefixes only matter when doing exclusive XML
canonicalization.
2. I'm afraid that future canonicalization algorithms may add
additional, per-document options, like a list of attributes to inherit,
and I don't want the number of methods to grow exponentially to support
them all.
One question that arises though. now that it's a separate option, should
I just use addInclusiveNamespacePrefix and
removeInclusiveNamespacePrefix, each of which would operate on a single
prefix, rather than setting them all at once as a string?
In practice, I don't expect more than two or three inclusive prefixes
per document at most; and most applications will probably reuse the same
prefixes repeatedly.
--
Elliotte Rusty Harold elharo at metalab.unc.edu
XML in a Nutshell 3rd Edition Just Published!
http://www.cafeconleche.org/books/xian3/
http://www.amazon.com/exec/obidos/ISBN=0596007647/cafeaulaitA/ref=nosim
More information about the XOM-interest
mailing list