rdblue commented on a change in pull request #432: Allow writers to control
size of files generated
URL: https://github.com/apache/incubator-iceberg/pull/432#discussion_r326406321
##########
File path: spark/src/main/java/org/apache/iceberg/spark/source/Writer.java
##########
@@ -303,96 +305,179 @@ public String toString() {
}
}
}
+
+ private class EncryptedOutputFileFactory implements
OutputFileFactory<EncryptedOutputFile> {
+ private final int partitionId;
+ private final long taskId;
+ private final long epochId;
+ // The purpose of this uuid is to be able to know from two paths that
they were written by the same operation.
+ // That's useful, for example, if a Spark job dies and leaves files in
the file system, you can identify them all
+ // with a recursive listing and grep.
+ private final String uuid = UUID.randomUUID().toString();
+ private int fileCount;
+
+ EncryptedOutputFileFactory(int partitionId, long taskId, long epochId) {
+ this.partitionId = partitionId;
+ this.taskId = taskId;
+ this.epochId = epochId;
+ this.fileCount = 0;
+ }
+
+ private synchronized String generateFilename() {
+ return format.addExtension(String.format("%05d-%d-%s-%05d",
partitionId, taskId, uuid, fileCount++));
+ }
+
+ /**
+ * Generates EncryptedOutputFile for UnpartitionedWriter.
+ */
+ public EncryptedOutputFile newOutputFile() {
+ OutputFile file =
fileIo.newOutputFile(locations.newDataLocation(generateFilename()));
+ return encryptionManager.encrypt(file);
+ }
+
+ /**
+ * Generates EncryptedOutputFile for PartitionedWriter.
+ */
+ public EncryptedOutputFile newOutputFile(PartitionKey key) {
+ OutputFile rawOutputFile =
fileIo.newOutputFile(locations.newDataLocation(spec, key, generateFilename()));
+ return encryptionManager.encrypt(rawOutputFile);
+ }
+ }
}
private interface AppenderFactory<T> {
FileAppender<T> newAppender(OutputFile file, FileFormat format);
}
- private static class UnpartitionedWriter implements DataWriter<InternalRow>,
Closeable {
- private final FileIO fileIo;
- private FileAppender<InternalRow> appender = null;
- private Metrics metrics = null;
- private List<Long> offsetRanges = null;
- private final EncryptedOutputFile file;
+ private interface OutputFileFactory<T> {
+ T newOutputFile();
+ T newOutputFile(PartitionKey key);
+ }
- UnpartitionedWriter(
- EncryptedOutputFile outputFile,
- FileFormat format,
- AppenderFactory<InternalRow> factory,
- FileIO fileIo) {
+ @SuppressWarnings("checkstyle:VisibilityModifier") // direct access desired
from sub-classes for performance.
+ private abstract static class BaseWriter implements DataWriter<InternalRow> {
+ protected static final int ROWS_DIVISOR = 1000;
+
+ protected final Set<PartitionKey> completedPartitions = Sets.newHashSet();
+ protected final List<DataFile> completedFiles = Lists.newArrayList();
+ protected final PartitionSpec spec;
+ protected final FileFormat format;
+ protected final AppenderFactory<InternalRow> appenderFactory;
+ protected final OutputFileFactory<EncryptedOutputFile> fileFactory;
+ protected final PartitionKey key;
+ protected final FileIO fileIo;
+ protected final long targetFileSize;
+ protected PartitionKey currentKey = null;
+ protected FileAppender<InternalRow> currentAppender = null;
+ protected EncryptedOutputFile currentFile = null;
+ protected long currentRows;
Review comment:
It is also strange that this is only used in child classes.
Can this class provide a `writeInternal` method that updates `currentRows`,
writes to the appender, and checks when to open and close the current appender?
That would be cleaner and would no longer require all the protected fields.
Does performance really degrade when the subclasses are a bit more separated
from this base class?
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
