/**
* DataCleaner (community edition)
* Copyright (C) 2014 Neopost - Customer Information Management
*
* This copyrighted material is made available to anyone wishing to use, modify,
* copy, or redistribute it subject to the terms and conditions of the GNU
* Lesser General Public License, as published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
* or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License
* for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this distribution; if not, write to:
* Free Software Foundation, Inc.
* 51 Franklin Street, Fifth Floor
* Boston, MA 02110-1301 USA
*/
package org.datacleaner.spark.utils;
import java.io.IOException;
import java.net.URI;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.metamodel.util.FileHelper;
import org.apache.metamodel.util.FileResource;
import org.apache.metamodel.util.HdfsResource;
import org.apache.metamodel.util.Resource;
import org.apache.spark.api.java.JavaSparkContext;
import org.apache.spark.deploy.SparkHadoopUtil;
import org.datacleaner.util.HadoopResource;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import com.google.common.base.Strings;
/**
* Helper class for interacting with HDFS.
*/
public class HdfsHelper {
private static final Logger logger = LoggerFactory.getLogger(HdfsHelper.class);
private static Configuration _lastKnownConfiguration;
private final Configuration _hadoopConfiguration;
public HdfsHelper(final JavaSparkContext sparkContext) {
this(getHadoopConfigurationIfYarnMode(sparkContext));
}
public HdfsHelper(final Configuration configuration) {
if (configuration == null) {
logger.warn("Hadoop Configuration is null!", new Throwable());
} else {
_lastKnownConfiguration = configuration;
}
_hadoopConfiguration = configuration;
}
/**
* Creates a {@link HdfsHelper} without any configuration or context
* available. This is normally not the recommended way to obtain a
* {@link HdfsHelper} but may be necessary in executor functions where the
* {@link JavaSparkContext} is not in scope and not made available by Spark
* (at least spark's end-user API).
*
* @return
*/
public static HdfsHelper createHelper() {
Configuration configuration = _lastKnownConfiguration;
if (configuration == null) {
try {
final SparkHadoopUtil sparkHadoopUtil = SparkHadoopUtil.get();
if (sparkHadoopUtil.isYarnMode()) {
configuration = sparkHadoopUtil.conf();
}
} catch (final Exception e) {
// the above is developer API so we don't consider it very
// stable.
}
}
return new HdfsHelper(configuration);
}
private static Configuration getHadoopConfigurationIfYarnMode(final JavaSparkContext sparkContext) {
final String sparkMaster = sparkContext.getConf().get("spark.master");
if (Strings.isNullOrEmpty(sparkMaster) || "local".equals(sparkMaster)) {
return null;
}
return sparkContext.hadoopConfiguration();
}
/**
* Clears up the statically cached reference to a {@link Configuration} object, which is used an
* HdfsHelper is instantiated without an explicit {@link Configuration} object.
*
* Note: Only use if you want to start fresh and make sure no lingering objects are used.
*/
public static void clear() {
_lastKnownConfiguration = null;
}
public String readFile(final URI filepath) {
return readFile(filepath, false);
}
public String readFile(final URI filepath, final boolean failOnNoData) {
final Resource resourceInUse = getResourceToUse(filepath);
if (failOnNoData && resourceInUse == null) {
throw new IllegalArgumentException("Could not resolve resource: " + filepath);
}
return readResource(resourceInUse);
}
public String readResource(final Resource resource) {
final Resource resourceInUse = getResourceToUse(resource);
if (resourceInUse == null) {
return null;
}
return resourceInUse.read(in -> {
return FileHelper.readInputStreamAsString(in, FileHelper.DEFAULT_ENCODING);
});
}
public Resource getResourceToUse(final Resource resource) {
if (resource == null) {
return null;
}
if (_hadoopConfiguration == null || resource instanceof HadoopResource) {
return resource;
}
if (resource instanceof HdfsResource) {
// wrap the resource with our known configuration
return new HadoopResource(resource, _hadoopConfiguration, HadoopResource.DEFAULT_CLUSTERREFERENCE);
}
if (resource instanceof FileResource) {
// this may very well be a path that was mis-interpreted as a local
// file because no scheme was defined
if (resource.getQualifiedPath().startsWith("/")) {
return new HadoopResource(resource, _hadoopConfiguration, HadoopResource.DEFAULT_CLUSTERREFERENCE);
}
}
return resource;
}
public Resource getResourceToUse(final URI path) {
if (path == null) {
return null;
}
if (_hadoopConfiguration == null) {
if ("hdfs".equals(path.getScheme())) {
return new HdfsResource(path.toString());
}
return new FileResource(path.toString());
}
return new HadoopResource(path, _hadoopConfiguration, HadoopResource.DEFAULT_CLUSTERREFERENCE);
}
public boolean isDirectory(final URI path) {
final Resource resource = getResourceToUse(path);
if (!resource.isExists()) {
return false;
}
if (resource instanceof FileResource) {
return ((FileResource) resource).getFile().isDirectory();
}
if (resource instanceof HdfsResource) {
final FileSystem fileSystem = ((HdfsResource) resource).getHadoopFileSystem();
final Path hadoopPath = ((HdfsResource) resource).getHadoopPath();
try {
return fileSystem.isDirectory(hadoopPath);
} catch (final IOException e) {
throw new IllegalStateException(e);
}
}
// actually we don't know, but most likely it's not a directory
return false;
}
}