Standardises outputs (and in some cases inputs) as per #223:
agent -> agent_id
id/link -> link_id
I've also changed the way that geojson is written to disk, such that the geojsons maintain the name of the index field (previously they are renamed to "id"). This effectively removes the index from the geojson and adds a new field for our index with the correct name (called "link_id" for example).
I have not changed seq to seq_id because seq is supposed to be the sequence or "enumerator", not an index. Happy to argue this case.
Changes like this are very painful in elara, because they cascade throughout the handlers, the tests, the test fixtures and the input benchmarks (and the docs). These changes will also impact downstream users. In particular ersa and scripts using ersa. So please consider this change with care and plan ahead accordingly.
Standardises outputs (and in some cases inputs) as per #223:
I've also changed the way that geojson is written to disk, such that the geojsons maintain the name of the index field (previously they are renamed to "id"). This effectively removes the index from the geojson and adds a new field for our index with the correct name (called "link_id" for example).
I have not changed seq to seq_id because seq is supposed to be the sequence or "enumerator", not an index. Happy to argue this case.
Changes like this are very painful in elara, because they cascade throughout the handlers, the tests, the test fixtures and the input benchmarks (and the docs). These changes will also impact downstream users. In particular ersa and scripts using ersa. So please consider this change with care and plan ahead accordingly.