Editing huge files; recommended Java parsers?

[David Byrden <David@Byrden.com>]

For editing files of around 200MB without unusual
power or memory consumption, what is recommended
for the Java platform?
The edits are probably local, so the obvious approach
is to build a memory tree for specific subtrees of the
document. What parser would help? We'd like to do as
little custom programming as possible.
Another need is to compare pairs of files. At present we
use SAX, canonicalising them in an app-specific way, and
flattening (stripping out the markup) so we can get them
both into memory.

					David