0.1.3

Features

BigQuery

• Resumable uploads via write channel are now supported (#540)

  An example of uploading a CSV file in chunks of CHUNK_SIZE bytes:

  try (FileChannel fileChannel = FileChannel.open(Paths.get("/path/to/your/file"))) {
    ByteBuffer buffer = ByteBuffer.allocate(256 * 1024);
    TableId tableId = TableId.of("YourDataset", "YourTable");
    LoadConfiguration configuration =
        LoadConfiguration.of(tableId, FormatOptions.of("CSV"));
    WriteChannel writeChannel = bigquery.writer(configuration);
    long position = 0;
    long written = fileChannel.transferTo(position, CHUNK_SIZE, writeChannel);
    while (written > 0) {
      position += written;
      written = fileChannel.transferTo(position, CHUNK_SIZE, writeChannel);
    }
    writeChannel.close();
  }

• defaultDataset(String dataset) (in QueryJobInfo and QueryRequest) can be used to specify a default dataset (#567).

Storage

• The name of the method to submit a batch request has changed from apply to submit (#562).

Fixes

BigQuery

• hashCode and equals are now overridden in subclasses of BaseTableInfo (#565, #573).
• jobComplete is renamed to jobCompleted in QueryResult (#567).

Datastore

• The precondition check that cursors are UTF-8 strings has been removed (#578).

• EntityQuery, KeyQuery, and ProjectionEntityQuery classes have been introduced (#585). This enables users to use modify projections and group by clauses for projection entity queries after using toBuilder(). For example, this now works:

  ProjectionEntityQuery query = Query.projectionEntityQueryBuilder()
      .kind("Person")
      .projection(Projection.property("name"))
      .build();
  ProjectionEntityQuery newQuery =
      query.toBuilder().projection(Projection.property("favorite_food")).build();