I quick query.

A colleague and I had a bit of a 'discussion' (argument) the other day about search facilities and their ability to index web content.

My argument was that to index and search through a traditionally none-web-based document such, as a Word or Excel document, you needed a specialist search tool, such as e-swish or some of the other alternatives to that. I stated that for that reason, Google only return values from either html or pdf formats because it isn't designed to search through standard files.

His argument was that any search facility can index and search through word or Excel documents and the only reason that Google and the like does not return values from those file types is because webmasters do not traditionally publish such files for access over the Internet. In other words, it's simply coincidence that that I've never seen a search hit from a file type other than html or pdf.

Who's correct?

The reason I ask is that we have a large repository of Word documents that ideally we'd like the XOOPS search bar to be able to index and return search hits from. I said it can't do it and that XOOPS can only return values from data stored in the database and you can't make XOOPS go out and index an external Word file. He said it should be able to do it.

Who's correct?

Thanks

When a URL is requested, by a human or a search bot, the response isn't really a file such as .html or .pdf, but a document with a MIME content type and body. As long as the server (or the script running on the server) returns a content type and body that the search engine is happy with, I don't think the file extension matters.

Thanks Dave_L.

So could the XOOPS search be used to search through a Word document etc stored on our network using a direct UNC path?

No. The XOOPS search by itself only works by searching via SQL queries. However, I think that the DMS module uses the Swish-E file indexer, which can index the contents of Word, Excel, PDF, and others.