How to Grep all your tinderbox documents

Spotlight metadata importers are written by the developer of the product. E.g., Eastgate. They are part of the software. This is what DEVONthink does when you click the “Create spotlight index” option in a DEVONthink database’s property sheet. The spotlight metadata files are placed in a cache in ~/Library/Caches/Metadata. The importers themselves are .mdimporter files and can be found in various places in the file system including inside .app packages. (There’s a method for third parties to write importers, sometimes called Spotlight plugins.)

If you have DEVONthink and create the Spotlight index you will notice when you do a Spotlight search that there are documents listed in the results with Kind == DEVONthink Pro Office Document. Those are the metadata files – extension in this case is .dt2. On my machine other apps created metadata caches – MacJournal, Papers, VoodooPad, OmniFocus, etc. These are the files that FoxTrot uses when it builds its index. In that case, FoxTrot is doing nothing more than Spotlight and cannot give better answers to search queries than the data that the developer placed in their metadata cache files.

The problem of searching TB file with Foxtrot is that the whole database is going to be considered as one file.

It might be helpful for some kinds of searches, like the proximity search. But, since the individual notes of TB are not visible to Foxtrot, the notes cannot be ranked with Foxtrot’s ranking system. Having a large TB database, Foxtrot is going to rank that database over the other files in the system, all the time.

Exporting each of the notes individually and indexing them in Foxtrot sounds more practical to me.

1 Like

Thank you for the script. This is neat. I find some cool sentences I never knew they existed.

Excellent. Then it was worth my time to figure it out and post it. Thanks for the note.

1 Like

Just came across this. Very useful! Thanks.