I've been fiddling with Web.RSS in Pike 7.6, and I'm wondering what people are using it for. From what I can tell, it's RSS 1.0 only, and I've not been able to get it to successfully parse any of the RSS 1.0 files I've fed it. I'm not sure if I'm doing anything wrong or not. Anyone out there have any first hand knowledge of using it?
On a side note, I wrote an RSS 2.0 parser; it's not complete yet, but I've had good luck with it so far. I'd be happy to pass it around if anyone's interested in helping out with it.
Bill
I wrote it and I've only been using it for very specific "academic" purposes. It is however a good proof that essentially no one writes correct RSS, since when I stumble upon problems it is almost always in the RSS data and not the implementation.
I figured that it was an academic venture; I don't doubt that there's a large amount of malformed data out there; RSS 1.0 is a good bit more complicated than 2.0, but there are a lot of sources out there, and they all throw different errors :/
Does it make sense to have a certain amount of lenience when parsing? I know I'd find it more useful if it could parse data already available.
Bill
On Fri, 25 Feb 2005, Martin Nilsson (DivX Networks) @ Pike (-) developers forum wrote:
I wrote it and I've only been using it for very specific "academic" purposes. It is however a good proof that essentially no one writes correct RSS, since when I stumble upon problems it is almost always in the RSS data and not the implementation.
My RSS parser is geared towards being an application of RDF. My purpose was to implement RSS, not to actually use it for anything. If you are going to use RSS for practical purposes it is probably best to not regard it as RDF, and possibly not even XML (Not that it is valid XML to begin with, only well formed).
[Moving to Pike general list...]
Fair enough. I've extended my RSS parser to understand both 1.0 and 2.0 files. Right now it's just the core format; I don't handle any namespace extensions, but is able to read a number of files that appear to follow the specs.
If anyone is interested, you can get it here:
http://hww3.riverweb.com/dist/pike_modules/RSS.pmod
If there's sufficient interest, I'll pretty it up and package it for modules.gotpike.org.
Bill
On Fri, 25 Feb 2005, Martin Nilsson (DivX Networks) @ Pike (-) developers forum wrote:
My RSS parser is geared towards being an application of RDF. My purpose was to implement RSS, not to actually use it for anything. If you are going to use RSS for practical purposes it is probably best to not regard it as RDF, and possibly not even XML (Not that it is valid XML to begin with, only well formed).
You should use Parser.XML.Tree instead of Parser.XML.NSTree since Tree also has namespaces now. I intend to rewrite everything that depends on NSTree to use Tree in the future.
pike-devel@lists.lysator.liu.se
-
Bill Welliver -
Martin Nilsson (DivX Networks) ＠ Pike (-) developers forum