From 338d6a6c4554a9f7afcb2f572943f276f7912995 Mon Sep 17 00:00:00 2001
From: Martin Czygan <martin.czygan@gmail.com>
Date: Thu, 1 Apr 2021 00:57:17 +0200
Subject: find doi in url

---
 skate/README.md                           |  3 +++
 skate/cmd/skate-from-unstructured/main.go | 19 ++++++++++++-------
 skate/verify.go                           |  6 +++---
 3 files changed, 18 insertions(+), 10 deletions(-)

(limited to 'skate')

diff --git a/skate/README.md b/skate/README.md
index 1962dc6..cc7e238 100644
--- a/skate/README.md
+++ b/skate/README.md
@@ -56,6 +56,9 @@ After this step:
 * cluster, e.g. `skate-cluster ...`
 
 ### skate-from-unstructured
+
+
+
 ### skate-ref-to-release
 ### skate-to-doi
 ### skate-verify
diff --git a/skate/cmd/skate-from-unstructured/main.go b/skate/cmd/skate-from-unstructured/main.go
index 0208d91..8ebc613 100644
--- a/skate/cmd/skate-from-unstructured/main.go
+++ b/skate/cmd/skate-from-unstructured/main.go
@@ -33,16 +33,10 @@ func main() {
 		if err := json.Unmarshal(p, &ref); err != nil {
 			return nil, err
 		}
-		// TODO: ref
 		if err := parseUnstructured(&ref); err != nil {
 			return nil, err
 		}
-		b, err := json.Marshal(ref)
-		if err != nil {
-			return nil, err
-		}
-		b = append(b, bytesNewline...)
-		return b, nil
+		return skate.JsonMarshalLine(&ref)
 	})
 	pp.NumWorkers = *numWorkers
 	pp.BatchSize = *batchSize
@@ -74,6 +68,17 @@ func parseUnstructured(ref *skate.Ref) error {
 	if v != "" && ref.Biblio.DOI == "" {
 		ref.Biblio.DOI = v
 	}
+	// DOI in URL
+	prefixes := []string{"http://doi.org/", "https://doi.org/", "http://dx.doi.org/", "https://dx.doi.org/"}
+	for _, prefix := range prefixes {
+		if ref.Biblio.DOI != "" && strings.HasPrefix(ref.Biblio.Url, prefix) {
+			ref.Biblio.DOI = strings.Replace(ref.Biblio.Url, prefix, "", -1)
+		}
+	}
+	v = PatDOINoHyphen.FindString(ref.Key)
+	if v != "" && ref.Biblio.DOI == "" {
+		ref.Biblio.DOI = v
+	}
 	// Arxiv
 	vs = PatArxivPDF.FindStringSubmatch(uns)
 	if len(vs) != 0 && ref.Biblio.ArxivId == "" {
diff --git a/skate/verify.go b/skate/verify.go
index cfe31ec..40ee4e2 100644
--- a/skate/verify.go
+++ b/skate/verify.go
@@ -123,8 +123,8 @@ var (
 	PatPages           = regexp.MustCompile(`([0-9]{1,})-([0-9]{1,})`)
 )
 
-// jsonMarshalLine marshals a value as JSON and adds a newline.
-func jsonMarshalLine(v interface{}) ([]byte, error) {
+// JsonMarshalLine marshals a value as JSON and adds a newline.
+func JsonMarshalLine(v interface{}) ([]byte, error) {
 	b, err := json.Marshal(v)
 	if err != nil {
 		return nil, err
@@ -230,7 +230,7 @@ func RefClusterToBiblioRef(p []byte) ([]byte, error) {
 				continue // Assume we already have the DOI matches.
 			}
 			br = generateBiblioRef(re, pivot, result, "fuzzy")
-			return jsonMarshalLine(br)
+			return JsonMarshalLine(br)
 		default:
 			continue
 		}
-- 
cgit v1.2.3