Reduce errors in ICU4J Collator executor results (#185)

* Fix parsing for Collator for ICU4J executor

* Refactor ICU4J executor run configs

* Add tests for the ICU4J executor for collator

* Update ICU4J executor code for Collator to directly call the superclass method

* Ensure that ICU4J runs Collator with normalization turned on

* Remove unit test based on upstream test data gen bug re: double escaping

executors/icu4j/73/executor-icu4j/src/main/java/org/unicode/conformance/testtype/collator/CollatorTester.java

@@ -39,27 +39,8 @@ public ITestTypeInputJson inputMapToJson(Map<String, Object> inputMapData) {
 
     result.locale = (String) inputMapData.get("locale", null);
 
-    boolean ignorePunctuation = false;
-    Optional<Object> ignorePunctuationStr = inputMapData.get("ignorePunctuation");
-    try {
-      if (ignorePunctuationStr.isPresent()) {
-        ignorePunctuation = Boolean.parseBoolean((String) ignorePunctuationStr.get());
-      }
-    } catch (Exception e) {
-      // do nothing, default is false
-    }
-    result.ignorePunctuation = ignorePunctuation;
-
-    int line = 0;
-    Optional<Object> lineStr = inputMapData.get("line");
-    try {
-      if (lineStr.isPresent()) {
-        line = Integer.parseInt((String) lineStr.get());
-      }
-    } catch (Exception e) {
-      // do nothing, default is 0
-    }
-    result.line = line;
+    result.ignorePunctuation = (boolean) inputMapData.get("ignorePunctuation", false);
+    result.line = (int) ((double) inputMapData.get("line", 0.0));
 
     result.compare_type = (String) inputMapData.get("compare_type", null);
     result.test_description = (String) inputMapData.get("test_description", null);
@@ -158,7 +139,7 @@ public Collator getCollatorForInput(CollatorInputJson input) {
 
     if (input.locale == null) {
       if (input.rules == null) {
-        result = (RuleBasedCollator) RuleBasedCollator.getInstance();
+        result = (RuleBasedCollator) Collator.getInstance(ULocale.ROOT);
       } else {
         try {
           result = new RuleBasedCollator(input.rules);
@@ -168,7 +149,7 @@ public Collator getCollatorForInput(CollatorInputJson input) {
       }
     } else {
       ULocale locale = ULocale.forLanguageTag(input.locale);
-      result = (RuleBasedCollator) RuleBasedCollator.getInstance(locale);
+      result = (RuleBasedCollator) Collator.getInstance(locale);
       if (input.rules != null) {
         String defaultRules = result.getRules();
         String newRules = defaultRules + input.rules;
@@ -180,6 +161,10 @@ public Collator getCollatorForInput(CollatorInputJson input) {
       }
     }
 
+    // ensure that ICU performs decomposition before collation in order to get proper results,
+    // per documentation: https://unicode-org.github.io/icu-docs/apidoc/dev/icu4j/com/ibm/icu/text/Collator.html
+    result.setDecomposition(Collator.CANONICAL_DECOMPOSITION);
+
     if (input.ignorePunctuation) {
       result.setAlternateHandlingShifted(true);
     }

executors/icu4j/73/executor-icu4j/src/test/java/org/unicode/conformance/collator/CollatorTest.java

@@ -0,0 +1,33 @@
+package org.unicode.conformance.collator;
+
+import static org.junit.Assert.assertTrue;
+
+import org.junit.Test;
+import org.unicode.conformance.testtype.collator.CollatorOutputJson;
+import org.unicode.conformance.testtype.collator.CollatorTester;
+
+public class CollatorTest {
+
+  @Test
+  public void testIgnorePunctuation() {
+    String testInput =
+        "{\"label\":\"0001424\",\"s1\":\" ?\",\"s2\":\"　?\",\"line\":59,\"ignorePunctuation\":true}";
+
+    CollatorOutputJson output =
+        (CollatorOutputJson) CollatorTester.INSTANCE.getStructuredOutputFromInputStr(testInput);
+
+    assertTrue(output.result);
+  }
+
+  @Test
+  public void testNonEscaped() {
+    String testInput =
+        "\t{\"label\":\"0006747\",\"s1\":\"̴?\",\"s2\":\"̴̲\",\"line\":5382,\"ignorePunctuation\":true}";
+
+    CollatorOutputJson output =
+        (CollatorOutputJson) CollatorTester.INSTANCE.getStructuredOutputFromInputStr(testInput);
+
+    assertTrue(output.result);
+  }
+
+}

run_config.json

@@ -247,21 +247,7 @@
       "icu_version": "icu73",
       "exec": "icu4j",
       "test_type": [
-        "collation_short"
-      ],
-      "per_execution": 10000
-    }
-  },
-  {
-    "prereq": {
-      "name": "mvn-icu4j-73-shaded",
-      "version": "73",
-      "command": "mvn -f ../executors/icu4j/73/executor-icu4j/pom.xml package"
-    },
-    "run": {
-      "icu_version": "icu73",
-      "exec": "icu4j",
-      "test_type": [
+        "collation_short",
         "lang_names",
         "likely_subtags",
         "number_fmt"