NeoDIVA .lab Phoneme Error Checker

This Python script crawls a directory, finds subfolders containing .lab files, and validates the phonemes within them based on a language tag in the folder's name. This script uses all the phoneme sets included in the NeoDIVA DiffSinger model, but should be applicable to most datasets.

It checks for:

• Typos: Phonemes not found in the global list or the specified language's phoneme set. (AP, SP, vf, q, cl, ctrash, vtrash and trash)
• Suggestions: pau (suggests SP) and br (suggests AP).
• Tagged Phonemes: Validates phonemes with language tags (e.g., ja/a).

1. Supported Languages

The script recognizes folder names ending in the following language tags (case-insensitive):

• _en English (Arpasing+ Phoneme Set)
• _es Spanish (Namine Criss Phoneme Set)
• _fr French (Millefeuille Phoneme Set)
• _it Italian (Gianloop Phoneme Set)
• _ja Japanese (Standard Phoneme Set)
• _ka Georgian (Datawave Phoneme Set)
• _ko Korean (Team CODA Phoneme Set)
• _pl Polish (PixPrucer Phoneme Set)
• _pt Portuguese (Brazillian, Team BRAPA Phoneme Set)
• _ro Romanian (NeoDIVA Phoneme Set)
• _ru Russian (Lunai Phoneme Set)
• _th Thai (PrintMOV Phoneme Set)
• _zh Chinese (TGM Phoneme Set)
• _hy Armenian (amihopeutau Phoneme Set)

2. Save the Script

Save the Python file anywhere on your computer, ideally in documents in it's own folder for organization.

3. Prepare Your Folders

This is the most important step. The script is designed to find language codes in your folder names.

• A folder's name must end with an underscore (_) followed by one of the two-letter language codes listed above (e.g., _en, _es, _ja).
• The script will skip any folder that does not match this format.

Correct Folder Naming Examples:

C:\My_Data\Speaker_01_en
C:\My_Data\Recordings_fr
C:\My_Data\Section_4\Speaker_10_ko

Incorrect Folder Naming Examples:

• C:\My_Data\English_Files (The script will skip this folder because it has no language tag).
• C:\My_Data\Speaker_pt_files (The script will skip this because the _pt tag is not at the very end of the folder name).

All your .lab files should be placed directly inside these language-tagged folders.

4. Run the Script

1. Open your terminal (like Command Prompt, PowerShell, or Windows Terminal).

2. Use the cd command to move to the folder where you saved the check_phonemes.py file.

   # Example:
   cd C:\Users\Example\Documents\LabelChecker

3. Once you are in the correct folder, run the script:

   python check_phonemes.py

   (You may need to use py instead of python depending on your setup).

5. Enter the Root Folder Path

1. The script will ask you:

   Please enter the path to the root folder to scan:
   

2. This is the main folder that contains all your language-tagged folders (like C:\My_Data from the example above).
3. Copy the full path to this root folder, paste it into the terminal, and press Enter.

6. Check the Log Files

The script will scan all the subfolders.

• When it is finished, look inside your root folder (the path you just gave the script).
• For every language-tagged folder it processed (e.g., Speaker_01_en), it will create a new log file in that root directory (e.g., Speaker_01_en_errors.log).
• Open these .log files to see the results.
  • If any typos, suggestions (pau/br), or malformed lines were found, they will be listed with the file name and line number.
  • If a folder's labels were all correct, its log file will simply say "No errors found."