The collection of supported languages is empty

[HerGeo]

Description
After a clean install, Audiveris asks every restart for downloading the tessdata files. Actually, they are in the config/tessdata and get overwritten at every download. When processing a pdf, I get The collection of supported languages is empty for every page. I have no Tesseract installed by myself (at least I think so).
I checked TESSDATA_PREFIX, but it's not set / null.
Probably due to this bug, no text is recognized.

To Reproduce
Clean install of Audiveris, download languages by checking checkboxes and restart Audiveris.
Alternatively, transcribe a pdf and see the message in blue.

Expected behavior
Not to be asked for language installation once they are installed

Audiveris environment:
from log:
TESSDATA_PREFIX: null
Environment:

• Audiveris: 5.6.2:85cdc9e
• OS: Windows 11 10.0
• Architecture: amd64
• Java VM: OpenJDK 64-Bit Server VM (build 21.0.8+9-LTS, mixed mode)
• OCR Engine: Tesseract OCR, version 5.3.1
  Alias patterns: [(IMSLP[0-9])-.]
  Audiveris version 5.6.2

from Help>About
Audiveris
Role: Optical Music Recognition
Version: 5.6.2:85cdc9e
Container: file:/C:/Program%20Files/Audiveris/app/audiveris.jar
Home page: http://www.audiveris.org
Dev. project: https://github.com/Audiveris/audiveris
Software license: GNU Affero GPL v3
OCR engine: Tesseract OCR, version 5.3.1
OCR folder: C:\Users\XXX\AppData\Roaming\AudiverisLtd\audiveris\config\tessdata
Java vendor: Azul Systems, Inc.
Java version: 21.0.8
Java runtime: OpenJDK Runtime Environment (build 21.0.8+9-LTS)
Java VM: OpenJDK 64-Bit Server VM (build 21.0.8+9-LTS, mixed mode)
Operating System: Windows 11 10.0
Architecture: amd64

[hbitteur]

@HerGeo

Could you post the content of your OCR folder at

OCR folder: C:\Users\XXX\AppData\Roaming\AudiverisLtd\audiveris\config\tessdata

I suspect it is still empty.

To populate this folder, you must do it interactively, via the menu Tools | Languages. See this handbook section.

[HerGeo]

Exactly, thats the thing: I download the languages by checking the checkboxes in the "Installation of OCR languages" dialog. The log says, that the respective file is stored in this exact folder. When I open it, I can see the downloaded files so this isn't the problem, see screenshot:

But if I transcribe a pdf, I get the warning "The collection of supported languages is empty" and if I restart, Audiveris asks me to download the languages again, although the files are still in the folder as shown in the screenshot. If I then check the checkboxes again, the file is downloaded again and replaces the old one.

[hbitteur]

Just in case, could you verify the value of the constant useOCR.
To do so, use the menu Tools | Constants and type useOCR in the Search field.
Verify the checkbox for this boolean value:

[hbitteur]

But if I transcribe a pdf, I get the warning "The collection of supported languages is empty" and if I restart, Audiveris asks me to download the languages again, although the files are still in the folder as shown in the screenshot.

If I read you correctly:

1. After a fresh installation, you get the warning "The collection of supported languages is empty" each time you launch the program. This the normal initial behavior.
2. You have downloaded a few languages. OK.
3. From now on, if you exit and restart the program, you don't get the warning. Is this correct?
4. But if you transcribe a .pdf file, you get the warning.

Please confirm the point 3.

[HerGeo]

Sorry for the confusion: On restart, I get the following message, every time regardless of what I do:

I also checked the constant useOCR and it is checked.
Seems, that Audiveris downloads the file into the correct folder, but tesseract tries to read another one/has no access or whatsoever. Maybe this has to do something with not being admin user?

[HerGeo]

Indeed, when using admin user, after downloading the files once, Audiveris can find the files and does not ask for installation of languages at restart. Tried again with non-admin user, with no success.
My assumption is, that the path for tesseract gets set during installation (with admin rights?) and therefore cannot be accessed in normal operation as non admin-user.

[hbitteur]

My assumption is, that the path for tesseract gets set during installation (with admin rights?) and therefore cannot be accessed in normal operation as non admin-user.

The target for Tesseract language files is:

OCR folder: C:\Users\XXX\AppData\Roaming\AudiverisLtd\audiveris\config\tessdata

Where "XXX" is your user's name.

Could you check who is the owner of this tessdata folder?
You shouldn't have to use admin rights to access it.

When you installed Audiveris, were you a standard user or an admin?
When you launch Audiveris, I suppose you are a standard user (except the case you tried with admin rights).
And when you download language files, still as a standard user, Audiveris is able to write to the tessdata folder. So why can't you access this folder for reading?

I'm lost.

[hbitteur]

According to ChatGPT:

To find out who owns a folder in Windows, you can use the following steps:

Method 1: Using Properties
Right-click on the folder.
Select Properties.
Go to the Security tab.
Click on the Advanced button.
In the Advanced Security Settings window, you’ll see the Owner listed at the top.
If you want to change the owner, click on the Change link next to the owner's name.

Method 2: Using Command Prompt (takeown)
Open Command Prompt as Administrator. (Right-click on Start > Command Prompt (Admin) or PowerShell (Admin)).

Run the following command to check ownership:

icacls "C:\path\to\your\folder"

This will list the folder's permissions, including the owner.

Alternatively, to change the owner, you can run:

takeown /f "C:\path\to\your\folder" /r /d y

This command changes the ownership to the current user.