Tesseract couldn't load any languages

[cheeck97]

Hello, I recently encountered such a problem that a tesseract cannot load / initialize * .traineddata.

But by the time I start receiving this message, the tesseract already recognized about 50 iterations in 1-2 hours. *.traineddata is in two places: in the root of the launch and in the TESSDATA_PREFIX environment (/usr/local/share/tessdata).

If I reload * .traineddata, the problem will be solved only temporarily
Help me solve the problem please

OS : Ubuntu 18.04.3
NodeJS: 13.8.0
NPM: 6.13.6
Tesseract.js: 2.1.1

[Balearica]

If this issue is still active, can you post a snippet of the code your were running and/or link to the code in a repo? There seems to be similar issues stemming from multiple processes writing to the same .trainneddata file, so I'm wondering if this is related (see #576).

[Balearica]

I was able to reproduce this bug in version 2, but not the current version (v3). If this issue is still active, please confirm you still encounter this bug in the latest version and provide a reproducible example. Other users confirmed still an issue in v3.

[Balearica]

I implemented changes that should resolve this in Version 4. Specifically, the cache used to be overwritten with each recognize job, which I believe lead to it being randomly corrupted seemingly at random (often after hours of recognition jobs). If anybody encounters this error using v4, please open a new issue.