From 25a4517383c03540bd3d7695e14f60c1bb20da2d Mon Sep 17 00:00:00 2001
From: Martin Czygan <martin.czygan@gmail.com>
Date: Tue, 1 Jun 2021 17:20:22 +0200
Subject: add mapper: MapperReleaseContainerName

---
 skate/cmd/skate-map/main.go | 19 ++++++++++---------
 skate/map.go                | 27 ++++++++++++++++++++++-----
 2 files changed, 32 insertions(+), 14 deletions(-)

(limited to 'skate')

diff --git a/skate/cmd/skate-map/main.go b/skate/cmd/skate-map/main.go
index 572ecec..060e324 100644
--- a/skate/cmd/skate-map/main.go
+++ b/skate/cmd/skate-map/main.go
@@ -67,15 +67,16 @@ func main() {
 	availableMappers := map[string]skate.Mapper{
 		// Add new mapper functions here. TODO: add more docs, and improve
 		// composability, e.g. like middleware.
-		"id":  skate.Identity,
-		"ff":  skate.CreateFixedMapper(*extraValue),
-		"ti":  skate.MapperTitle,
-		"tn":  skate.MapperTitleNormalized,
-		"ty":  skate.MapperTitleNysiis,
-		"ts":  skate.MapperTitleSandcrawler,
-		"ru":  skate.MapperURLFromRef,
-		"cni": skate.MapperContainerName,
-		"cns": skate.MapperContainerNameSandcrawler,
+		"id":   skate.Identity,
+		"ff":   skate.CreateFixedMapper(*extraValue),
+		"ti":   skate.MapperTitle,
+		"tn":   skate.MapperTitleNormalized,
+		"ty":   skate.MapperTitleNysiis,
+		"ts":   skate.MapperTitleSandcrawler,
+		"ru":   skate.MapperURLFromRef,
+		"cni":  skate.MapperContainerName,
+		"cns":  skate.MapperContainerNameSandcrawler,
+		"rcns": skate.MapperReleaseContainerName,
 	}
 	if *logFile != "" {
 		f, err := os.OpenFile(*logFile, os.O_CREATE|os.O_APPEND, 0644)
diff --git a/skate/map.go b/skate/map.go
index f812464..a06b5ae 100644
--- a/skate/map.go
+++ b/skate/map.go
@@ -24,16 +24,16 @@ type TitleDoc struct {
 	Title string `json:"title"`
 }
 
-// ContainerNameDoc is a document with a container title.
+// ContainerNameDoc is a (ref) document with a container title.
 type ContainerNameDoc struct {
 	Biblio struct {
 		ContainerName string `json:"container_name"`
 	} `json:"biblio"`
 }
 
-// PartialDoc for ref docs, that do not have DOI or title. E.g. we found 49701699
-// (NCVY), 36401044 (NCVYU), 29668363 (NCUY), and so on. Some examples: XXX
-type PartialDoc struct {
+// PartialRef for ref docs, that do not have DOI or title. E.g. we found 49701699
+// (NCVY), 36401044 (NCVYU), 29668363 (NCUY), and so on. Some examples: XXX.
+type PartialRef struct {
 	ContainerName string `json:"container_name"`
 	Contribs      []struct {
 		// XXX: Need a way to sensibly compare sets of author names.
@@ -200,7 +200,7 @@ func MapperTitleSandcrawler(p []byte) (fields [][]byte, err error) {
 // MapperContainerName extracts (container_name, doc).
 func MapperContainerName(p []byte) ([][]byte, error) {
 	var (
-		doc PartialDoc
+		doc PartialRef
 		key []byte
 	)
 	if err := json.Unmarshal(p, &doc); err != nil {
@@ -240,6 +240,23 @@ func MapperURLFromRef(p []byte) (fields [][]byte, err error) {
 	return fields, nil
 }
 
+// MapperReleaseContainerName extracts a normalized container name.
+func MapperReleaseContainerName(p []byte) (fields [][]byte, err error) {
+	var (
+		doc Release
+		key []byte
+	)
+	if err := json.Unmarshal(p, &doc); err != nil {
+		return nil, err
+	}
+	if doc.Container.Name != "" {
+		key = []byte(sandcrawlerSlugify(wsReplacer.Replace(strings.TrimSpace(doc.Container.Name))))
+	} else if doc.ContainerName != "" {
+		key = []byte(sandcrawlerSlugify(wsReplacer.Replace(strings.TrimSpace(doc.ContainerName))))
+	}
+	return [][]byte{key, p}, nil
+}
+
 // MapperPartial works on partial documents.
 func MapperPartial(p []byte) (fields [][]byte, err error) {
 	return nil, nil
-- 
cgit v1.2.3