[HIVEMALL-145] Merge Brickhouse functions

core/src/main/java/hivemall/tools/sanity/AssertUDF.java

@@ -25,8 +25,10 @@
 
 @Description(name = "assert",
         value = "_FUNC_(boolean condition) or _FUNC_(boolean condition, string errMsg)"
-                + "- Throws HiveException if condition is not met")
-@UDFType(deterministic = true, stateful = false)
+                + "- Throws HiveException if condition is not met",
+        extended = "SELECT count(1) FROM stock_price WHERE assert(price > 0.0);\n"
+                + "SELECT count(1) FROM stock_price WHRE assert(price > 0.0, 'price MUST be more than 0.0')")
+@UDFType(deterministic = false, stateful = false)
 public final class AssertUDF extends UDF {
 
     public boolean evaluate(boolean condition) throws HiveException {

core/src/main/java/hivemall/tools/sanity/RaiseErrorUDF.java

@@ -18,21 +18,48 @@
  */
 package hivemall.tools.sanity;
 
+import org.apache.commons.lang.StringUtils;
 import org.apache.hadoop.hive.ql.exec.Description;
-import org.apache.hadoop.hive.ql.exec.UDF;
+import org.apache.hadoop.hive.ql.exec.UDFArgumentException;
+import org.apache.hadoop.hive.ql.exec.UDFArgumentLengthException;
 import org.apache.hadoop.hive.ql.metadata.HiveException;
 import org.apache.hadoop.hive.ql.udf.UDFType;
+import org.apache.hadoop.hive.ql.udf.generic.GenericUDF;
+import org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector;
+import org.apache.hadoop.hive.serde2.objectinspector.primitive.PrimitiveObjectInspectorFactory;
 
-@Description(name = "raise_error", value = "_FUNC_() or _FUNC_(string msg) - Throws an error")
-@UDFType(deterministic = true, stateful = false)
-public final class RaiseErrorUDF extends UDF {
+@Description(name = "raise_error", value = "_FUNC_() or _FUNC_(string msg) - Throws an error",
+        extended = "SELECT product_id, price, raise_error('Found an invalid record') FROM xxx WHERE price < 0.0")
+@UDFType(deterministic = false, stateful = false)
+public class RaiseErrorUDF extends GenericUDF {
 
-    public boolean evaluate() throws HiveException {
-        throw new HiveException();
+    @Override
+    public ObjectInspector initialize(ObjectInspector[] argOIs) throws UDFArgumentException {
+        if (argOIs.length != 0 && argOIs.length != 1) {
+            throw new UDFArgumentLengthException(
+                "Expected one or two arguments for raise_error UDF: " + argOIs.length);
+        }
+
+        return PrimitiveObjectInspectorFactory.writableBooleanObjectInspector;
+    }
+
+    @Override
+    public Object evaluate(DeferredObject[] arguments) throws HiveException {
+        if (arguments.length == 1) {
+            Object arg0 = arguments[0].get();
+            if (arg0 == null) {
+                throw new HiveException();
+            }
+            String msg = arg0.toString();
+            throw new HiveException(msg);
+        } else {
+            throw new HiveException();
+        }
     }
 
-    public boolean evaluate(String errorMessage) throws HiveException {
-        throw new HiveException(errorMessage);
+    @Override
+    public String getDisplayString(String[] children) {
+        return "raise_error(" + StringUtils.join(children, ',') + ')';
     }
 
 }

resources/ddl/define-all-as-permanent.hive

@@ -812,3 +812,9 @@ CREATE FUNCTION to_json as 'hivemall.tools.json.ToJsonUDF' USING JAR '${hivemall
 
 DROP FUNCTION IF EXISTS from_json;
 CREATE FUNCTION from_json as 'hivemall.tools.json.FromJsonUDF' USING JAR '${hivemall_jar}';
+
+DROP FUNCTION IF EXISTS assert;
+CREATE FUNCTION assert as 'hivemall.tools.sanity.AssertUDF' USING JAR '${hivemall_jar}';
+
+DROP FUNCTION IF EXISTS raise_error;
+CREATE FUNCTION raise_error as 'hivemall.tools.sanity.RaiseErrorUDF' USING JAR '${hivemall_jar}';

resources/ddl/define-all.hive

@@ -818,3 +818,9 @@ create temporary function to_json as 'hivemall.tools.json.ToJsonUDF';
 
 drop temporary function if exists from_json;
 create temporary function from_json as 'hivemall.tools.json.FromJsonUDF';
+
+drop temporary function if exists assert;
+create temporary function assert as 'hivemall.tools.sanity.AssertUDF';
+
+drop temporary function if exists raise_error;
+create temporary function raise_error as 'hivemall.tools.sanity.RaiseErrorUDF';

resources/ddl/define-all.spark

@@ -771,3 +771,9 @@ sqlContext.sql("CREATE TEMPORARY FUNCTION to_json AS 'hivemall.tools.json.ToJson
 
 sqlContext.sql("DROP TEMPORARY FUNCTION IF EXISTS from_json")
 sqlContext.sql("CREATE TEMPORARY FUNCTION from_json AS 'hivemall.tools.json.FromJsonUDF'")
+
+sqlContext.sql("DROP TEMPORARY FUNCTION IF EXISTS assert")
+sqlContext.sql("CREATE TEMPORARY FUNCTION assert AS 'hivemall.tools.sanity.AssertUDF'")
+
+sqlContext.sql("DROP TEMPORARY FUNCTION IF EXISTS raise_error")
+sqlContext.sql("CREATE TEMPORARY FUNCTION raise_error AS 'hivemall.tools.sanity.RaiseErrorUDF'")

resources/ddl/define-udfs.td.hql

@@ -230,3 +230,7 @@ create temporary function sessionize as 'hivemall.tools.datetime.SessionizeUDF';
 create temporary function to_json as 'hivemall.tools.json.ToJsonUDF';
 
 create temporary function from_json as 'hivemall.tools.json.FromJsonUDF';
+
+create temporary function assert as 'hivemall.tools.sanity.AssertUDF';
+
+create temporary function raise_error as 'hivemall.tools.sanity.RaiseErrorUDF';