Hazel 6 OCR on-the-fly contents

Thanks for 6.0.1. Update complete.

The account number in my bank statement still isn't getting picked up. To refresh your memory, that account number appears when viewing the .pdf, but is removed from the embedded text in the .pdf. The rest of the contents are in the embedded text. I was hoping Hazel would scan the document separately and match the account number. Is that the expected behavior?

If the document has text, then text recognition does not come into play. A user above request an way to use text recognition in these cases which I'm looking into.

Mr_Noodle wrote:If the document has text, then text recognition does not come into play. A user above request an way to use text recognition in these cases which I'm looking into.

Understood. Thanks.

@sascha I don't understand how your rule applies to OCR at all. It does nothing to match against contents.

Not everyone needs full OCR and aside from PDF, other image format have nowhere to store the text. Sometimes it's enough to do a quick organization by contents without needing anything more. Or to do an initial check before using a more dedicated OCR program.

Hi there,

As another data point I've also been unable to make use of text content matching.
Text Contents are always empty in my case.
For Use text recognition I've tried both always and as needed.

For reference I used the Noodlesoft purchase invoice, trying to match the word "Receipt"
Hazel matches on the Contents rule, but not on the Text Contents.
Image at: https://cln.sh/R8NQh8N48wdmBGGSY6Hq

This was with Hazel 6.0.2 on macOS 15.1.

Thanks, and let me know if more details are needed.

Why are you using Text Contents at all? Use Contents.

Mr_Noodle wrote:Why are you using Text Contents at all? Use Contents.

Thanks.
For reference I had a PDF which didn't have embedded text, just outlined fonts.
Thought the OCR results would show up in the Text Contents attribute, but I now see that it actually gets set on the Contents.