vonhabsi writes

Is it possible to run recoll with an alternate database, set of directories, and options from the command line?

I am looking to run a recoll configuration with options like ' --root-dir=~/some_directory --database=db_directory --config=some_other_recoll.conf --debug=on --debug-log=otherfile"

I basically want to try out recoll on small set of files to see how it works.

PS. Could the Kind* values in the bug submission form include help?

medoc writes

Unfortunately, I don’t know how to change the list of Kinds of issues. Bug/major seems a bit overblown though. Why not use task ? Or use the mailing list.

Anyway, no problem. The normal approach for doing what you want is to create another configuration directory (e.g. ~/.recoll-other), and use the -c option:

recollindex -c ~/.recoll-other
recoll -c ~/.recoll-other

You can also use the RECOLL_CONFDIR environement variable to the same effect.

But really: you are asking questions which are almost answered in the manual TOC !

vonhabsi writes

Actually I am trying to run a small configuration where I can get kind of output referenced in this issue - https://bitbucket.org/medoc/recoll/issues/355/recoll-is-always-indexing-non-updated, some more detailed information on what recoll is doing on a small run.

vonhabsi writes

I am trying the changes described in the wiki - https://bitbucket.org/medoc/recoll/wiki/WhyIsMyFileNotIndexed and notice need for logfilename = stderr to capture python logs.

Does that mean there is no way for python errors to be captured to a file and saved later?

logfilename = stderr
thrQSizes = -1 -1 -1

medoc writes

You need to redirect both stdout and stderr to a file:

recollindex  > /tmp/trace 2 >&1