Skip to content

Custom Metadata Fields

APM supports extending job metadata through two mechanisms: the Admin UI (no code required) and direct database manipulation.

  1. Go to Admin → Metadata Options.
  2. Select the group where the new key belongs:
  3. Dublin Core — dc:* namespace fields
  4. eBraille Profile — production-specific fields
  5. METS / PREMIS — archival metadata fields
  6. Click + Add Key.
  7. Enter the metadata key (e.g. dc:audience) and display label.
  8. Click Save.

The key appears immediately in all metadata editor dropdowns and the Additional Allowed Fields section of job detail views.

Via Database

Insert directly into material_category:

INSERT INTO material_category (section, value, label, sort_order)
VALUES ('metadata_dublin_core', 'dc:audience', 'Audience', 16);

Valid sections: metadata_dublin_core, metadata_ebraille_profile, metadata_mets_premis.

Storing Values

Metadata is stored in the job_metadata table as key-value pairs:

INSERT INTO job_metadata (job_type, job_id, meta_key, meta_value)
VALUES ('braille', 42, 'dc:audience', 'Students Grade 7-9');

The Python helper:

from accessibility_mgr.db import queries as Q
Q.set_job_metadata("braille", 42, "dc:audience", "Students Grade 7-9")

Querying Custom Metadata

# Get all metadata for a job
meta = Q.list_job_metadata("braille", 42)
# {'dc:title': '...', 'dc:audience': 'Students Grade 7-9', ...}

# Get a single field
val = Q.get_job_metadata("braille", 42, "dc:audience")

# Search across all jobs
results = Q.search_all("Students Grade 7")
# results["metadata"] contains matching key-value pairs

Backfilling Typo Keys

If historical data has inconsistent key names, use the backfill tool (Admin → Metadata Options → Backfill Typo Keys) to preview and apply fuzzy-matched normalisation.

See Categories for the full workflow.