Management Commands

Available Commands

check_image_files

This command aims to repair incorrect file extensions.

In certain error cases a warning will be issued:

• image file not found.

• image cannot be verified by Pillow.

clean_empty_media_folders

This command deletes empty media folders. The media directory can be found in ./src/assets/media, or for a docker based setup it’s the directory configured as MEDIA_DIR in the .env file.

clean_artwork_fields

This command checks the fields title & title_en of the Artwork model and if present removes non-printable characters by replacing them with spaces.

create_image_fullsize

This command generates an image_fullsize for every artwork, by converting the image_original to an image of the same size and in a (standardized) JPEG format with Pillow.

import_external_metadata

This command maps identifiers from external sources (e.g., GND, Getty, Wikidata) for Persons, Locations and Keywords via CSV files, and updates corresponding entries in the database with external data. For more information, please read the External Metadata documentation.

• It’s important to note that the gnd_overwrite-flag is set a bit differently:

  • By default the flag will be turned on.

  • In case of a name mismatch between the database entry and the fetched GND data, all data will be overwritten, except for the name field and the overwrite flag will be turned off.

  • The differing name retrieved from the GND response is added to the first index of the synonyms field instead.

  • Editors can review these synonyms to decide whether to update the primary name manually (or by activating the overwrite flag again).

Arguments

Optional

• -s, --skip-header Skips the first row of the CSV file if it contains headers.

• -p, --progress Displays progress updates while processing the file.

Positional

• type

  The type of data to be imported. Must be one of:

  • artist

  • location

  • keyword

• file The full path to the CSV file containing the mapping of labels to external source IDs. The file should have the format:

  name;source_id

Usage examples

A common usage might look like:

python manage.py import_external_metadata --progress artist artist_data.csv

This assumes artist_data.csv contains a list of person names and their corresponding GND IDs (without any header line in the CSV). For a local development setup, you might want to use a relative path, such as ./data/artist_data.csv.

For a containerized setup, (temporarily) copy your CSV file (e.g. artist_data.csv) into the src/ directory of the repo and execute the command

docker compose exec image-django python manage.py import_external_metadata artist artist_data.csv

repair_image_paths

This command checks all images, if the filenames conform to the latest naming schema. If the filename does not conform, the file (and its references in the database) will be updated to the current schema.

Note

In a 2.x install this command should not be needed. But if you are migrating an older Image instance (e.g. from version 1.x), you might need this.

update_search_vector

This command updates the search vector for all artworks, which is used for a performant full-text search on artworks. This is useful when a new version with an updated search logic is deployed, as well as in local development.