protected void writeAndValidate(Schema schema) throws IOException { List<Record> expected = RandomGenericData.generate(schema, 100, 0L); File testFile = temp.newFile(); Assert.assertTrue("Delete should succeed", testFile.delete()); try (FileAppender<Record> writer = Avro.write(Files.localOutput(testFile)) .schema(schema) .createWriterFunc(DataWriter::create) .named("test") .build()) { for (Record rec : expected) { writer.add(rec); } } List<Record> rows; try (AvroIterable<Record> reader = Avro.read(Files.localInput(testFile)) .project(schema) .createReaderFunc(DataReader::create) .build()) { rows = Lists.newArrayList(reader); } for (int i = 0; i < expected.size(); i += 1) { DataTestHelpers.assertEquals(schema.asStruct(), expected.get(i), rows.get(i)); } } }
protected Record writeAndRead(String desc, Schema writeSchema, Schema readSchema, Record record) throws IOException { File file = temp.newFile(desc + ".avro"); file.delete(); try (FileAppender<Record> appender = Avro.write(Files.localOutput(file)) .schema(writeSchema) .createWriterFunc(DataWriter::create) .build()) { appender.add(record); } Iterable<Record> records = Avro.read(Files.localInput(file)) .project(readSchema) .createReaderFunc(DataReader::create) .build(); return Iterables.getOnlyElement(records); } }
private InputFile writeFile(String location, String filename, List<Record> records) throws IOException { Path path = new Path(location, filename); FileFormat format = FileFormat.fromFileName(filename); Preconditions.checkNotNull(format, "Cannot determine format for file: %s", filename); switch (format) { case AVRO: try (FileAppender<Record> appender = Avro.write(fromPath(path, CONF)) .schema(SCHEMA) .createWriterFunc(DataWriter::create) .named(format.name()) .build()) { appender.addAll(records); } return HadoopInputFile.fromPath(path, CONF); case PARQUET: try (FileAppender<Record> appender = Parquet.write(fromPath(path, CONF)) .schema(SCHEMA) .createWriterFunc(GenericParquetWriter::buildWriter) .build()) { appender.addAll(records); } return HadoopInputFile.fromPath(path, CONF); default: throw new UnsupportedOperationException("Cannot write format: " + format); } }
.createWriterFunc(ignored -> new SparkAvroWriter(schema)) .setAll(properties) .schema(schema)