OCRmyPDF - Wired error message from tesseract

Question

I get a strange error message when running a OCRmyPDF command

My setup:

macOS Sequoia 15.2
OCRmyPDF 16.8.0 (installed by Brew)
tesseract 5.5.0 (installed by Brew)
Command: ocrmypdf -l deu+fra+eng --clean --force-ocr test.pdf test-out.pdf 2>> debugOCR.txt

I have to say that the command is triggered by the software NoodleSoft Hazel, and as far as i understand Hazel executes the shell commands in a dedicated environment. However, my setup worked fine for a few weeks, but within the processing of a batch of PDF files, the following error started to occur. Since then I was not able to bring it back to work.

The debug file debugOCR.txt shows the following error:

1 [tesseract] Error in fopenReadStream: failed to open locally with tail 000001_ocr.png for filename /tmp/ocrmypdf.io.81a_o2mw/000001_ocr.png
1 [tesseract] Leptonica Error in findFileFormat: image file not found: /tmp/ocrmypdf.io.81a_o2mw/000001_ocr.png
1 [tesseract] Error in fopenReadStream: failed to open locally with tail PNG for filename PNG
1 [tesseract] Leptonica Error in pixRead: image file not found: PNG
1 [tesseract] Image file PNG cannot be read!
1 [tesseract] Error during processing.
SubprocessOutputError

In the folder /tmp i can't find any subfolder like /tmp/ocrmypdf.io.81a_o2mw/.

I also have to mention that when executing the following commands directly in Apple Terminal, they work fine:

ocrmypdf -l deu+fra+eng --clean --force-ocr test.pdf test-out.pdf 2>> debugOCR.txt
tesseract test.tiff output --oem 1 -l eng pdf

Any hints where I have to dig deeper? Is ocrmypdf or tesseract missing some environment variables in the Hazel environment? Other hints?

Thanks a lot

AJ

the two commands are just to show that ocrmypd and tesseract work when used separately directly in the Apple Terminal. The error message I get when I call ocrmypdf from Hazel. In Hazel i don't call tesseract directly, tesseract is being called by ocrmypdf as a sub-process, as far as I understand. — user9803071
– user9803071, Commented Jan 30 at 7:18

Aman Rusia · Accepted Answer · 2025-02-04 19:31:59Z

0

https://github.com/tesseract-ocr/tesseract/issues/4333

This is likely the issue.

I faced the same while using wcgw mcp which also has a separate terminal evironment.

Setting TMPDIR to //tmp helped me.

answered Feb 4 at 19:31

Aman Rusia

161 bronze badge

Sign up to request clarification or add additional context in comments.

1 Comment

user9803071 Feb 6 at 6:45

I already had ... export TMPDIR="/tmp" ... in my script. but changing it to ... export TMPDIR="//tmp" ... did the trick... thanks a lot for the hint...

Collectives™ on Stack Overflow

OCRmyPDF - Wired error message from tesseract

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related