OCR rename multiple image files

[jkingno1]

Hi, 

I'm looking for suggestions on how to make this flow variable and OCR scan more than 1 specific file. I've tried changing the 'files' in the text detection but it brings up an error. Also the text detection response variable changes between 1-Xnumber for where the Job number is detected in each image. The number is always only 5 Numbers. Would colour coding be a good strategy to always extract the right information I need? Any ideas? I've attached an image of the flow of got working currently. 

 

[momlo]

Hi @jkingno1 

Get files action gets you list of files, now you need to iterate it with "For each Loop".

Then, each iteration willl execute code in the loop will be executed against 1 file from the list stored in a %CurrentItem% variable.

In your example, text detection and rename actions need to be moved inside the loop

Text detection needs to reference the name of the %CurrentItem%

In rename action you get the filenaem from JSON, so if this value is constant you will have namign conflict so you need to make it dynamic (filename or path).

You can learn loops here, also there are quite a few threads on this forum talking about this particular topic - renaming files in loop so take a look at those too.

https://learn.microsoft.com/en-us/power-automate/desktop-flows/use-loops#for-each-loop

[jkingno1]

Hi @momlo, 

Thanks for your response. So far I have implemented a loop that works to an extent, the only problem I get is that the text detection sometimes detects multiple 5 numbers and chooses the highest number which is not correct. If I add a "JN:" next to the Job number e.g. "JN:34738" how would I get a regex to detect for that?

I also need a way to append the data with a (1),(2),(3) etc and I'm not sure how to do that. I tried append text but I believe it isn't the correct option. Any ideas? I've attached an image of my flow. Thanks heaps!

 

[momlo]

Hi @jkingno1 

You can use lookaheads to match digits only where there is specific content before or after digits:

Match 1 or more digits after "JN:"

(?<=JN:)\d+

Match 1 or more digits between "JN:" and "text after number"

(?<=JN:)\d+(?=text after number)

[jkingno1 · Accepted solution]

Thanks for your help @momlo 

I've implemented a JN\d{5} regular expression which works really well, I realized the : cant be used when renaming the file, and I also worked out myself how to create a parenthesis (If that's what its called when a (1),(2) (3) is added at the end of the 'same name', which was done by generating a random number at the end. Here's the flow I got for anyone in future who has the same problem. 

 

[MichaelAnnis]

I don't know Jason, so I'm just going to reiterate what I think you are trying to do, and give you an answer for that assumption.

Do you just need the last 5?  If so, you don't need a regex.

For each %CurrentItem% in files

    Get subtext from %CurrentItem.NameWithoutExtension% starting from %CurrentItem.NameWithoutExtension.Length - 5% to end of text 

Then for Rename Files, looking at your loop, the If statement (steps 8-11) would rename the %CurrentItem% and then following that step 12 would try to rename the item again, but if %CurrentItem% has already been renamed, then you aren't going to find %CurrentItem% where it used to be, so step 12 will always error out if the if statement is true.

If you are trying to use one rename or the other, you want to use IF -> Else.  It goes like this:

If (calculation is true)

    'do something

Else (if not true)

    'do the other thing

End (if)

Also, %Match% is just 5 digits, correct?  If so, you can't use Match for the If file exists parameter.  It needs to be a whole file path:  something like %FromPath%\%Match%.jpg

Hope this all helps.

[jkingno1]

Thanks for your reply, i appreciate the help. Luckily i just figured it out which you will see as the marked solution. 🙂