As part of #2673, task 3 was to ascertain if there was a test that could be written to determine if the wrong organisation was assigned to an entity. This ask came off the back of Drew's ticket, where he found instances of entities where WOK should be WOI, and CHW should be CHE.
I conducted some analysis to figure out how widespread this problem was. The notebook can be found here: https://github.com/digital-land/jupyter-analysis/tree/main/analysis/2026-07_correct_organisations_of_entities
My analysis found that the problem wasn't as widespread as originally thought:
This ticket will require:
Note: If source.csv does need updating, a new source hash will have to be generated. This is generated by source = MD5("collection|organisation|endpoint") (one MD5 over the whole combined string), i.e. a single 32-char hex value, no pipes. E.g. title-boundary|government-organisation:D69|7058af9c6a0f85d6101c59484d83aa411ebb108ff0c653c2274aac2a7839168f
would generate a source hash of: bf8c84e7748d3579ed73ab7358aeae45
As part of #2673, task 3 was to ascertain if there was a test that could be written to determine if the wrong organisation was assigned to an entity. This ask came off the back of Drew's ticket, where he found instances of entities where WOK should be WOI, and CHW should be CHE.
I conducted some analysis to figure out how widespread this problem was. The notebook can be found here: https://github.com/digital-land/jupyter-analysis/tree/main/analysis/2026-07_correct_organisations_of_entities
My analysis found that the problem wasn't as widespread as originally thought:
This ticket will require: