Advanced PDF searching

I want to be able to use Boolean searching, for example, to find Baalbek NEAR/10 time not just for those documents which match the search criteria (ie that combination appears at least once) , but to find every instance within that document where those two words appear in proximity. IMHO it is only the latter ability that is truly useful to search large volumes of documents for contextual connections.

I last raised this issue about two years ago and I believe it was not possible then. My question for forum users is:

  1. has that been implemented in Devonthink Pro yet? If so how do I do it?
  2. is it likely ever to be, or are there insurmountable technical difficulties?
  3. if the answer to 1. and 2. is no, can anybody recommend me some Mac software that will play nicely with the Devonthink that will allow me to keep DT as my data repository but will usefully search my large PDF library for the research project I’m about to commence for these types of Boolean searches
  4. is there a script that has been or can be written that would allow DT to do these useful Boolean searches of databases.

FWIW, I would also welcome some answers to the above…

Searching supports such boolean operators but highlighting of found occurrences doesn’t. At least plain/rich text formats might support this in the future, PDF documents aren’t that likely anytime soon due to limitations of Mac OS X’s PDFkit framework.


So disappointed to hear this Christian. It was perhaps my biggest wish for the next version of DT. Thanks for all your hard work of you and the debs anyway Christian. I am sure there will be much in DT3 to surprise and delight.

Summarised and highlighted search results is perhaps one of my favourite features of Acrobat.

I referenced an example here: [url]What does the future hold for DEVONthink?]

David, you might try looking at Acrobat Pro. It has some advanced searching parameters including stemming and proximity. Ideally for these to work well, the pdf documents have to be catalogued by Acrobat, a process which doesn’t work very well with files stored inside a DT database. You have to rather export and index the files you want to catalogue with Acrobat.

Its a while since I have used it but I recall that Foxtrot Professional also has fairly sophisticated search parameters and an ability to display the results in a list format.


Stay tuned :slight_smile:

Thanks for the replies. My last foray into Adobe was a few years ago and I remember it being sluggish and unstable on my Mac. Maybe its improved, but I would really like to stay within a Devonthink/Tinderbox environment.

Boolean location of hits would be such a great feature for DT if its technically possible - especially for those of us with very large pdf libraries.