Many of us here are academicians. We have packs of pdf files lying around (organized in Be or other). It is very important to have a powerful searching tool to find that specific concept or idea, or term.

So far, in the Mac hemisphere, the above 4 apps seem the leaders.

Have u guys tried these apps?
Which one do you find the most effective, and why?

I own DevonSphere and it does a serviceable job but I rarely need to exclusively search PDFs and I don't find it that pleasurable to use. (One of the reasons I don't use DevonThink anymore.)

I have however completely replaced Spotlight with HoudahSpot for everyday searches of all kinds, including pdfs, and even changed the default Command-Spacebar to invoke HoudahSpot instead of Spotlight.

You can even craft a Keyboard Maestro macro to take a HoudahSpot PDF result and open the document in a PDF viewer like Skim and take you to the result, which you can see here.

https://www.youtube.com/watch?v=Rnc2TkP812M

I've used HoudahSpot since 2008, and according to my records the app's total cost including two upgrades was about $45. Worth every penny.

Never used FoxTrot but it has a very good reputation, although I think you'd need to spend $120 on the FoxTrot Professional version to match HoudahSpot's search capabilities. HoudahSpot leverages the built in Spotlight index, not sure if FoxTrot does the same or builds its own.

Back in the late 80s, the creator of Lotus 1-2-3, Mitch Kapor, started a company called On Technology, and its first Mac product was a Desk Accessory (a mini app that ran in the old cooperative-multitasking Mac OS of the time) called On Location that was a personal Mac search engine. It had its niche because search was terrible on old Macs, but the app had to build it own indices for each volume searched - floppy, HD, CD-ROM. The background indexing took a long time, and consumed up to 5% of the drive's space. If I remember right you could see a stamp-sized text window in results and you needed to open the documents themselves then search within them to find your search term. It also would only index a beginning portion of long documents, which made it useless for me. Price in 1990 was around $130, which would be around twice that today.

I own DevonSphere and it does a serviceable job but I rarely need to exclusively search PDFs and I don't find it that pleasurable to use. (One of the reasons I don't use DevonThink anymore.)

I have however completely replaced Spotlight with HoudahSpot for everyday searches of all kinds, including pdfs, and even changed the default Command-Spacebar to invoke HoudahSpot instead of Spotlight.

You can even craft a Keyboard Maestro macro to take a HoudahSpot PDF result and open the document in a PDF viewer like Skim and take you to the result, which you can see here.

https://www.youtube.com/watch?v=Rnc2TkP812M

I've used HoudahSpot since 2008, and according to my records the app's total cost including two upgrades was about $45. Worth every penny.

Never used FoxTrot but it has a very good reputation, although I think you'd need to spend $120 on the FoxTrot Professional version to match HoudahSpot's search capabilities. HoudahSpot leverages the built in Spotlight index, not sure if FoxTrot does the same or builds its own.

Back in the late 80s, the creator of Lotus 1-2-3, Mitch Kapor, started a company called On Technology, and its first Mac product was a Desk Accessory (a mini app that ran in the old cooperative-multitasking Mac OS of the time) called On Location that was a personal Mac search engine. It had its niche because search was terrible on old Macs, but the app had to build it own indices for each volume searched - floppy, HD, CD-ROM. The background indexing took a long time, and consumed up to 5% of the drive's space. If I remember right you could see a stamp-sized text window in results and you needed to open the documents themselves then search within them to find your search term. It also would only index a beginning portion of long documents, which made it useless for me. Price in 1990 was around $130, which would be around twice that today.

Disclaimer: IANAA as such, but I am a professional researcher and regularly edit academic papers on behalf of – usually German – doctoral students through to senior professors.

I’ve used DEVONsphere, but not for some time. Didn’t make much of an impression, to be honest.

I recently tried out the latest version of HoudahSpot, which I thought was really quite good. In fact, I might have been persuaded to buy it if not for…

FoxTrot Pro (please note that I use the Pro version, which is extraordinarily powerful), which I’ve been using since version 5.0, and is amazing. This is the closest thing on macOS to the academic’s friend, dtSearch (the best desktop search app on Windows, and indeed on a web server). FoxTrot Pro is now at version 7.0, but I confess I’m still using the last update to version 6.0, which is already powerful enough for my needs. FoxTrot Pro does all the things dtSearch does, at least as well, and includes its own viewers so you can open multiple found files in separate windows using the FoxTrot viewer rather than the originating app (if you want to).

I’ve started moving all my extensive databases of Stuff from document management programs like DEVONthink Pro Office, Keep It, Together, MacJournal, Yojimbo and so on into simple folders and then indexing the latter with FoxTrot Pro. You can build as many indices as you like; you can also apply complex Boolean criteria if you want to, or use FoxTrot’s ingenious “near” switches instead to make things easier. If you’re searching on multiple terms, FoxTrot will assign a different colour to each of them so you can skip from hit term to hit term and see at a glance which specific term has been highlighted. I believe you can download a trial – I’d strongly recommend doing so.

For academic (and indeed general or specialised research) purposes, FoxTrot Pro wins hands-down.

As to satis´ point, I also have tried HoudahSpot some time ago. Honestly, I find this app almost useless because it relies on metadata.

What I really want is a deep dig into the documents; not their metadata. I want a searching tool that digs all the pdf documents in my drive, and pick up that specific term (idea). In this case, to rely on applications like HoudahSpot, you need extensively tag the documents. I have too much documents to tag each manually.


I love the proximity search in Foxtrot. It is incredible.
I also have tried to use the proximity searches in Devonthink, but, requires a lot of work.

The beautify of the proximity search in Foxtrot is just the simple button to drag to expand and contract the search areas. It is so effective.

But, I was also curios about the new app: PDF-search that seems to do similar proximity search. I havn

Yes, I haven’t really used my copy of PDF Search very much, because I use FoxTrot almost exclusively, but it’s a very powerful search engine.

Reminds me I must dig it out and give it another go!

PDF-search seems to create much bigger database.

My collection of pdf files about 11.8 GB in finder.

That turn out to be just 797mb in Foxtrot`s index.
Just putting the same folder, PDF-search already piled up 9GB database. it seems like PDF-search copies the whole thing; while Foxtro does some magic.

@dellu – yes, I remember finding the same thing. FoxTrot extracts all text to generate its indices (and that’s the basis of the FoxTrot mobile app for iOS; it uses the text-based indices, not the actual files, although you can transfer the actual files if you want to and have enough storage space on your iOS/iPadOS device).

Whereas PDF Search, well, clearly doesn’t!

I've used all 3. Foxtrot (and again, this is the Pro version), just works and is powerful. It's also crazy fast. The only caveat for me is that it doesn't seem to do so well with emails (I have 85K emails). For this reason I have Devonthink Pro. However, Devonthink is slow in opening and if I were using it constantly, there are too many beachballs. A tremendous positive with Devonthink is if you also want you data on iOS. You can do this with Foxtrot, but it's not so elegant. DevonSPHERE left me all meh.

After experimenting with pDF-search for a while, I find it pretty fascinating. The results in PDF-search are much more surprising, than the ones I find in Foxtrot or Devonthink. The reason might be because it uses compilations of algorithms, to rank the pages.

PDF Search algorithm calculates a rank value for every page in documents according to keywords you entered. These ranks are computed as follows:.
Keyword Distance : Pages containing keywords nearer to each other have a higher rank.
Keyword Density : Pages containing more keywords have a higher rank.
Importance : Pages that contain the keywords in the title or that are rendered in bold or a larger font will have a higher rank.
Document Date: Pages within more recent files are ranked more highly than older files.

I’ve only used PDF Search on an iPad Pro. I find the results surprising as well but in a bad way. Often it misses the documents that I’m sure to contain the keywords. This makes a an untrustworthy search engine. For now I’ve reverted to rely on DEVONthink to index PDFs.

Hm, that’s interesting. I know I’ve had suspicions about PDF Search’s accuracy on macOS before (I haven’t used the iOS version – actually, I didn’t realised there was one), but they released an update recently which claimed to have resolved “certain issues”.

Still, no worries: I’m perfectly happy with FoxTrot!

I have a friend who wants to search and relate PDF files on a Windows machine.

Devonthink's "see also" function is pretty cool, but he needs that kind of functionality in Windows.

Windows-based suggestions appreciated.

Well, the ultimate tool on Windows is dtSearch (even now), but it’s relatively expensive, so maybe he could look at e.g. X1 or Copernic (I used Copernic when I was still on Windows). Having said that, an oldie but goldie favourite is the professional version of “Agent Ransack” (itself a very good freebie search tool): FileLocator Pro. It’s been going for a long time and is very precise (also supports a whole bunch of search functions like regex etc.)

Cheap ’n’ cheerful options include, as already mentioned, Agent Ransack and the excellent DocFetcher (also an oldie but goldie).

Interestingly, a couple of recent search-tool reviews I read rate FileLocator Pro above Copernic, especially if you want to build big indices. But IANAWU, so can’t advise you further.