Re: GSoc 2007 Weekly report (Browser Extension Rewrite)

["Lukas Lipka" <lukaslipka gmail com>]

Hi!

All the files in ~/.beagle/ToIndex get handled by our IndexingServiceQueryable. How this queryable works is that the ToIndex directory is monitored by inotify and every time you drop a file in there, it gets handled by the queryable itself and indexed. 

Alongside with the actual file to index you drop in a metadata file (sidecar) which defines the URI, HitType, MimeType and properties of the file you want to index. The metadata filename is prefixed with a '.' (so if you want to index "
foo.html", the metadata filename is ".foo.html").

Structure of the metadata file:
* 1.line - URI
* 2.line - HitType
* 3. line - MimeType
* All following lines are properties in "t:key=value" format

Hope that helps! ;-)

Best,
Lukas

On 6/2/07, Tao Fei <filia tao gmail com> wrote:

I have read the old extension's code. I'm a little confused.
It seems that the extension write the meta data (url, title, etc) of
an web page into a file (in ~/.beagle/ToIndex ) and then overwritten
it with the content of the web page.
Then it works. The page is indexed. It doesn't call beagle-index-url.
I need more information about how beagle index the file. How does
beagle know the file  is an web page ?
(I have read the wiki, but failed to found any information about it)
Any links?