I have not looked at the IO interfaces for this library, but it seems clear that there should be a good way of passing, or using, a ruby IO stream as a source to the SAX parser.
Which forks me to another topic, we need to eval and clean up the task list on rubyforge and seed it with new or reformulated feature/bug tasks we we can manage to get it done. It is also the right place to be attaching patches and requirements. I am starting to loose track of it. Thanks to everyone for your efforts and thoughts. Dan On Dec 9, 2007, at 04:30, Jochen Topf wrote: > > On Fri, Dec 07, 2007 at 07:24:49PM -0500, Todd Fisher wrote: >> On Dec 7, 2007 4:30 AM, Jochen Topf <[EMAIL PROTECTED]> wrote: >>> I have some rather large XML files which I want to read with the SAX >>> parser. Unfortunately it can only read from files or strings, not >>> from >>> a Ruby IO stream or so. I want to pipe in the XML file directly from >>> a process decompressing the file or use some Ruby lib to do the >>> decompression on the fly. >> >> It doesn't look like it's supported. It could be we'd have to use >> xmlParseChunk. It shouldn't be too hard the tricky part would be >> how to >> expose that feature in a easy to use ruby API... > > Hm. I think we need something the other way round: Whenever the > Parser needs > more characters to parse, it should call back into Ruby code which > will read > some bytes from an IO object and give those bytes to the parser. > > Jochen > -- > Jochen Topf [EMAIL PROTECTED] http://www.remote.org/jochen/ > +49-721-388298 > > _______________________________________________ > libxml-devel mailing list > libxml-devel@rubyforge.org > http://rubyforge.org/mailman/listinfo/libxml-devel _______________________________________________ libxml-devel mailing list libxml-devel@rubyforge.org http://rubyforge.org/mailman/listinfo/libxml-devel
