Merge pull request #1302 from libris/feature/softer-work-matching

Feature/softer work matching
libris · Sep 8, 2023 · e7aa105 · e7aa105
2 parents ee2e597 + 12639b7
commit e7aa105
Show file tree

Hide file tree

Showing 9 changed files with 65 additions and 48 deletions.
diff --git a/librisworks/scripts/display-works.groovy b/librisworks/scripts/display-works.groovy
@@ -2,7 +2,7 @@ import se.kb.libris.mergeworks.Doc
 import se.kb.libris.mergeworks.Html
 import se.kb.libris.mergeworks.WorkComparator
 
-import static se.kb.libris.mergeworks.Util.partition
+import static se.kb.libris.mergeworks.Util.workClusters
 
 htmlReport = getReportWriter('works.html')
 
@@ -28,15 +28,6 @@ new File(System.getProperty('clusters')).splitEachLine(~/[\t ]+/) { cluster ->
 
 htmlReport.println(Html.END)
 
-Collection<Collection<Doc>> workClusters(Collection<Doc> docs, WorkComparator c) {
-    docs.each { it.addComparisonProps() }
-
-    def workClusters = partition(docs, { Doc a, Doc b -> c.sameWork(a, b) })
-            .each { work -> work.each { doc -> doc.removeComparisonProps() } }
-
-    return workClusters
-}
-
 Doc createNewWork(Map workData) {
     workData['@id'] = "TEMPID#it"
     Map data = [

diff --git a/librisworks/scripts/merge-works.groovy b/librisworks/scripts/merge-works.groovy
@@ -2,7 +2,7 @@ import se.kb.libris.mergeworks.Html
 import se.kb.libris.mergeworks.WorkComparator
 import se.kb.libris.mergeworks.Doc
 
-import static se.kb.libris.mergeworks.Util.partition
+import static se.kb.libris.mergeworks.Util.workClusters
 
 maybeDuplicates = getReportWriter("maybe-duplicate-linked-works.tsv")
 multiWorkReport = getReportWriter("multi-work-clusters.html")
@@ -46,7 +46,7 @@ new File(System.getProperty('clusters')).splitEachLine(~/[\t ]+/) { cluster ->
             uniqueWorksAndTheirInstances.add(new Tuple2(linkedWorks.find(), localWorks))
         } else {
             maybeDuplicates.println(linkedWorks.collect { it.shortId() }.join('\t'))
-            System.err.println("Local works ${localWorks.collect { it.shortId() }} match multiple linked works: ${linkedWorks.collect { it.shortId() }}. Duplicate linked works?")
+            System.err.println("Local works ${localWorks.collect { it.shortId() }} match multiple linked works: ${linkedWorks.collect { it.shortId() }}. Duplicated linked works?")
         }
     }
 
@@ -102,19 +102,6 @@ void saveAndLink(Doc workDoc, Collection<Doc> instanceDocs = [], boolean existsI
     }
 }
 
-Collection<Collection<Doc>> workClusters(Collection<Doc> docs, WorkComparator c) {
-    docs.each {
-        if (it.instanceData) {
-            it.addComparisonProps()
-        }
-    }
-
-    def workClusters = partition(docs, { Doc a, Doc b -> c.sameWork(a, b) })
-            .each { work -> work.each { doc -> doc.removeComparisonProps() } }
-
-    return workClusters
-}
-
 Doc createNewWork(Map workData) {
     workData['@id'] = "TEMPID#it"
     Map data = [

diff --git a/librisworks/scripts/swedish-fiction.groovy b/librisworks/scripts/swedish-fiction.groovy
@@ -17,6 +17,7 @@ new File(System.getProperty('clusters')).splitEachLine(~/[\t ]+/) {cluster ->
                 && !doc.isDrama()
                 && !doc.isThesis()
                 && !doc.isInSb17Bibliography()
+                && !doc.intendedForMarcPreAdolescent()
     }
 
     def swedish = { Doc doc ->

diff --git a/librisworks/src/main/groovy/se/kb/libris/mergeworks/Doc.groovy b/librisworks/src/main/groovy/se/kb/libris/mergeworks/Doc.groovy
@@ -144,6 +144,10 @@ class Doc {
         asList(workData['genreForm'])
     }
 
+    List<Map> intendedAudience() {
+        asList(workData['intendedAudience'])
+    }
+
     List<Map> publication() {
         asList(instanceData?.publication)
     }
@@ -211,6 +215,10 @@ class Doc {
                 || hasRelationshipWithContribution()
     }
 
+    boolean intendedForMarcPreAdolescent() {
+        intendedAudience().contains(['@id': 'https://id.kb.se/marc/PreAdolescent'])
+    }
+
     boolean hasPart() {
         workData['hasPart'] || instanceData['hasTitle'].findAll { it['@type'] == 'Title' }.any {
             it.hasPart?.size() > 1 || it.hasPart?.any { p -> asList(p.partName).size() > 1 || asList(p.partNumber).size() > 1 }

diff --git a/librisworks/src/main/groovy/se/kb/libris/mergeworks/Util.groovy b/librisworks/src/main/groovy/se/kb/libris/mergeworks/Util.groovy
@@ -5,6 +5,8 @@ import whelk.Whelk
 import whelk.util.DocumentUtil
 import whelk.util.Unicode
 
+import static se.kb.libris.mergeworks.compare.IntendedAudience.preferredComparisonOrder
+
 class Util {
     static def titleComponents = ['mainTitle', 'titleRemainder', 'subtitle', 'hasPart', 'partNumber', 'partName', 'marc:parallelTitle', 'marc:equalTitle']
 
@@ -298,4 +300,17 @@ class Util {
                 ? normalize("${agent.givenName} ${agent.familyName}")
                 : agent.name ? normalize("${agent.name}") : null
     }
+
+    static Collection<Collection<Doc>> workClusters(Collection<Doc> docs, WorkComparator c) {
+        docs.each {
+            if (it.instanceData) {
+                it.addComparisonProps()
+            }
+        }.with { preferredComparisonOrder(it) }
+
+        def workClusters = partition(docs, { Doc a, Doc b -> c.sameWork(a, b) })
+                .each { work -> work.each { doc -> doc.removeComparisonProps() } }
+
+        return workClusters
+    }
 }
diff --git a/librisworks/src/main/groovy/se/kb/libris/mergeworks/WorkComparator.groovy b/librisworks/src/main/groovy/se/kb/libris/mergeworks/WorkComparator.groovy
@@ -11,10 +11,10 @@ class WorkComparator {
 
     Map<String, FieldHandler> comparators = [
             'classification'  : new Classification(),
-            'contentType'     : new SameOrEmpty('https://id.kb.se/term/rda/Text'),
+            'contentType'     : new ContentType(),
             'genreForm'       : new GenreForm(),
             'hasTitle'        : new WorkTitle(),
-            'intendedAudience': new SameOrEmpty('https://id.kb.se/marc/Juvenile'),
+            'intendedAudience': new IntendedAudience(),
             '_numPages'       : new Extent(),
             'subject'         : new Subject(),
             'summary'         : new StuffSet(),

diff --git a/librisworks/src/main/groovy/se/kb/libris/mergeworks/compare/ContentType.groovy b/librisworks/src/main/groovy/se/kb/libris/mergeworks/compare/ContentType.groovy
@@ -0,0 +1,12 @@
+package se.kb.libris.mergeworks.compare
+
+import static se.kb.libris.mergeworks.Util.asList
+
+class ContentType extends StuffSet {
+    private static def allowedValues = ['https://id.kb.se/term/rda/StillImage', 'https://id.kb.se/term/rda/Text']
+
+    @Override
+    boolean isCompatible(Object a, Object b) {
+        asList(a).every { it['@id'] in allowedValues } && asList(b).every { it['@id'] in allowedValues }
+    }
+}
diff --git a/librisworks/src/main/groovy/se/kb/libris/mergeworks/compare/IntendedAudience.groovy b/librisworks/src/main/groovy/se/kb/libris/mergeworks/compare/IntendedAudience.groovy
@@ -0,0 +1,24 @@
+package se.kb.libris.mergeworks.compare
+
+import se.kb.libris.mergeworks.Doc
+
+import static se.kb.libris.mergeworks.Util.asList
+
+class IntendedAudience extends StuffSet {
+    private static def GENERAL = ['@id': 'https://id.kb.se/marc/General']
+    private static def ADULT = ['@id': 'https://id.kb.se/marc/Adult']
+
+    @Override
+    boolean isCompatible(Object a, Object b) {
+        !a || !b || asList(a) == [GENERAL] || asList(b) == [GENERAL]
+                || !(asList(a) + asList(b)).findResults { it == ADULT }.containsAll([true, false])
+    }
+
+    static void preferredComparisonOrder(Collection<Doc> docs) {
+        docs.sort { Doc d ->
+            d.intendedAudience().with {
+                it.isEmpty() || it == [GENERAL] || it == [ADULT]
+            }
+        }.reverse(true)
+    }
+}
diff --git a/librisworks/src/main/groovy/se/kb/libris/mergeworks/compare/SameOrEmpty.groovy b/librisworks/src/main/groovy/se/kb/libris/mergeworks/compare/SameOrEmpty.groovy