rsimon / bm-egypt

Data enrichment for the BM Egyptian place vocabulary

BM Egypt

Data enrichment utilities for the BM Egyptian place vocabulary.

GeoNames

Harvester scripts that queries the GeoNames API endpoint with the 'Place name' value, and collects the results into a folder of JSON files. Each file contains the original record and all result candidates from the API. Naming convention is that the JSON file is named after the BM System ID number .json. There are two scripts:

• Harvester for the standard GeoNames searchJSON endpoint
• Harvester for the wikipediaSearchJSON GeoNames Wikipedia search
• Utility scripts reconcile_geonames.py and reconcile_wikipedia.py that filter the API responses, pick the most plausible one for each place, and produce a flat CSV result file.

Pleiades

Utilities to build a searchable index of Pleiades places using Meilisearch and the Pleiades GIS data download.

• The setup uses data from the Pleiades places.csv and names.csv. The script create-index-dataset.py merges the relevant data from the tables and produces a JSON dump for indexing in Meilisearch.
• Deployment of Meilisearch via a docker-compose.yml file. Run docker compose up, Meilisearch is at http://localhost:7700. With Meilisearch running, run the index-pleiades.py script.
• The harvest-pleiades-meili.py script harvests candidate results for each place name in the BM CSV.
• Utility script reconcile_pleiades.py which filters the responses, picks the most plausible one and produces a flat CSV result file.

Joined Results

The script join_results.py will read the three individual result files (GeoNames, Wikipedia, Pleiades) and produce one joint CSV.