[MKSearch-dev] test pages for crawler

Fri Nov 26 09:30:28 GMT 2004

On 24 Nov 2004, at 11:02, Bruno Postle wrote:

> Phil needs a small set of test pages that can be used to highlight
> particular indexing operations.  This is something Chris can do.
> 
> I'm assuming that initially the indexer is going to require valid XML
> and that it only indexes metadata within the document <head> contained
> in <meta> tags with a short set of names like so:

Yes, I would like a branch of the static test site to be purely valid 
XHTML with (ultimately) a broad range of test cases, starting with 
absolutely basic single metadatum documents for each of the DC 
elements, later e-GIF, etc. Ideally, the metadata values should be 
unique so we can un-ambiguously identify the source document from a 
given search result.

Then, perhaps in a set of subdirectories, test cases that mix up the 
metadata, repeat elements, etc. The whole thing would need to follow 
a scheme that could ultimately cover a very broad range of 
permutations.

I dare say the task would be more intellectually stimulating if the 
file set could be generated by a script!

Best regards,

Phil