RECEIPTS RENAMING PROCESS/WORKFLOW

Hello,

I’d like to index PDFs (receipts) in DEVONthink. They come from ExactSCAN Pro, are already readable, and are saved on my external Disk as e.g.: “0000-00-00_0000_ACCOUNT_00,00_OBJECT_MERCHANT_.00.0.pdf”,

I guess I am not the only user who’d like to process this kind of files for indexing-purposes:

Is it possible to get any script for Hazel in order to extract some information out of the OCRed receipt, which I’d like to appear in the name? As an example, a filename would become: “2013-08-05_0135_SOFTWARE_28,00_NoodleSoftStore_Hazel.33.1.pdf”

It means extracting the expenditure date (28,00) out of the readable PDF in order to have the filename renamed/actualized with this information.

And is there any script- or not-script-solution who could assist in doing the rest of the renaming job?:
- Filling the Account out of a predifined list
- Filling out the Object out of a predifined list
- Filling the Merchant out of a predifined list

Any hints would be much appreciated

G

First off, please just email OR post. Doing it both places makes me have to reply in both places. From here on out, I'll reply here so I don't have to type everything twice.

If the account, object and merchant are in the name, you can use "Name matches" to pull that info out. Search the help on "match patterns" on how to specify a pattern/custom token. Any custom tokens you create can then be used in a renaming pattern.

If any of the data you want is in the contents (like expenditure date), you can try using "Contents contains match" to find and match it.

If all the data is in the name or contents and you are able to match against them, I'm not sure why you would need the predefined list in this case.

Hello Rob,

sorry, it took me a while to answer as I had to go deeper into a tutorial. I now try to answer in the same order:

Account/Object/MERCHANT: considering particularely the “Merchant”, does using “Name matches” mean I have to, FIRST, create a single new rule or sub-rule for each new merchant which is going to be OCRed AFTER this step? Or is it possible to generate/extract the Merchant’s name dynamically with a single generic rule who would suit for any new type of receipt (or invoice) involved into the hazelation (renaming proccess)?

Matching is kind of tricky in several situations. I created an overview of different cases who bring up a few questions. It also answers the reason for the predifined-list-idea:

http://www.2shared.com/photo/aUsSXLOB/2013-08-10_HAZEL_RENAMING-WORK.html

(…For better legibility, I kept the original size of the .png = 13 MB)