Yitzchak Gale wrote: > Another question about HaXML and HXT - > what is the level of XML spec. compliance?
Implementing the XML 1.0 Standard was one of the goals of HXT when starting the project. This includes full support of DTD processing, which turned out to be the hardest part of the whole parsing and validating stuff. Another goal was, to stay as near as posible to the XML spec, that meant, separate the tasks in a clean way and do it step by step: reading, decoding, parsing, substituting entiies, implementing the include mechanism wiht external references, validating and normalizing. In a second step we added a HTML parser, again with parsec, to be able to process none standard XML. Again we hadn't in mind to process "very large" XML documents. There is no technical reason of adding 3. parser (or a 4. one) accepting something like XML, perhaps without the DTD suff, which works lazily. The only reason not yet having done this was lack of time and manpower. So, dear Haskeller, feel free to participate to HXT by developping a lazy parser and we will integarte it into HXT. This still does not solve the processing of "very very large" XML document. I doubt, whether we can do this with a DOM like approach, as in HXT or HaXml. Lazy input does not solve all problems. A SAX like parser could be a more useful choice for very large documents. Uwe _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe