From b2caf2a36c9b77e421ceeca257af53ea0ed5ed9f Mon Sep 17 00:00:00 2001
From: Martin Czygan <martin.czygan@gmail.com>
Date: Thu, 7 Nov 2024 10:18:57 +0100
Subject: [PATCH] crossref: #21429

---
 formats/crossref/document_test.go | 7 +++++++
 formats/finc/solr.go              | 6 +++++-
 2 files changed, 12 insertions(+), 1 deletion(-)

diff --git a/formats/crossref/document_test.go b/formats/crossref/document_test.go
index b0d397bc..9d098a18 100644
--- a/formats/crossref/document_test.go
+++ b/formats/crossref/document_test.go
@@ -35,6 +35,13 @@ func TestDocumentCombinedTitle(t *testing.T) {
 			},
 			result: "Hello Sub",
 		},
+		{
+			doc: &Document{
+				Title:    []string{"Deutsche Nationalkataloge – Herausforderungen an das deutsche Bibliothekssystem"},
+				Subtitle: []string{"Was aus der Perspektive der Digital Humanities zu tun wäre"},
+			},
+			result: "Deutsche Nationalkataloge – Herausforderungen an das deutsche Bibliothekssystem : Was aus der Perspektive der Digital Humanities zu tun wäre",
+		},
 	}
 	for _, c := range cases {
 		result := c.doc.CombinedTitle()
diff --git a/formats/finc/solr.go b/formats/finc/solr.go
index fde7e296..8dbc81f8 100644
--- a/formats/finc/solr.go
+++ b/formats/finc/solr.go
@@ -140,7 +140,11 @@ func (s *Solr5Vufind3) convert(is IntermediateSchema, withFullrecord bool) error
 	}
 
 	s.SourceID = is.SourceID
-	s.Subtitle = is.ArticleSubtitle
+	if len(is.ArticleSubtitle) > 20 && !strings.Contains(s.Title, is.ArticleSubtitle) {
+		// refs #21429, avoid duplications, e.g. crossref assembles a title
+		// from both title and subtitle already
+		s.Subtitle = is.ArticleSubtitle
+	}
 	s.TitleSort = is.SortableTitle()
 	s.Topics = is.Subjects