Re: Problems getting right results


On Sun, 2007-01-21 at 16:24 -0500, Carlos Moffat wrote:
> I'm using beagle 0.2.12 (Debian/Sid). I'm having trouble getting the
> right results to show up. One example is this: I have several PDF files
> authored by smith, so smith is both in the filename, say smith03.pdf,
> and frequently as text within the pdf file. Also, I have emails from
> this person.

One thing to try is to run beagle-extract-content on the file to see if
the text is being correctly extracted from the file.

> However, if search for 'smith' in either beagle-query or beagle-search,
> I get only the contact information and a few files (including 1 or 2 of
> the PDFs), but none of the emails.
> If, however, use 'smith ext:pdf' or 'mailfrom:smith', then the results
> show up correctly. Any ideas?

How many results are you seeing?  There is by default a limit of 100
results per backend, so if you have potentially hundreds of results for
"smith", it's quite possible you won't see them.

Another thing to try is to move the files you're missing into somewhere
in /tmp and set up a sandbox.  Let's say you create /tmp/sandbox and
move a few PDF files that aren't matching into there.  You could then
set up a daemon which worked only in that directory:

        BEAGLE_HOME=/tmp/sandbox BEAGLE_EXERCISE_THE_DOG=1 beagled --fg --debug --backend files
and watch as it indexes the files, when it's done, run a query in
another terminal:

        BEAGLE_HOME=/tmp/sandbox beagle-query --verbose smith
and see if things match.  We'll go from there. :)


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]