Andy Halterman

Thanks for the catch on the encoding. I added the encoding to the [code](https://github.com/openeventdata/es-geonames/commit/155f7c977ebc384cea2b7fe91b251f840563cea0) so that shouldn't happen for other people in the future. On the main issue, the ingestion...

Thanks for the catch on the encoding. I added the encoding to the [code](https://github.com/openeventdata/es-geonames/commit/155f7c977ebc384cea2b7fe91b251f840563cea0) so that shouldn't happen for other people in the future. On the main issue, the ingestion...

Good catch! [Here's](https://github.com/ahalterman/mordecai3/blob/bb05976f03d543b2b57543c5fb1dada8b5787247/mordecai3/geoparse.py#L136) where the bug is. I define it as `"end_char": ent[-1].idx + len(ent.text)}`, but it should be `"end_char": ent[-1].idx + len(ent[-1].text)}`. Basically, I'm adding the whole length of...

The [`spacy-3-8`](https://github.com/ahalterman/mordecai3/tree/spacy-3-8) branch has a working fix for this bug. @andybega is doing some awesome work cleaning up the code and adding some tests. Once that's done, we'll merge the...

Should be closed by #34.

Hey Dominik, Thanks for these comments and suggestions. This code all grew out of some specific use cases, so there's some path dependency baked in that we could probably revisit....

I've had a lot of success using small LLMs with structured JSON output (e.g., Qwen3-4B or even 0.6B) and I think that's a great idea! (As an aside, I lived...