Add filetype to all images in the catalog DB

[obulat]

Current Situation

There are currently 563 004 660 images without a file type in the database.
We need to add the filetype information to all images.

Suggested Improvement

There are several things that need to be done here:

• Audit the provider scripts for filetype extraction. Add the necessary steps, if necessary.
• Identify the providers that don't have filetype data in the database.
• Create a DAG for re-downloading the data from each of the providers from the previous step, and run it to refresh the data. This will probably need to be coordinated with other data collection (filesize, width, height).

Benefit

Data consistency.

Additional context

We are currently also extracting extensions during the Elasticsearch indexing:
https://github.com/WordPress/openverse-api/blob/2e85caf7aede8aaf9d77cd5cb050f50b860ee58e/ingestion_server/ingestion_server/elasticsearch_models.py#L135-L144
This should be done entirely in the catalog.

Here, we should also make sure that we don't create duplicate filetype for types such as jpg/jpeg.

Implementation

• 🙋 I would be interested in implementing this feature.