Remove Hive integration tests for CDH 5

.github/workflows/ci.yml

@@ -176,7 +176,6 @@ jobs:
         config:
           - config-empty
           - config-hdp3
-          # TODO: config-cdh5
           # TODO: config-apache-hive3
     timeout-minutes: 60
     steps:

docs/src/main/sphinx/connector/hive.rst

@@ -40,8 +40,7 @@ The Hive connector requires a Hive metastore service (HMS), or a compatible
 implementation of the Hive metastore, such as
 `AWS Glue Data Catalog <https://aws.amazon.com/glue/>`_.
 
-Apache Hadoop 2.x and 3.x are supported, along with derivative distributions,
-including Cloudera CDH 5 and Hortonworks Data Platform (HDP).
+Apache Hadoop HDFS 2.x and 3.x are supported.
 
 Many distributed storage systems including HDFS,
 :doc:`Amazon S3 <hive-s3>` or S3-compatible systems,

lib/trino-parquet/src/test/java/io/trino/parquet/writer/TestParquetWriter.java

@@ -17,8 +17,6 @@
 import org.testng.annotations.Test;
 
 import java.io.IOException;
-import java.util.regex.Matcher;
-import java.util.regex.Pattern;
 
 import static org.assertj.core.api.Assertions.assertThat;
 
@@ -31,21 +29,12 @@ public void testCreatedByIsParsable()
         String createdBy = ParquetWriter.formatCreatedBy("test-version");
         // createdBy must start with "parquet-mr" to make Apache Hive perform timezone conversion on INT96 timestamps correctly
         // when hive.parquet.timestamp.skip.conversion is set to true.
-        // Apache Hive 3.2 and above, and CDH 5 enable hive.parquet.timestamp.skip.conversion by default
+        // Apache Hive 3.2 and above enable hive.parquet.timestamp.skip.conversion by default
         assertThat(createdBy).startsWith("parquet-mr");
         VersionParser.ParsedVersion version = VersionParser.parse(createdBy);
         assertThat(version).isNotNull();
         assertThat(version.application).isEqualTo("parquet-mr-trino");
         assertThat(version.version).isEqualTo("test-version");
         assertThat(version.appBuildHash).isEqualTo("n/a");
-
-        // Ensure that createdBy field is parsable in CDH 5 to avoid the exception "parquet.io.ParquetDecodingException: Cannot read data due to PARQUET-246: to read safely, set parquet.split.files to false";
-        // the pattern is taken from https://github.com/cloudera/parquet-mr/blob/cdh5-1.5.0_5.15.1/parquet-common/src/main/java/parquet/VersionParser.java#L34
-        Pattern pattern = Pattern.compile("(.+) version ((.*) )?\\(build ?(.*)\\)");
-        Matcher matcher = pattern.matcher(createdBy);
-        assertThat(matcher.matches()).isTrue();
-        assertThat(matcher.group(1)).isEqualTo("parquet-mr-trino");
-        assertThat(matcher.group(3)).isEqualTo("test-version");
-        assertThat(matcher.group(4)).isEqualTo("n/a");
     }
 }

plugin/trino-hive/src/main/java/io/trino/plugin/hive/util/HiveUtil.java

@@ -358,7 +358,7 @@ public static Optional<CompressionCodec> getCompressionCodec(TextInputFormat inp
     private static Class<? extends InputFormat<?, ?>> getInputFormatClass(JobConf conf, String inputFormatName)
             throws ClassNotFoundException
     {
-        // CDH uses different names for Parquet
+        // legacy names for Parquet
         if ("parquet.hive.DeprecatedParquetInputFormat".equals(inputFormatName) ||
                 "parquet.hive.MapredParquetInputFormat".equals(inputFormatName)) {
             return MapredParquetInputFormat.class;
@@ -462,7 +462,7 @@ public static Deserializer getDeserializer(Configuration configuration, Properti
 
     private static Class<? extends Deserializer> getDeserializerClass(String name)
     {
-        // CDH uses different names for Parquet
+        // legacy name for Parquet
         if ("parquet.hive.serde.ParquetHiveSerDe".equals(name)) {
             return ParquetHiveSerDe.class;
         }