Updated discovery for new FHIR resources #300
Conversation
☂️ Python Coverage
Overall Coverage
New FilesNo new covered files... Modified Files
|
tests/conftest.py
Outdated
| @@ -178,6 +178,7 @@ def debug_table_head(cursor, table, rows=3, cols="*"): | |||
| def debug_diff_tables(cols, found, ref, pos=0): | |||
| cols = cols if len(cols) > 0 else [] | |||
| found = found[pos] if len(found) > pos else [] | |||
| print(str(ref)) | |||
tests/test_discovery.py
Outdated
| for i in range(0, len(table_rows)): | ||
| table_rows[i] = tuple([x if x else "" for x in table_rows[i]]) |
There was a problem hiding this comment.
nit: it's odd to my Python eyes to see an actual index loop 😄 maybe something like this? Could be cut over several lines for clarity.
table_rows = [tuple(x or "" for x in row) for row in table_rows]
There was a problem hiding this comment.
i personally find this more obtuse, but i agree it's more pythonic, so i'll make this change.
There was a problem hiding this comment.
I only generally like that construction with good formatting:
table_rows = [
tuple(x or "" for x in row)
for row in table_rows
]
Which I find readable enough - but if you vibe your approach for readability, I think that's a good enough reason to keep it as-is.
tests/test_discovery.py
Outdated
| "/test_data/discovery/discovery__code_sources.csv", | ||
| "w", | ||
| ) as f: | ||
| writer = csv.writer(f, quoting=csv.QUOTE_MINIMAL) | ||
| for row in table_rows: | ||
| writer.writerow(row) |
There was a problem hiding this comment.
Should comment this block out yeah?
There was a problem hiding this comment.
I could've sworn i'd done this.
| '', | ||
| '', | ||
| '' |
There was a problem hiding this comment.
This reference file has a lot of blanks - is there no coding data in the test data for allergies or these others? Procedure.reasonCode has something. But it's a rarity. Even encounter.hospitalization.dischargedisposition lost data.
There was a problem hiding this comment.
This is probably due to running against my test db - my internet access has been so bad from home over the last few days on VPN that i got bizaare timeouts trying to run this against cerner. I'll try it again tomorrow when i'm onsite.
There was a problem hiding this comment.
I think in CONTRIBUTING.md, we have a one-liner to run against the test data - does that generate good-enough results?
There was a problem hiding this comment.
updated w/ local db - still pretty blank-ful.
This adds allergyintolerance, diagnosticreport, and procedure to resources the discovery study checks for coding information.
It also lightly redoes test data generation/validation logic due to some issues with commas inside fields.
Checklist
docs/needs to be updatedgenerate-mdcorestudy fields that not in US Core, update our list of those incore-study-details.mdmanifest.toml