Updates based on further discussion in PR #56

* Update datacite transform to fix errors and add validation to publicationYear * Update dspace_dim transform for code consistency
MITLibraries · Mar 14, 2023 · 885bd58 · 885bd58
1 parent 30a81c6
commit 885bd58
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 7 deletions.
diff --git a/transmogrifier/sources/datacite.py b/transmogrifier/sources/datacite.py
@@ -107,11 +107,14 @@ def get_optional_fields(self, xml: Tag) -> Optional[dict]:
 
         # dates
         if publication_year := xml.metadata.find("publicationYear", string=True):
-            fields["dates"] = [
-                timdex.Date(
-                    kind="Publication date", value=publication_year.string.strip()
-                )
-            ]
+            publication_year = str(publication_year.string.strip())
+            if validate_date(
+                publication_year,
+                source_record_id,
+            ):
+                fields["dates"] = [
+                    timdex.Date(kind="Publication date", value=publication_year)
+                ]
         else:
             logger.warning(
                 "Datacite record %s missing required Datacite field publicationYear",
@@ -120,7 +123,8 @@ def get_optional_fields(self, xml: Tag) -> Optional[dict]:
 
         for date in xml.metadata.find_all("date"):
             d = timdex.Date()
-            if date_value := str(date.string):
+            if date_value := date.string:
+                date_value = str(date_value)
                 if "/" in date_value:
                     split = date_value.index("/")
                     gte_date = date_value[:split].strip()

diff --git a/transmogrifier/sources/dspace_dim.py b/transmogrifier/sources/dspace_dim.py
@@ -87,7 +87,7 @@ def get_optional_fields(self, xml: Tag) -> Optional[dict]:
             )
 
         # dates
-        for date in [d for d in xml.find_all("dim:field", element="date") if d.string]:
+        for date in xml.find_all("dim:field", element="date", string=True):
             date_value = str(date.string.strip())
             if validate_date(date_value, source_record_id):
                 if date.get("qualifier") == "issued":