[GitHub] [iceberg] huaxingao commented on a diff in pull request #5302: Add SparkV2Filters

[GitBox <gi...@apache.org>]

huaxingao commented on code in PR #5302:
URL: https://github.com/apache/iceberg/pull/5302#discussion_r936030720


##########
spark/v3.3/spark/src/main/java/org/apache/iceberg/spark/SparkV2Filters.java:
##########
@@ -0,0 +1,375 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *   http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing,
+ * software distributed under the License is distributed on an
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+ * KIND, either express or implied.  See the License for the
+ * specific language governing permissions and limitations
+ * under the License.
+ */
+package org.apache.iceberg.spark;
+
+import static org.apache.iceberg.expressions.Expressions.and;
+import static org.apache.iceberg.expressions.Expressions.equal;
+import static org.apache.iceberg.expressions.Expressions.greaterThan;
+import static org.apache.iceberg.expressions.Expressions.greaterThanOrEqual;
+import static org.apache.iceberg.expressions.Expressions.in;
+import static org.apache.iceberg.expressions.Expressions.isNaN;
+import static org.apache.iceberg.expressions.Expressions.isNull;
+import static org.apache.iceberg.expressions.Expressions.lessThan;
+import static org.apache.iceberg.expressions.Expressions.lessThanOrEqual;
+import static org.apache.iceberg.expressions.Expressions.not;
+import static org.apache.iceberg.expressions.Expressions.notIn;
+import static org.apache.iceberg.expressions.Expressions.notNull;
+import static org.apache.iceberg.expressions.Expressions.or;
+import static org.apache.iceberg.expressions.Expressions.startsWith;
+
+import java.util.Map;
+import java.util.Objects;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+import java.util.stream.Collectors;
+import java.util.stream.Stream;
+import org.apache.iceberg.expressions.Expression;
+import org.apache.iceberg.expressions.Expressions;
+import org.apache.iceberg.relocated.com.google.common.base.Preconditions;
+import org.apache.iceberg.relocated.com.google.common.collect.ImmutableMap;
+import org.apache.iceberg.util.NaNUtil;
+import org.apache.spark.sql.connector.expressions.LiteralValue;
+import org.apache.spark.sql.connector.expressions.NamedReference;
+import org.apache.spark.sql.connector.expressions.filter.And;
+import org.apache.spark.sql.connector.expressions.filter.Not;
+import org.apache.spark.sql.connector.expressions.filter.Or;
+import org.apache.spark.sql.connector.expressions.filter.Predicate;
+import org.apache.spark.unsafe.types.UTF8String;
+
+public class SparkV2Filters {
+
+  private static final Pattern BACKTICKS_PATTERN = Pattern.compile("([`])(.|$)");
+
+  private SparkV2Filters() {}
+
+  private static final Map<String, Expression.Operation> FILTERS =

Review Comment:
   I think the main reason that Spark uses a String for the V2 Filter name is because Spark introduces [`GeneralScalarExpression`](https://github.com/apache/spark/blob/master/sql/catalyst/src/main/java/org/apache/spark/sql/connector/expressions/GeneralScalarExpression.java) in 3.3 and pushes down a lot of functions such as `+`, `-`, `*`, `/`, `ABS`, `LOG`, ... Spark doesn't want to have a class for each of these Expressions so it uses a more genetic way. V2 [`Predicate`](https://github.com/apache/spark/blob/master/sql/catalyst/src/main/java/org/apache/spark/sql/connector/expressions/filter/Predicate.java) is a subclass of `GeneralScalarExpression` so it also uses the field `name` for Filter name. 
   
   I am actually wondering if any of these function push down can benefit Iceberg. It seems to me that for both the data predicate push down and partition pruning, Iceberg only supports predicate `attri cmp lit`, so we can't really take advantage of any of these new function push down in Spark. In Spark built-in file source such as Parquet, partition pruning uses Catalyst Expression so Spark can prune partition such as `partCol1 + partCol2 > 3`. 
   
   



##########
spark/v3.3/spark/src/main/java/org/apache/iceberg/spark/SparkV2Filters.java:
##########
@@ -0,0 +1,375 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *   http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing,
+ * software distributed under the License is distributed on an
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+ * KIND, either express or implied.  See the License for the
+ * specific language governing permissions and limitations
+ * under the License.
+ */
+package org.apache.iceberg.spark;
+
+import static org.apache.iceberg.expressions.Expressions.and;
+import static org.apache.iceberg.expressions.Expressions.equal;
+import static org.apache.iceberg.expressions.Expressions.greaterThan;
+import static org.apache.iceberg.expressions.Expressions.greaterThanOrEqual;
+import static org.apache.iceberg.expressions.Expressions.in;
+import static org.apache.iceberg.expressions.Expressions.isNaN;
+import static org.apache.iceberg.expressions.Expressions.isNull;
+import static org.apache.iceberg.expressions.Expressions.lessThan;
+import static org.apache.iceberg.expressions.Expressions.lessThanOrEqual;
+import static org.apache.iceberg.expressions.Expressions.not;
+import static org.apache.iceberg.expressions.Expressions.notIn;
+import static org.apache.iceberg.expressions.Expressions.notNull;
+import static org.apache.iceberg.expressions.Expressions.or;
+import static org.apache.iceberg.expressions.Expressions.startsWith;
+
+import java.util.Map;
+import java.util.Objects;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+import java.util.stream.Collectors;
+import java.util.stream.Stream;
+import org.apache.iceberg.expressions.Expression;
+import org.apache.iceberg.expressions.Expressions;
+import org.apache.iceberg.relocated.com.google.common.base.Preconditions;
+import org.apache.iceberg.relocated.com.google.common.collect.ImmutableMap;
+import org.apache.iceberg.util.NaNUtil;
+import org.apache.spark.sql.connector.expressions.LiteralValue;
+import org.apache.spark.sql.connector.expressions.NamedReference;
+import org.apache.spark.sql.connector.expressions.filter.And;
+import org.apache.spark.sql.connector.expressions.filter.Not;
+import org.apache.spark.sql.connector.expressions.filter.Or;
+import org.apache.spark.sql.connector.expressions.filter.Predicate;
+import org.apache.spark.unsafe.types.UTF8String;
+
+public class SparkV2Filters {
+
+  private static final Pattern BACKTICKS_PATTERN = Pattern.compile("([`])(.|$)");
+
+  private SparkV2Filters() {}
+
+  private static final Map<String, Expression.Operation> FILTERS =
+      ImmutableMap.<String, Expression.Operation>builder()
+          .put("ALWAYS_TRUE", Expression.Operation.TRUE)
+          .put("ALWAYS_FALSE", Expression.Operation.FALSE)
+          .put("=", Expression.Operation.EQ)
+          .put("<=>", Expression.Operation.EQ)
+          .put(">", Expression.Operation.GT)
+          .put(">=", Expression.Operation.GT_EQ)
+          .put("<", Expression.Operation.LT)
+          .put("<=", Expression.Operation.LT_EQ)
+          .put("IN", Expression.Operation.IN)
+          .put("IS_NULL", Expression.Operation.IS_NULL)
+          .put("IS_NOT_NULL", Expression.Operation.NOT_NULL)
+          .put("AND", Expression.Operation.AND)
+          .put("OR", Expression.Operation.OR)
+          .put("NOT", Expression.Operation.NOT)
+          .put("STARTS_WITH", Expression.Operation.STARTS_WITH)
+          .build();
+
+  public static Expression convert(Predicate[] predicates) {
+    Expression expression = Expressions.alwaysTrue();
+    for (Predicate predicate : predicates) {
+      Expression converted = convert(predicate);
+      Preconditions.checkArgument(
+          converted != null, "Cannot convert predicate to Iceberg: %s", predicate);
+      expression = Expressions.and(expression, converted);
+    }
+    return expression;
+  }
+
+  @SuppressWarnings({"checkstyle:CyclomaticComplexity", "checkstyle:MethodLength"})
+  public static Expression convert(Predicate predicate) {
+    if (checkIfPredicateValid(predicate) == null) {
+      return null;

Review Comment:
   The idea is that if the Spark V2 filter is not valid, then we simply return a `null` and skip the convert code, then later on we don't need to bind and push down this predicate (https://github.com/apache/iceberg/blob/master/spark/v3.3/spark/src/main/java/org/apache/iceberg/spark/source/SparkScanBuilder.java#L120)
   



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org