Can't load tessdata_fast data packs

[trisolari]

Hey, I was having issues converting the regular .traineddata files to the fast variants, but for some reason could not initialise the data packs.

I was able to repro it in the Console example from https://github.com/charlesw/tesseract-samples by just adding the fast datapacks to the tessdata folder and adding those languages to the Engine constructor.

Just wondering if this is a case of the tessdata_fast being incompatible with the current version (v5?) ? I wasn't able to see anything indicating the sort on the tesseract doc website.

Here is a fork for your convenience: https://github.com/trisolari/tesseract-samples

using (var engine = new TesseractEngine(@"./tessdata", "eng+Latin", EngineMode.Default))

Error opening data file ./tessdata/Latin.traineddata
Please make sure the TESSDATA_PREFIX environment variable is set to your "tessdata" directory.
Failed loading language 'Latin'
Mean confidence: 0.95
Text (GetText):
This is a lot of 12 point text to test the
ocr code and see if it works on all types
of file format.

The quick brown dog jumped over the
...
...

Would appreciate any help!

[trisolari]

Hmm appears Linux only? Most users will want tessdata_fast and that is what will be shipped as part of Linux distributions.

[trisolari]

Didn't download the files properly