Regex support

[spasavanje]

Hi there,

Just wanted to say this app is amazing, great work! As I was testing it, it occurred to me that regex support would make even more useful. I don't if that's even possible, just wanted to share it with you.

[seanpedrick-case]

Thanks! I've been putting an update together for this week to put in some VLM support. I'll try to add in regex support at the same time.

[seanpedrick-case]

Hi,

As of version 1.5.2, regex searches are possible in two places. First, in the csv supplied as a 'deny list' for the redaction process (uploaded on the Redaction settings tab).

After initial redaction when you are reviewing outputs on the Review tab, you can now use regex using the interface on the right to search through text. You can enter regex terms in the 'Multi-word text search' box, and then enable use regex under 'search options' just below that. When you hit search, it should now identify all words in the ocr outputs that match the regex expression.

Let me know if you have any trouble getting it to work!

Sean

[spasavanje]

Hey Sean,

Thank you very much, that was really quick. I will try it out first thing tomorrow morning, I am sure it'll work just fine... The reason I thought it would be a good idea, is phone numbers in my country. They have to be redacted but aren't recognized currently, so I thought regex might sort it out.

Do you mind if I test it with some of my colleagues and let you know if we get some constructive idea or suggestion from the front lines???

[seanpedrick-case]

Good idea! Yes I have tested it with a regex form like \d\d\d-\d\d\d\d (i.e three digits, a dash, then four digits), and it worked, so hopefully you are working with something similar. That's with a deny list input via csv on the redaction settings page, or by using the 'Multi-word text search' on the Redaction Review tab, which you can find on the right of the screen. Let me know if you have any trouble with it!

And yes, absolutely fine for anyone you want to test it out. Thanks for using it.