Add workaround if date in 008 is too long #2114

hbz · Jan 6, 2025 · 8562998 · 8562998
1 parent 9f1528b
commit 8562998
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 3 deletions.
diff --git a/src/main/resources/alma/fix/describedBy.fix b/src/main/resources/alma/fix/describedBy.fix
@@ -31,10 +31,16 @@ add_field("describedBy.resultOf.instrument.label","Software lobid-resources")
 copy_field("almaMmsId","describedBy.resultOf.object.id")
 prepend("describedBy.resultOf.object.id","https://lobid.org/marcxml/")
 
-# MNG is a ALMA-specific element (MNG  .b only states the indexing date into ALMA, while 008 is the initial cataloguing date.)
+# 008/00-05 has the initial cataloguing date. We use MNG info as fallback.
+# MNG is a ALMA-specific element (MNG  .b only states the indexing date into ALMA.)
 
 copy_field("008","@initialCataloguingDate")
-substring("@initialCataloguingDate","0","6")
+if any_match("@initialCataloguingDate", "^\\d{8}.*") # Due to cataloguing errors there seem to be year not represented by two but by four digits. The first step is a workaround.
+  substring("@initialCataloguingDate","2","6")
+elsif any_match("@initialCataloguingDate", "^\\d{6}.*") # 008/00-05 is the correct form for the cataloguing date in MARC.
+  substring("@initialCataloguingDate","0","6")
+end
+
 if any_match("@initialCataloguingDate","^[0-4].*")
   prepend("@initialCataloguingDate","20")
 elsif any_match("@initialCataloguingDate","\\d*")

diff --git a/src/test/resources/alma-fix/99374515437806441.json b/src/test/resources/alma-fix/99374515437806441.json
@@ -36,7 +36,7 @@
       },
       "object" : {
         "id" : "https://lobid.org/marcxml/99374515437806441",
-        "dateCreated" : "2020-24-08",
+        "dateCreated" : "2024-08-16",
         "dateModified" : "2024-10-30",
         "type" : [ "DataFeedItem" ],
         "label" : "hbz-Ressource 99374515437806441 im Exportformat MARC21 XML",