[GitHub] [hudi] bvaradar commented on a change in pull request #3808: [HUDI-2560] introduce id_based schema to support full schema evolution.

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

bvaradar commented on a change in pull request #3808:
URL: https://github.com/apache/hudi/pull/3808#discussion_r808603391



##########
File path: hudi-common/src/main/java/org/apache/hudi/internal/schema/io/FileBasedInternalSchemaStorageManager.java
##########
@@ -0,0 +1,169 @@
+/*
+ * 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.hudi.internal.schema.io;
+
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.fs.FSDataInputStream;
+import org.apache.hadoop.fs.FileSystem;
+import org.apache.hadoop.fs.Path;
+import org.apache.hudi.common.table.HoodieTableMetaClient;
+import org.apache.hudi.common.table.timeline.HoodieActiveTimeline;
+import org.apache.hudi.common.table.timeline.HoodieInstant;
+import org.apache.hudi.common.util.FileIOUtils;
+import org.apache.hudi.common.util.Option;
+import org.apache.hudi.exception.HoodieException;
+import org.apache.hudi.exception.HoodieIOException;
+import org.apache.hudi.internal.schema.InternalSchema;
+import org.apache.hudi.internal.schema.utils.SerDeHelper;
+
+import org.apache.log4j.LogManager;
+import org.apache.log4j.Logger;
+
+import java.io.IOException;
+import java.nio.charset.StandardCharsets;
+import java.util.Arrays;
+import java.util.List;
+import java.util.TreeMap;
+import java.util.stream.Collectors;
+
+import static org.apache.hudi.common.table.timeline.HoodieTimeline.SAVE_SCHEMA_ACTION;
+
+public class FileBasedInternalSchemaStorageManager extends AbstractInternalSchemaStorageManager {
+  private static final Logger LOG = LogManager.getLogger(FileBasedInternalSchemaStorageManager.class);
+
+  public static final String SCHEMA_NAME = ".schema";
+  private final Path baseSchemaPath;
+  private Configuration conf;
+  private HoodieTableMetaClient metaClient;
+
+  public FileBasedInternalSchemaStorageManager(Configuration conf, Path baseTablePath) {
+    Path metaPath = new Path(baseTablePath, ".hoodie");
+    this.baseSchemaPath = new Path(metaPath, SCHEMA_NAME);
+    this.conf = conf;
+    this.metaClient = HoodieTableMetaClient.builder().setBasePath(metaPath.getParent().toString()).setConf(conf).build();
+  }
+
+  public FileBasedInternalSchemaStorageManager(HoodieTableMetaClient metaClient) {
+    Path metaPath = new Path(metaClient.getBasePath(), ".hoodie");
+    this.baseSchemaPath = new Path(metaPath, SCHEMA_NAME);
+    this.conf = metaClient.getHadoopConf();
+    this.metaClient = metaClient;
+  }
+
+  @Override
+  public void persistHistorySchemaStr(String instantTime, String historySchemaStr) {
+    cleanResidualFiles();
+    HoodieActiveTimeline timeline = metaClient.getActiveTimeline();
+    HoodieInstant hoodieInstant = new HoodieInstant(HoodieInstant.State.REQUESTED, SAVE_SCHEMA_ACTION, instantTime);
+    timeline.createNewInstant(hoodieInstant);
+    byte[] writeContent = historySchemaStr.getBytes(StandardCharsets.UTF_8);
+    timeline.transitionRequestedToInflight(hoodieInstant, Option.empty());
+    timeline.saveAsComplete(new HoodieInstant(HoodieInstant.State.INFLIGHT, hoodieInstant.getAction(), hoodieInstant.getTimestamp()), Option.of(writeContent));
+    LOG.info(String.format("persist history schema success on commit time: %s", instantTime));
+  }
+
+  private void cleanResidualFiles() {
+    List<String> validateCommits = getValidInstants();
+    try {
+      FileSystem fs = baseSchemaPath.getFileSystem(conf);
+      if (fs.exists(baseSchemaPath)) {
+        List<String> candidateSchemaFiles = Arrays.stream(fs.listStatus(baseSchemaPath)).filter(f -> f.isFile())
+            .map(file -> file.getPath().getName()).collect(Collectors.toList());
+        List<String> residualSchemaFiles = candidateSchemaFiles.stream().filter(f -> !validateCommits.contains(f.split("\\.")[0])).collect(Collectors.toList());
+        // clean residual files
+        residualSchemaFiles.forEach(f -> {
+          try {
+            fs.delete(new Path(metaClient.getSchemaFolderName(), f));
+          } catch (IOException o) {
+            throw new HoodieException(o);
+          }
+        });
+      }
+    } catch (IOException e) {
+      throw new HoodieException(e);
+    }
+  }
+
+  public void cleanOldFiles(List<String> validateCommits) {
+    try {
+      FileSystem fs = baseSchemaPath.getFileSystem(conf);
+      if (fs.exists(baseSchemaPath)) {
+        List<String> candidateSchemaFiles = Arrays.stream(fs.listStatus(baseSchemaPath)).filter(f -> f.isFile())
+            .map(file -> file.getPath().getName()).collect(Collectors.toList());
+        List<String> validateSchemaFiles = candidateSchemaFiles.stream().filter(f -> validateCommits.contains(f.split("\\.")[0])).collect(Collectors.toList());
+        for (int i = 0; i < validateSchemaFiles.size(); i++) {
+          fs.delete(new Path(validateSchemaFiles.get(i)));
+        }
+      }
+    } catch (IOException e) {
+      throw new HoodieException(e);
+    }
+  }
+
+  private List<String> getValidInstants() {
+    metaClient.reloadActiveTimeline();
+    return metaClient.getCommitsTimeline()
+        .filterCompletedInstants().getInstants().map(f -> f.getTimestamp()).collect(Collectors.toList());
+  }
+
+  @Override
+  public String getHistorySchemaStr() {

Review comment:
       We need a proper class to manage in-memory representation of Schema history instead of keeping it as string and passing everywhere as string.

##########
File path: hudi-common/src/main/java/org/apache/hudi/internal/schema/action/SchemaMerger.java
##########
@@ -0,0 +1,133 @@
+/*
+ * 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.hudi.internal.schema.action;
+
+import org.apache.hudi.common.util.StringUtils;
+import org.apache.hudi.internal.schema.InternalSchema;
+import org.apache.hudi.internal.schema.Type;
+import org.apache.hudi.internal.schema.Types;
+
+import java.util.ArrayList;
+import java.util.List;
+
+/**
+ * auxiliary class.
+ * help to merge file schema and query schema to produce final read schema for avro/parquet file
+ */
+public class SchemaMerger {
+  private final InternalSchema fileSchema;
+  private final InternalSchema querySchema;
+  // now there exist some bugs when we use spark update/merge api,
+  // those operation will change col nullability from optional to required which is wrong.
+  // Before that bug is fixed, we need to do adapt.
+  // if mergeRequiredFiledForce is true, we will ignore the col's required attribute.
+  private final Boolean mergeRequiredFiledForce;
+  public SchemaMerger(InternalSchema fileSchema, InternalSchema querySchema, Boolean mergeRequiredFiledForce) {

Review comment:
       nit: rename mergeRequiredFiledForce to ignoreRequiredAttribute

##########
File path: hudi-common/src/main/java/org/apache/hudi/internal/schema/action/SchemaMerger.java
##########
@@ -0,0 +1,133 @@
+/*
+ * 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.hudi.internal.schema.action;
+
+import org.apache.hudi.common.util.StringUtils;
+import org.apache.hudi.internal.schema.InternalSchema;
+import org.apache.hudi.internal.schema.Type;
+import org.apache.hudi.internal.schema.Types;
+
+import java.util.ArrayList;
+import java.util.List;
+
+/**
+ * auxiliary class.
+ * help to merge file schema and query schema to produce final read schema for avro/parquet file
+ */
+public class SchemaMerger {

Review comment:
       Rename to InternalSchemaMerger




-- 
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: commits-unsubscribe@hudi.apache.org

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