Metrics QA fixes #4527

Removed concat usage Added correct datasetsBySubject query Updated docs on default metrics timeout
IQSS · May 22, 2018 · a8fcd3e · pdurbin · May 22, 2018 · a8fcd3e
1 parent 4e1471b
commit a8fcd3e
Show file tree

Hide file tree

Showing 2 changed files with 27 additions and 19 deletions.
diff --git a/doc/sphinx-guides/source/installation/config.rst b/doc/sphinx-guides/source/installation/config.rst
@@ -1283,6 +1283,6 @@ Enable the collection of provenance metadata on Dataverse via the provenance pop
 :MetricsCacheTimeoutMinutes
 +++++++++++++++++++++++++++
 
-Sets how long a cached metrics result is used before re-running the query for a request. Note this only effects queries on the current month, previous months queries are cached indefinitely.
+Sets how long a cached metrics result is used before re-running the query for a request. Note this only effects queries on the current month, previous months queries are cached indefinitely. The default timeout is 7 days (10080 minutes).
 
 ``curl -X PUT -d 10080 http://localhost:8080/api/admin/settings/:MetricsCacheTimeoutMinutes``
diff --git a/src/main/java/edu/harvard/iq/dataverse/metrics/MetricsServiceBean.java b/src/main/java/edu/harvard/iq/dataverse/metrics/MetricsServiceBean.java
@@ -51,9 +51,9 @@ public long datasetsByMonth(String yyyymm) throws Exception {
         Query query = em.createNativeQuery(""
                 + "select count(*)\n"
                 + "from datasetversion\n"
-                + "where concat(datasetversion.dataset_id,':', datasetversion.versionnumber + (.1 * datasetversion.minorversionnumber)) in \n"
+                + "where datasetversion.dataset_id || ':' || datasetversion.versionnumber + (.1 * datasetversion.minorversionnumber) in \n"
                 + "(\n"
-                + "select concat(datasetversion.dataset_id,':', max(datasetversion.versionnumber + (.1 * datasetversion.minorversionnumber))) as max \n"
+                + "select datasetversion.dataset_id || ':' || max(datasetversion.versionnumber + (.1 * datasetversion.minorversionnumber)) as max \n"
                 + "from datasetversion\n"
                 + "join dataset on dataset.id = datasetversion.dataset_id\n"
                 + "where versionstate='RELEASED' \n"
@@ -75,9 +75,9 @@ public long filesByMonth(String yyyymm) throws Exception {
                 + "select count(*)\n"
                 + "from filemetadata\n"
                 + "join datasetversion on datasetversion.id = filemetadata.datasetversion_id\n"
-                + "where concat(datasetversion.dataset_id,':', datasetversion.versionnumber + (.1 * datasetversion.minorversionnumber)) in \n"
+                + "where datasetversion.dataset_id || ':' || datasetversion.versionnumber + (.1 * datasetversion.minorversionnumber) in \n"
                 + "(\n"
-                + "select concat(datasetversion.dataset_id,':', max(datasetversion.versionnumber + (.1 * datasetversion.minorversionnumber))) as max \n"
+                + "select datasetversion.dataset_id || ':' || max(datasetversion.versionnumber + (.1 * datasetversion.minorversionnumber)) as max \n"
                 + "from datasetversion\n"
                 + "join dataset on dataset.id = datasetversion.dataset_id\n"
                 + "where versionstate='RELEASED'\n"
@@ -120,20 +120,28 @@ public List<Object[]> dataversesByCategory() throws Exception {
 
     public List<Object[]> datasetsBySubject() {
         Query query = em.createNativeQuery(""
-                + "SELECT strvalue, count(dataset.id)\n"
-                + "FROM datasetfield_controlledvocabularyvalue \n"
-                + "JOIN controlledvocabularyvalue ON controlledvocabularyvalue.id = datasetfield_controlledvocabularyvalue.controlledvocabularyvalues_id\n"
-                + "JOIN datasetfield ON datasetfield.id = datasetfield_controlledvocabularyvalue.datasetfield_id\n"
-                + "JOIN datasetfieldtype ON datasetfieldtype.id = controlledvocabularyvalue.datasetfieldtype_id\n"
-                + "JOIN datasetversion ON datasetversion.id = datasetfield.datasetversion_id\n"
-                + "JOIN dvobject ON dvobject.id = datasetversion.dataset_id\n"
-                + "JOIN dataset ON dataset.id = datasetversion.dataset_id\n"
-                + "WHERE\n"
-                + "datasetfieldtype.name = 'subject'\n"
-                + "AND dvobject.publicationdate is NOT NULL\n"
-                + "AND dataset.harvestingclient_id IS NULL\n"
-                + "GROUP BY strvalue\n"
-                + "ORDER BY count(dataset.id) desc;"
+               + "SELECT strvalue, count(dataset.id)\n"
+               + "FROM datasetfield_controlledvocabularyvalue \n"
+               + "JOIN controlledvocabularyvalue ON controlledvocabularyvalue.id = datasetfield_controlledvocabularyvalue.controlledvocabularyvalues_id\n"
+               + "JOIN datasetfield ON datasetfield.id = datasetfield_controlledvocabularyvalue.datasetfield_id\n"
+               + "JOIN datasetfieldtype ON datasetfieldtype.id = controlledvocabularyvalue.datasetfieldtype_id\n"
+               + "JOIN datasetversion ON datasetversion.id = datasetfield.datasetversion_id\n"
+               + "JOIN dvobject ON dvobject.id = datasetversion.dataset_id\n"
+               + "JOIN dataset ON dataset.id = datasetversion.dataset_id\n"
+               + "WHERE\n"
+               + "datasetversion.dataset_id || ':' || datasetversion.versionnumber + (.1 * datasetversion.minorversionnumber) in \n"
+               + "(\n"
+               + "select datasetversion.dataset_id || ':' || max(datasetversion.versionnumber + (.1 * datasetversion.minorversionnumber)) as max \n"
+               + "from datasetversion\n"
+               + "join dataset on dataset.id = datasetversion.dataset_id\n"
+               + "where versionstate='RELEASED'\n"
+               + "and date_trunc('month', releasetime) <=  to_date('2018-05','YYYY-MM')\n"
+               + "and dataset.harvestingclient_id is null\n"
+               + "group by dataset_id \n"
+               + ")\n"
+               + "AND datasetfieldtype.name = 'subject'\n"
+               + "GROUP BY strvalue\n"
+               + "ORDER BY count(dataset.id) desc;"
         );
         logger.info("query: " + query);