Best method for importing large number of people tags in digiKam

[Marc Palaus]

I'd maybe write a script that reads that information from your SQLite 
table, and use exiftool to check for existing tags in each of the 
images, and write to them if necessary. I wouldn't involve Digikam in 
that process. Once all the images have the proper tags and face regions 
in them, you can let Digikam read all that metadata and update its database.

El 4/7/2024 a les 16:39, Sepp A ha escrit:
> Dear all,
>
> I have an SQLite table with approximately 43,000 filenames as the 
> rows, each associated with a column listing people names (separated by 
> a comma and space). I am aiming to import these "/People" tags into 
> digiKam. Some tags already exist, so I need a script that will check 
> for existing tags, create new ones if needed, and tag the 
> corresponding files accordingly.
>
> I am relatively new to working with databases, and the directory 
> "\.mysql.digikam\db_data" contains many confusing files. I would 
> greatly appreciate any guidance or assistance on how to approach this 
> task. Furthermore, I've read a bit about XMP sidecar files, but I 
> think adding and cleaning them up afterward might be a bit tedious. 
> Would you recommend using XMP sidecar files over directly modifying 
> the digiKam database?
>
> Here are some details about my digiKam setup:
> - digiKam version: 8.3.0
> - Database: MySQL Internal (MariaDB) on Windows 10, stored on a SSD
> - Database backend: QMYSQL
> - Total items: ~195,000
> - Albums: ~1,200
> - Tags: ~2,500
>
> Despite using MariaDB, digiKam takes a considerable time to start and 
> often freezes when handling large albums.
>
> Thank you in advance for your help.