Allow original casing of clinical data values (but group them case in…

…sensitive)

src/main/java/org/cbioportal/service/util/StudyViewColumnarServiceUtil.java

@@ -27,11 +27,14 @@ public static List<ClinicalDataCountItem> mergeClinicalDataCounts(
     ) {
         items.forEach(attr -> {
             Map<String, List<ClinicalDataCount>> countsPerType = attr.getCounts().stream()
-                .collect(Collectors.groupingBy(ClinicalDataCount::getValue));
+                .collect(Collectors.groupingBy(a -> a.getValue().toLowerCase()));
             List<ClinicalDataCount> res = countsPerType.entrySet().stream().map(entry -> {
                 ClinicalDataCount mergedCount = new ClinicalDataCount();
                 mergedCount.setAttributeId(attr.getAttributeId());
-                mergedCount.setValue(entry.getKey());
+                // we are just going to choose the value of the first item
+                // due to failure in data normalization in source files, we may find values
+                // have inconsistent casing.  we choose to merge and choose an arbitrary casing
+                mergedCount.setValue(entry.getValue().stream().findFirst().get().getValue());
                 mergedCount.setCount(entry.getValue().stream().mapToInt(ClinicalDataCount::getCount).sum());
                 return mergedCount;
             }).toList();

src/main/resources/org/cbioportal/persistence/mybatisclickhouse/StudyViewMapper.xml

@@ -190,7 +190,7 @@
         WITH clinical_data_query AS (
         SELECT
         attribute_name AS attributeId,
-        upper(attribute_value) AS value,
+        attribute_value AS value,
         cast(count(*) AS INTEGER) as count
         FROM clinical_data_derived
         <where>

src/test/java/org/cbioportal/service/util/StudyViewColumnarServiceUtilTest.java

@@ -31,9 +31,11 @@ public void testMergeClinicalDataCounts() {
         count1.setValue("value1");
         count1.setCount(1);
 
+        // not that the caseing of the value is different
+        // our merge code should ignore that and choose the first item
         ClinicalDataCount count2 = new ClinicalDataCount();
         count2.setAttributeId("attr1");
-        count2.setValue("value1");
+        count2.setValue("VALUE1");
         count2.setCount(2);
 
         ClinicalDataCount count3 = new ClinicalDataCount();