feat(ci): Use cached ncbi dataset package to isolate tests

Resolves #46

ingest/build-configs/ci/config.yaml

@@ -1,5 +1,5 @@
-# TODO: If the ingest workflow ever runs too long, we should figure out a way
-# to subset the ingest data. Currently, the CI just runs the default ingest workflow.
+# Use cached ncbi datasets package to speed up tests and isolate from ncbi servers
+mock_fetch: true
 
 # Snakemake requires at least one top level key in a config file, so including
 # a bogus key here that should not be used anywhere in the Snakemake workflow

ingest/rules/fetch_from_ncbi.smk

@@ -45,9 +45,18 @@ rule fetch_ncbi_dataset_package:
         """
 
 
+def get_ncbi_dataset_package_path():
+    """
+    Use cached data package in ci to isolate from ncbi server
+    """
+    if config.get("mock_fetch", False):
+        return "test_data/ncbi_dataset.zip"
+    return "data/ncbi_dataset.zip"
+
+
 rule extract_ncbi_dataset_sequences:
     input:
-        dataset_package="data/ncbi_dataset.zip",
+        dataset_package=get_ncbi_dataset_package_path(),
     output:
         ncbi_dataset_sequences=temp("data/ncbi_dataset_sequences.fasta"),
     benchmark:
@@ -61,7 +70,7 @@ rule extract_ncbi_dataset_sequences:
 
 rule format_ncbi_dataset_report:
     input:
-        dataset_package="data/ncbi_dataset.zip",
+        dataset_package=get_ncbi_dataset_package_path(),
     output:
         ncbi_dataset_tsv=temp("data/ncbi_dataset_report.tsv"),
     params: