Closed Ferrisx4 closed 3 years ago
Oh @Ferrisx4 . I'm sorry. The problems here are related to the OBO Loader. There is a PR to fix it here: https://github.com/tripal/tripal/pull/1117.
Perhaps you could now, pull that code from the PR and retest it? That would be sufficient as a functional test for that PR
Hi @spficklin, I pulled the PR and tested it. It was successful other than a few warnings during step #4:
Step 4: Loading terms... %. Memory: 59,168,560 bytes.
Cannot add an Alt ID without an accession: 'topic_0197'
Cannot add an Alt ID without an accession: 'topic_3076'
Cannot add an Alt ID without an accession: 'topic_3040'
Step 5: Cleanup...100.00 %. Memory: 59,433,896 bytes. ```
Yeah, those warnings are from badly formatted alt_ids in the OBO. I'm not sure if those were historically made incorrectly or if there were entered into the OBO incorrect. But it doesn't inhibit loading of the vocabulary.
Would you mind leaving a comment about the successful test on the PR? I'll mark it has having been approved and that should just trigger one more code review.
I think this issue is now resolved.
When launching the job to load the EDAM vocabulary, the server gets a bad response from EBI:
This continues on for every term (~3000 - http://edamontology.org/EDAM.obo), and then the same for 'format', 'operation', and 'topic'.
Letting it run to completion (instead of cancelling it) results in a Division by Zero error from TripalImporter.inc:646 followed by evidently not loading any type defs or terms:
The double slash in the URL does not seem to be the cause.