harness updates to accommodate DatabaseUpdater

microbiomedata · Dec 27, 2024 · ae220be · ae220be
1 parent 8af1493
commit ae220be
Show file tree

Hide file tree

Showing 4 changed files with 71 additions and 11 deletions.
diff --git a/nmdc_runtime/site/graphs.py b/nmdc_runtime/site/graphs.py
@@ -474,8 +474,10 @@ def nmdc_study_to_ncbi_submission_export():
 
 @graph
 def fill_missing_data_generation_data_object_records():
-    study_id = get_database_updater_inputs()
-    database = missing_data_generation_repair(study_id)
+    (study_id, gold_nmdc_instrument_mapping_file_url) = get_database_updater_inputs()
+    gold_nmdc_instrument_map_df = get_df_from_url(gold_nmdc_instrument_mapping_file_url)
+
+    database = missing_data_generation_repair(study_id, gold_nmdc_instrument_map_df)
 
     database_dict = nmdc_schema_object_to_dict(database)
     filename = nmdc_study_id_filename(study_id)

diff --git a/nmdc_runtime/site/ops.py b/nmdc_runtime/site/ops.py
@@ -1255,19 +1255,49 @@ def nmdc_study_id_filename(nmdc_study_id: str) -> str:
 
 
 @op(
-    config_schema={"nmdc_study_id": str},
-    out={"nmdc_study_id": Out(str)},
+    config_schema={
+        "nmdc_study_id": str,
+        "gold_nmdc_instrument_mapping_file_url": str,
+    },
+    out={
+        "nmdc_study_id": Out(str),
+        "gold_nmdc_instrument_mapping_file_url": Out(str),
+    },
 )
-def get_database_updater_inputs(context: OpExecutionContext) -> str:
-    return context.op_config["nmdc_study_id"]
+def get_database_updater_inputs(context: OpExecutionContext) -> Tuple[str, str]:
+    return (
+        context.op_config["nmdc_study_id"],
+        context.op_config["gold_nmdc_instrument_mapping_file_url"],
+    )
 
 
-@op(required_resource_keys={"runtime_api_user_client"})
+@op(
+    required_resource_keys={
+        "runtime_api_user_client",
+        "runtime_api_site_client",
+        "gold_api_client",
+    }
+)
 def missing_data_generation_repair(
-    context: OpExecutionContext, nmdc_study_id: str
+    context: OpExecutionContext,
+    nmdc_study_id: str,
+    gold_nmdc_instrument_map_df: pd.DataFrame,
 ) -> nmdc.Database:
-    client: RuntimeApiUserClient = context.resources.runtime_api_user_client
-    database_updater = DatabaseUpdater(client, nmdc_study_id)
-    database = database_updater.get_database()
+    runtime_api_user_client: RuntimeApiUserClient = (
+        context.resources.runtime_api_user_client
+    )
+    runtime_api_site_client: RuntimeApiSiteClient = (
+        context.resources.runtime_api_site_client
+    )
+    gold_api_client: GoldApiClient = context.resources.gold_api_client
+
+    database_updater = DatabaseUpdater(
+        runtime_api_user_client,
+        runtime_api_site_client,
+        gold_api_client,
+        nmdc_study_id,
+        gold_nmdc_instrument_map_df,
+    )
+    database = database_updater.create_missing_dg_records()
 
     return database
diff --git a/nmdc_runtime/site/repository.py b/nmdc_runtime/site/repository.py
@@ -940,12 +940,28 @@ def database_record_repair():
                                 "password": {"env": "API_ADMIN_PASS"},
                             },
                         },
+                        "runtime_api_site_client": {
+                            "config": {
+                                "base_url": {"env": "API_HOST"},
+                                "client_id": {"env": "API_SITE_CLIENT_ID"},
+                                "client_secret": {"env": "API_SITE_CLIENT_SECRET"},
+                                "site_id": {"env": "API_SITE_ID"},
+                            },
+                        },
+                        "gold_api_client": {
+                            "config": {
+                                "base_url": {"env": "GOLD_API_BASE_URL"},
+                                "username": {"env": "GOLD_API_USERNAME"},
+                                "password": {"env": "GOLD_API_PASSWORD"},
+                            },
+                        },
                     },
                 ),
                 "ops": {
                     "get_database_updater_inputs": {
                         "config": {
                             "nmdc_study_id": "",
+                            "gold_nmdc_instrument_mapping_file_url": "https://raw.githubusercontent.com/microbiomedata/nmdc-schema/refs/heads/main/assets/misc/gold_seqMethod_to_nmdc_instrument_set.tsv",
                         }
                     },
                     "export_json_to_drs": {"config": {"username": ""}},

diff --git a/nmdc_runtime/site/resources.py b/nmdc_runtime/site/resources.py
@@ -370,6 +370,18 @@ def fetch_study(self, id: str) -> Union[Dict[str, Any], None]:
             return None
         return results[0]
 
+    def fetch_projects_by_biosample(self, biosample_id: str) -> List[Dict[str, Any]]:
+        id = self._normalize_id(biosample_id)
+        results = self.request("/projects", params={"biosampleGoldId": id})
+        return results
+
+    def fetch_biosample_by_biosample_id(
+        self, biosample_id: str
+    ) -> List[Dict[str, Any]]:
+        id = self._normalize_id(biosample_id)
+        results = self.request("/biosamples", params={"biosampleGoldId": id})
+        return results
+
 
 @resource(
     config_schema={