49 lines
1.2 KiB
Python
49 lines
1.2 KiB
Python
import polars as pl
|
|
|
|
from pcdigitizer import Annotation, PubChemAPI
|
|
|
|
|
|
def test_source_df(mock_session):
|
|
df = PubChemAPI.get_sources(session=mock_session)
|
|
|
|
columns = df.columns
|
|
assert len(columns) == 23
|
|
assert columns[0] == "Source Name"
|
|
assert columns[-3] == "License URL"
|
|
|
|
source_name = "Hazardous Substances Data Bank"
|
|
source_count = df.select(
|
|
(pl.col("Source Name").str.contains(source_name)).sum()
|
|
).item()
|
|
assert source_count == 1
|
|
|
|
|
|
def test_source_annotations(mock_session):
|
|
annotations = PubChemAPI.get_source_annotations(
|
|
"Hazardous Substances Data Bank (HSDB)", session=mock_session
|
|
)
|
|
assert len(annotations["Compound"]) == 246
|
|
|
|
|
|
def test_annotations(mock_session):
|
|
annotations = PubChemAPI.get_annotations(session=mock_session)
|
|
annotation_types = sorted(annotations.keys())
|
|
assert annotation_types == [
|
|
"Assay",
|
|
"Cell",
|
|
"Compound",
|
|
"Element",
|
|
"Gene",
|
|
"Pathway",
|
|
"Protein",
|
|
"Taxonomy",
|
|
]
|
|
assert "Dissociation Constants" in annotations["Compound"]
|
|
|
|
|
|
def test_get_data(mock_session):
|
|
data = PubChemAPI.get_data(
|
|
Annotation.DISSOCIATION_CONSTANTS, 1, session=mock_session
|
|
)
|
|
assert len(data) == 1000
|