Before I move on with this post, I'm going to make a couple of assumptions. I'm going to assume that you already understand that Hazel is an amazing application for macOS, that watches over folders in some form of virtual nanny-state, performing file actions on your behalf whilst you get that busy work done.
I'm also going to continue on the basis that you know about the awesome power of DEVONthink Pro Office, the Database/Paperless Office/Finder Replacement/All round swiss-army-knife-type-storage-app for macOS.
They are both amazing applications and I couldn't imagine having to live without either of them. However, after reading various forums and listening to some of my favourite podcasts, it seems that the general impression is that you can't have a workflow that includes them both. Using Hazel to rename files isn't possible within DEVONthink itself.
While the statement above is largely true, a small amount of lateral thinking can, however open up some possibilities.
1) All files downloaded to my Downloads folder (the default)
2) Hazel watches this folder and automatically performs OCR on any PDF files that appear, using AppleScript and PDFPen Pro
3) These files are then renamed and moved to a Holding Folder I have in Dropbox.
4) Using the built-in Folder Actions for DEVONthink, a script runs on all contents of the folder. This script will Import, OCR and Delete the original file.
Dropbox Holding Folder
The first stage is to set up a Holding Folder. Now this Holding Folder has one purpose and one purpose only. It will perform a Folder Action that will Import a file into DEVONthink, perform OCR (Optical Character Recognition, making the contents of the document searchable) and Delete the original. This folder is the final place files will reside before import into DEVONthink.
The OCR that will be completed in this stage will be using the ABBYY Fine Reader Engine, arguably the most powerful OCR software on the Mac and built-in to DEVONthink Pro Office. We're going to be performing a 'lighter' (yet still powerful) OCR at an earlier stage, so as to rename the file effectively, however the final OCR will take place here.
Why have I specified that this should be setup as a Dropbox folder? Well, firstly, any folder connected to a Cloud Service is fine (Google Drive, OneDrive, iCloud) - I just happen to use Dropbox extensively. Secondly, I find myself scanning and submitting business expense receipts from iOS a LOT. I use PDFPen Scan+ to scan the receipt, I rename with a TextExpander snippet before sending it directly to my Dropbox Holding Folder. When it reaches my Macbook, the Folder Action kicks off and the file is imported into DEVONthink.
Assigning Folder Actions
Assigning the Folder Action to your newly created Holding Folder is a simple process.
- Open Spotlight
- Search for Folder Action Setup
- Click the + button on the left hand pane. Navigate to the Holding Folder you wish to assign the action to.
- You will be presented with a list of scripts to assign. Select Devonthink - Import, OCR and Delete.scpt
Please note, you must have the scripts for DEVONthink installed. If you don't see any DEVONthink references in the box of assignable scripts, install them by clicking on DEVONthink Pro Office in the menu bar, select Install Add-Ons and tick the check box for Folder Actions Scripts.
Great stuff. You now have a folder that will act as a direct conduit between yourself and DEVONthink that is accessible from anywhere you have an Internet connection.
OK. Now, we need to set up some Hazel rules. The question we need to ask ourselves, however, is where should these rules be applied? We've set up a Holding Folder, which is going to Import, OCR and Delete thanks to the Folder Action we assigned, yet this folder needs to be the last stop for the file - the final conduit between Finder and DEVONthink. Therefore I apply the majority of my Hazel rules to the default Downloads folder. These rules will carry out some OCR (using PDFPen Pro from Smile Software and subsequently rename the file based on the results.
I'll refer to you an earlier post on Date Matching with Hazel which runs through the process of scripting the OCR, looking for a date within the contents of the document and subsequent renaming.
The only part you need to add to the rule, in order to import into DEVONthink, is to move the file into your Holding Folder once renaming/tagging etc is complete. (Remember, DEVONthink recognises Finder tags, so you can use Hazel to pre-tag your files before they hit your DEVONthink Inbox)
You can make a lot of changes to the workflow above to suit your specific needs. I know that this works for me and apart from the odd tweak here and there, I'm very happy with how it runs.
There will be a video posted soon that walks through this process in finer detail - sometimes screencasts work better than text ever can, so watch this space and I'll let you know when it's ready!