HarshitGupta11 commented on code in PR #7197:
URL: https://github.com/apache/hadoop/pull/7197#discussion_r1925856394
##########
hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/shell/BulkDeleteCommand.java:
##########
@@ -34,161 +34,174 @@
public class BulkDeleteCommand extends FsCommand {
- public static void registerCommands(CommandFactory factory) {
- factory.addClass(BulkDeleteCommand.class, "-bulkDelete");
+ public static void registerCommands(CommandFactory factory) {
+ factory.addClass(BulkDeleteCommand.class, "-bulkDelete");
+ }
+
+ private static final Logger LOG =
LoggerFactory.getLogger(BulkDeleteCommand.class.getName());
+
+ public static final String NAME = "bulkDelete";
+
+ /**
+ * File Name parameter to be specified at command line.
+ */
+ public static final String READ_FROM_FILE = "readFromFile";
+
+ /**
+ * Page size parameter specified at command line.
+ */
+ public static final String PAGE_SIZE = "pageSize";
+
+
+ public static final String USAGE = "-[ " + READ_FROM_FILE + "] [<file>] [" +
PAGE_SIZE
+ + "] [<pageSize>] [<basePath> <paths>]";
+
+ public static final String DESCRIPTION = "Deletes the set of files under the
given <path>.\n" +
+ "If a list of paths is provided at command line then the paths are
deleted directly.\n" +
+ "User can also point to the file where the paths are listed as full
object names using the \"fileName\"" +
+ "parameter. The presence of a file name takes precedence over the
list of objects.\n" +
+ "Page size refers to the size of each bulk delete batch." +
+ "Users can specify the page size using \"pageSize\" command
parameter." +
+ "Default value is 1.\n";
+
+ private String fileName;
+
+ private int pageSize;
+
+ /**
+ * Making the class stateful as the PathData initialization for all args is
not needed
+ */
+ LinkedList<String> childArgs;
+
+ protected BulkDeleteCommand() {
+ this.childArgs = new LinkedList<>();
+ }
+
+ protected BulkDeleteCommand(Configuration conf) {
+ super(conf);
+ this.childArgs = new LinkedList<>();
+ this.pageSize = 1;
+ }
+
+ /**
+ * Processes the command line options and initialize the variables.
+ *
+ * @param args the command line arguments
+ * @throws IOException in case of wrong arguments passed
+ */
+ @Override
+ protected void processOptions(LinkedList<String> args) throws IOException {
+ CommandFormat cf = new CommandFormat(0, Integer.MAX_VALUE);
+ cf.addOptionWithValue(READ_FROM_FILE);
+ cf.addOptionWithValue(PAGE_SIZE);
+ cf.parse(args);
+ fileName = cf.getOptValue(READ_FROM_FILE);
+ if (cf.getOptValue(PAGE_SIZE) != null) {
+ pageSize = Integer.parseInt(cf.getOptValue(PAGE_SIZE));
+ } else {
+ pageSize = 1;
}
-
- private static final Logger LOG =
LoggerFactory.getLogger(BulkDeleteCommand.class.getName());
-
- public static final String name = "bulkDelete";
-
- /**
- * File Name parameter to be specified at command line
- */
- public static final String READ_FROM_FILE = "readFromFile";
-
- /**
- * Page size parameter specified at command line
- */
- public static final String PAGE_SIZE = "pageSize";
-
-
- public static final String USAGE = "-[ " + READ_FROM_FILE + "] [<file>] ["
+
- PAGE_SIZE + "] [<pageSize>] [<basePath> <paths>]";
-
- public static final String DESCRIPTION = "Deletes the set of files under
the given <path>.\n" +
- "If a list of paths is provided at command line then the paths are
deleted directly.\n" +
- "User can also point to the file where the paths are listed as
full object names using the \"fileName\"" +
- "parameter. The presence of a file name takes precedence over the
list of objects.\n" +
- "Page size refers to the size of each bulk delete batch." +
- "Users can specify the page size using \"pageSize\" command
parameter." +
- "Default value is 1.\n";
-
- private String fileName;
-
- private int pageSize;
-
- /*
- Making the class stateful as the PathData initialization for all args is
not needed
- */
- LinkedList<String> childArgs;
-
- protected BulkDeleteCommand() {
- this.childArgs = new LinkedList<>();
+ }
+
+ /**
+ * Processes the command line arguments and stores the child arguments in a
list.
+ *
+ * @param args strings to expand into {@link PathData} objects
+ * @return the base path of the bulk delete command.
+ * @throws IOException if the wrong number of arguments specified
+ */
+ @Override
+ protected LinkedList<PathData> expandArguments(LinkedList<String> args)
throws IOException {
+ if (fileName == null && args.size() < 2) {
+ throw new IOException("Invalid Number of Arguments. Expected more");
}
-
- protected BulkDeleteCommand(Configuration conf) {super(conf);}
-
- /**
- * Processes the command line options and initialize the variables
- * @param args the command line arguments
- * @throws IOException in case of wrong arguments passed
- */
- @Override
- protected void processOptions(LinkedList<String> args) throws IOException {
- CommandFormat cf = new CommandFormat(0, Integer.MAX_VALUE);
- cf.addOptionWithValue(READ_FROM_FILE);
- cf.addOptionWithValue(PAGE_SIZE);
- cf.parse(args);
- fileName = cf.getOptValue(READ_FROM_FILE);
- if(cf.getOptValue(PAGE_SIZE) != null) {
- pageSize = Integer.parseInt(cf.getOptValue(PAGE_SIZE));
- } else {
- pageSize = 1;
- }
+ LinkedList<PathData> pathData = new LinkedList<>();
+ pathData.add(new PathData(args.get(0), getConf()));
+ args.remove(0);
+ this.childArgs = args;
+ return pathData;
+ }
+
+ /**
+ * Deletes the objects using the bulk delete api.
+ *
+ * @param bulkDelete Bulkdelete object exposing the API
+ * @param paths list of paths to be deleted in the base path
+ * @throws IOException on error in execution of the delete command
+ */
+ void deleteInBatches(BulkDelete bulkDelete, List<Path> paths) throws
IOException {
+ Batch<Path> batches = new Batch<>(paths, pageSize);
+ while (batches.hasNext()) {
+ try {
+ List<Map.Entry<Path, String>> result =
bulkDelete.bulkDelete(batches.next());
+ LOG.debug("Deleted Result:{}", result.toString());
+ } catch (IllegalArgumentException e) {
+ LOG.error("Caught exception while deleting", e);
+ throw new IOException(e);
Review Comment:
done
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
