Re: [Texmacs-dev] Discussion about Search Engine Plugins

[Joris van der Hoeven]

Hi Ingolf,

On Sun, Aug 27, 2006 at 02:48:02PM +0200, Ingolf Schäfer wrote:
> I have started to do a SpotLight plugin, i.e. a Search Engine Plugin  
> for Mac OS X. There are some things which I would like to discuss  
> here, if possible. They should be the same for all other possible  
> architectures (Linux, BSD , Windows, ... ).
> 
> A search engine normally stores things like the name, date of  
> creation, and so on of a certain file. If you write a plugin, you can  
> import meta data. For instance the document title, the author's name  
> some keywords and other related things.
> 
> 1.) I would like to here some feed back on which meta tags, we should  
> make available to the search engines. At the moment, I have just made  
> a very crude version, which just grabs the document title and the  
> author.
> 
> .2) But even there is a problem, because I would like at flat text  
> representation, not a branch of the document's tree structure. Is  
> there any TeXmacs code, which I may reuse for this?
> 
> 3.) Another brute force option, would be to use the xml-format as  
> default for TeXmacs. There excellent parsers for xml-code out there,  
> that we could use in this case. This would make it _very_ easy to do  
> such search engine plugins.
> 
> What is your opinion on this?

I would prefer an even more brute force option: just scan TeXmacs for all
plain sentences or words and give them to spotlight. I think that it is
very useful to be able to grep the whole collection of all documents one
is writing. No real need for XML in order to do this.

Best wishes, Joris