fix: catch ETL module import errors (#165)

README.md

@@ -47,6 +47,12 @@ export DISEASE_NORM_DB_URL="postgresql://postgres@localhost:5432/disease_normali
 
 Use the `disease_norm_update` command in a shell to update the database.
 
+If you encounter an error message like the following, refer to the installation instructions above:
+
+```shell
+"Encountered ModuleNotFoundError attempting to import Mondo. Are ETL dependencies installed?"
+```
+
 #### Update source(s)
 
 The Disease Normalizer currently uses data from the following sources:

pyproject.toml

@@ -48,7 +48,7 @@ Source = "https://github.com/cancervariants/disease-normalization"
 "Bug Tracker" = "https://github.com/cancervariants/disease-normalization/issues"
 
 [project.scripts]
-disease_norm_update = "disease.cli:update_normalizer_db"
+disease_norm_update = "disease.cli:update_db"
 disease_norm_update_remote = "disease.cli:update_from_remote"
 disease_norm_dump = "disease.cli:dump_database"
 disease_norm_check_db = "disease.cli:check_db"

src/disease/cli.py

@@ -14,16 +14,8 @@
     DatabaseWriteException,
     create_db,
 )
-from disease.etl import DO, OMIM, Mondo, NCIt, OncoTree  # noqa: F401
-from disease.etl.merge import Merge
 from disease.schemas import SourceName
 
-# Use to lookup class object from source name. Should be one key-value pair
-# for every functioning ETL class.
-SOURCES_CLASS_LOOKUP = {
-    s.value.lower(): eval(s.value) for s in SourceName.__members__.values()
-}
-
 
 @click.command()
 @click.option("--db_url", help="URL endpoint for the application database.")
@@ -176,6 +168,13 @@ def _load_source(
     start_load = timer()
 
     # used to get source class name from string
+    try:
+        from disease.etl import DO, OMIM, Mondo, NCIt, OncoTree  # noqa: F401
+    except ModuleNotFoundError as e:
+        click.echo(
+            f"Encountered ModuleNotFoundError attempting to import {e.name}. Are ETL dependencies installed?"
+        )
+        click.get_current_context().exit()
     SourceClass = eval(n.value)  # noqa: N806
 
     source = SourceClass(database=db)
@@ -219,6 +218,15 @@ def _load_merge(db: AbstractDatabase, processed_ids: Set[str]) -> None:
         processed_ids = set()
         for source in SOURCES_FOR_MERGE:
             processed_ids |= db.get_all_concept_ids(source)
+
+    try:
+        from disease.etl.merge import Merge
+    except ModuleNotFoundError as e:
+        click.echo(
+            f"Encountered ModuleNotFoundError attempting to import {e.name}. Are ETL dependencies installed?"
+        )
+        click.get_current_context().exit()
+
     merge = Merge(database=db)
     click.echo("Constructing normalized records...")
     merge.create_merged_concepts(processed_ids)