From dea840992ea80288760dacb0c9c80bc8b085978f Mon Sep 17 00:00:00 2001
From: Martin Czygan <martin.czygan@gmail.com>
Date: Mon, 21 Jun 2021 20:13:48 +0200
Subject: add fallback self-link removal

This should be handled at e.g. doi matching time.
---
 skate/zippy.go | 13 +++++++++++++
 1 file changed, 13 insertions(+)

(limited to 'skate')

diff --git a/skate/zippy.go b/skate/zippy.go
index 2317b56..a846a3b 100644
--- a/skate/zippy.go
+++ b/skate/zippy.go
@@ -354,6 +354,7 @@ func ZippyBrefAugment(bref, raw io.Reader, w io.Writer) error {
 			// exact match, and twice unmatched).
 			// TODO: remove duplicates
 			matched = deduplicateBrefs(matched)
+			matched = removeSelfLinks(matched)
 			for _, bref := range matched {
 				stats.total++
 				if err := enc.Encode(bref); err != nil {
@@ -369,6 +370,18 @@ func ZippyBrefAugment(bref, raw io.Reader, w io.Writer) error {
 	return err
 }
 
+// removeSelfLinks removes self-referential links. Those should be caught
+// earlier at the root cause later.
+func removeSelfLinks(brefs []*BiblioRef) (result []*BiblioRef) {
+	for _, bref := range brefs {
+		if bref.SourceReleaseIdent == bref.TargetReleaseIdent {
+			continue
+		}
+		result = append(result, bref)
+	}
+	return result
+}
+
 // deduplicateBrefs deduplicates by the document id (for elasticsearch), which
 // may help filter out some duplicates but not all.
 func deduplicateBrefs(brefs []*BiblioRef) []*BiblioRef {
-- 
cgit v1.2.3