After negligible response to my last questions about this, I'm trying again... I've been modify parts of MnogoSearch and writing lots of my own code recently. Since I like XML as a transport medium, I have a way of pushing information into the search engine as XML, and for returning results as XML. First, Input: This works quite well. All the important bits of the pushing-info-into-the-engine can be found at http://www-student.cs.york.ac.uk/~gjb105/udm_xml.tar.gz Unfortunately, it currently only works on the crc-multi db schema. If anyone ever responded, and asked for support for other types, I'd look at it, but until then, it does what I want. Any advice/comment from people on this would be greatly appreciated. Second, Output: I can upload a tarball of the relevant files if required [PHP, but if you only want the XML file without the interesting HTTP header modification, then you just need the serach.html equivalent] This also works quite well. Internet Explorer loves it. But perl under solaris doesn't. OK. It does like it, but not all of the time. Basically, I index a lot of word documents, and HTML documents created by word. In that helpful way that word does, it uses various ugly character sets, mostly {windows-,cp}{1250,1251, and derivatives} Whenever I try and munge any output that has some of these non-ISO-friendly charaters in it [nost noticably, apostrophes, but copyright symbols appear a fair bit, too], it breaks. It throws a perl error of the form "scary character I don't recognise found at line XX column YY" I've not got no experience of perl character set munging, so I'm kinda having problems working stuff out; I really need some help on this, as it's currently the single biggest problem I'm having. I don't like database translations, since you lose original data; so this would probably have to be a post-processing thing, or even done in the PHP if possible. Again, I've no real experience of this... Thank-you very much, Gary (-; ___________________________________________ If you want to unsubscribe send "unsubscribe general" to [EMAIL PROTECTED]
