Code review is looking mostly good :-) Just one backwards incompatible change that we should resolve.

Manual Testing

GFF3 importer.

The change was in insertFeatureSeqs() so I tested

• using this test fixture tripal_chado/tests/fixtures/gff3_loader/small_gene_with_fasta.gff
• imported it with a default srcfeature type of scaffold
• created a scaffold content type and published it
• ✅ The residues and associated metadata showed up as expected

Newick Tree Loader

The change here was in the form when the sequence ontology is not present so I tested

• by updating the name of the sequence ontology to Fred which did trigger the code... However, apparently the original code was checking too late and other errors appeared.
• As such I make the code change in commit f73a8e6 with @dsenalik blessing
• ✅ Now loading the form shows the warning as expected.

Manual Testing

OBO Importer

The change here was in the stanza caching. This should be impacted on most obo loads so I tested by

• I used the local EDAM subset suggested by Doug
• On the obo imporer I selected "Add a New Ontology", included the path provided by Doug and set the same to "Local EDAM"
• ✅ I then ran the job and confirmed there were no errors as expected. If the changes had not worked, there would have been PDO errors.
  

Taxonomy Importer

Changes here were made in the following methods: findOrganism(), addOrganism(), and importRecord(). These will be hit on all jobs for this importer that has at least one organism. As such I tested by

• Created the organism Daucus muricatus which matches one of the ids in the next step through the Add Organism form (to test find when one exists).

• Imported the following ids 48026 48106 79170 79196 79200 180796 using the "NCBI Taxonomy Importer" form and selected "Import details for existing species"

• When I ran the job it failed with the following error showing the updated query in addOrganism() did not work.
  

• ❌ I fixed this by adding the field to be selected to one of the queries (see commit 37bbd17) but it still failed as follows. Note: While I did add the one organism ahead of time and it does have an organism_id of 5, it did not have a dbxref_id assigned via organism_dbxref before the import.

  sitedb=# select * from teapot.organism;
   organism_id | abbreviation |  genus   |  species   | common_name | infraspecific_name | type_id | comment 
  -------------+--------------+----------+------------+-------------+--------------------+---------+---------
             1 |              | Tripalus | databasica |             |                    |         | 
             5 |              | Daucus   | muricatus  |             |                    |         | 
  (2 rows)
  
  sitedb=# select * from teapot.organism_dbxref;
   organism_dbxref_id | organism_id | dbxref_id 
  --------------------+-------------+-----------
  (0 rows)

  

Manual Testing

✅ GFF3 Importer
✅ Newick Tree Loader
✅ OBO Importer
✅ Taxonomy importer
✅ Tree Generator
✅ FASTA Importer

Taxonomy Importer

Changes here were made in the following methods: findOrganism(), addOrganism(), and importRecord(). These will be hit on all jobs for this importer that has at least one organism. As such I tested by

• Created the organism Daucus muricatus which matches one of the ids in the next step through the Add Organism form (to test find when one exists).
• Imported the following ids 48026 48106 79170 79196 79200 180796 using the "NCBI Taxonomy Importer" form and selected "Import details for existing species"
• When I ran the job, it completed without error as expected :-)
• ✅ After finding the new fields and re-publishing, I was able to confirm all the details are present both on the pre-existing Daucus muricatus and one other organism.
  
  

Taxonomy Tree Generator

Changes were made in the run(), retrieveLineage(), removeTree(), and queryPhylonode(). I believe these changes will be best fit by regenerating an existing tree and trying to generate a tree that doesn't have an appropriate root organism.

• Tried to generate tree where the root taxon was "fabaceae" despite there being no organisms in the site for this family.
• ✅ Confirmed that the job ran with the expected error and that the initial record added by the importer was removed leaving an empty phylotree table.
  
• Used the importer to generate a tree for Daucus genus which there are 4 organism in the site.
• ✅ the job ran without error, confirmed there were 4 organism which should be included and 2 which should not. Selecting from the database shows there are the expected phylonode records 🎉
  

FASTA Importer

Changes are made to loadFastaFeature(), specifically with reference to updating an existing feature. As such we are testing by

• Import features using the GFF3 importer and /var/www/drupal/web/modules/contrib/tripal/tripal_chado/tests/fixtures/gff3_loader/gff_Citrus_sinensis-orange1.1g015632m.g.gff3
• Import the sequence for orange1.1g015632m using the FASTA importing using the /var/www/drupal/web/modules/contrib/tripal/tripal_chado/tests/fixtures/fasta_loader/Citrus_sinensis-orange1.1g015632m.g.fasta
• ✅ I confirmed that the sequence was imported and assigned to the residues field of the existing gene.