Lists Home |
Date Index |
the n X 100Mb dump is a bit daunting. ? is it possible to get the rdf presentation of individual pages? in the html which i looked at, the content of each
respective element appeared to be amenable to a single system-specific encoding, but i'c curious whether that's strictly a matter of chance or of convention.
underlying question is whether an application would expect to work with such "mixed encoding" content as multiple "single encoding" runs. i oberve that the DOM
specifies a normalization to a single run of utf-16 with surrogates in it's "interface", while java representes strings with a variation of utf-8. although i've
never tried to do things with complex fonts in java, that would seem to still be some distance to the representation which one would need for system calls to
get the correct glyphs on the screen.
Tim Bray wrote:
> james anderson wrote:
> > is there a collection of documents available which exercises unicode? something more than the few cases in the conformance suite? something which mixes texts
> > which would have distinct (eg language-specific) 16-bit encodings?
> The RDF dumps of http://dmoz.org, tons of stuff in every which language --
> Cheers, Tim Bray
> (ongoing fragmented essay: http://www.tbray.org/ongoing/)