Allow comma-separated parent_sample_id values in URL checking (#1320)

* allow comma-separated values in parent_sample_id field
hubmapconsortium · Apr 10, 2024 · 4fc83e6 · 4fc83e6
1 parent e25d263
commit 4fc83e6
Show file tree

Hide file tree

Showing 3 changed files with 40 additions and 22 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -4,6 +4,7 @@
 - Directory validation changes for "shared" uploads
 - Update Phenocycler directory schema
 - Remove bad paths from LC-MS directory schema
+- Allow multiple comma-separated parent_sample_id values
 
 ## v0.0.18
 

diff --git a/examples/dataset-examples/bad-cedar-assay-histology/README.md b/examples/dataset-examples/bad-cedar-assay-histology/README.md
@@ -13,9 +13,8 @@ Metadata TSV Validation Errors:
     examples/dataset-examples/bad-cedar-assay-histology/upload/bad-histology-metadata.tsv:
       URL Errors:
       - 'On row 2, column "parent_sample_id", value "wrong" fails because of error
-        "HTTPError": 401 Client Error: Unauthorized for url: https://entity.api.hubmapconsortium.org/entities/wrong.'
-      - 'On row 3, column "parent_sample_id", value "HBM854.FXDQ.783" fails because
-        of error "HTTPError": 401 Client Error: Unauthorized for url: https://entity.api.hubmapconsortium.org/entities/HBM854.FXDQ.783'
+        "HTTPError": Field value is not valid; URL https://entity.api.hubmapconsortium.org/entities/wrong
+        returned a 400 Error.'
       Validation Errors:
       - On row 0, column "parent_sample_id", value "wrong" fails because of error
         "invalidValueFormat".
@@ -27,4 +26,4 @@ Reference Errors:
 Hint: 'If validation fails because of extra whitespace in the TSV, try:
 
   src/cleanup_whitespace.py --tsv_in original.tsv --tsv_out clean.tsv.'
-```
+```
diff --git a/src/ingest_validation_tools/upload.py b/src/ingest_validation_tools/upload.py
@@ -567,26 +567,44 @@ def _check_matching_urls(
         for i, row in enumerate(rows):
             check = {k: v for k, v in row.items() if k in constrained_fields}
             for field, value in check.items():
-                try:
-                    url = constrained_fields[field] + value
-                    if field != "orcid_id":
-                        headers = self.app_context.get("request_header", {})
-                        headers["Authorization"] = f"Bearer {self.globus_token}"
-                    else:
-                        headers = {}
-                    response = requests.get(url, headers=headers)
-                    response.raise_for_status()
-                except Exception as e:
-                    error = {
-                        "errorType": type(e).__name__,
-                        "column": field,
-                        "row": i + 2,
-                        "value": value,
-                        "error_text": e.__str__(),
-                    }
-                    url_errors.append(self._get_message(error, report_type))
+                if field == "parent_sample_id":
+                    ids = value.split(",")
+                    for id in ids:
+                        error = self._check_single_url(field, id.strip(), constrained_fields, i)
+                        if error:
+                            url_errors.append(self._get_message(error, report_type))
+                else:
+                    error = self._check_single_url(field, value, constrained_fields, i)
+                    if error:
+                        url_errors.append(self._get_message(error, report_type))
         return url_errors
 
+    def _check_single_url(
+        self,
+        field: str,
+        value: str,
+        constrained_fields: Dict[str, str],
+        row_num: int,
+    ) -> Optional[Dict]:
+        try:
+            url = constrained_fields[field] + value
+            if field != "orcid_id":
+                headers = self.app_context.get("request_header", {})
+                headers["Authorization"] = f"Bearer {self.globus_token}"
+            else:
+                headers = {}
+            response = requests.get(url, headers=headers)
+            response.raise_for_status()
+        except Exception as e:
+            error = {
+                "errorType": type(e).__name__,
+                "column": field,
+                "row": row_num + 2,
+                "value": value,
+                "error_text": e.__str__(),
+            }
+            return error
+
     def _get_message(
         self,
         error: Dict[str, str],