将 SelectObjectContent 与 Amazon SDK 或命令行工具结合使用

基于 SQL 语句筛选 Amazon S3 对象的内容

以下 select-object-content 示例使用指定的 SQL 语句筛选 my-data-file.csv 对象并将输出发送到文件。

aws s3api select-object-content \
    --bucket my-bucket \
    --key my-data-file.csv \
    --expression "select * from s3object limit 100" \
    --expression-type 'SQL' \
    --input-serialization '{"CSV": {}, "CompressionType": "NONE"}' \
    --output-serialization '{"CSV": {}}' "output.csv"

此命令不生成任何输出。

以下示例演示了一个使用 JSON 对象的查询。该完整示例还演示了 CSV 对象的用法。

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import software.amazon.awssdk.core.async.AsyncRequestBody;
import software.amazon.awssdk.core.async.BlockingInputStreamAsyncRequestBody;
import software.amazon.awssdk.core.exception.SdkException;
import software.amazon.awssdk.services.s3.S3AsyncClient;
import software.amazon.awssdk.services.s3.model.CSVInput;
import software.amazon.awssdk.services.s3.model.CSVOutput;
import software.amazon.awssdk.services.s3.model.CompressionType;
import software.amazon.awssdk.services.s3.model.ExpressionType;
import software.amazon.awssdk.services.s3.model.FileHeaderInfo;
import software.amazon.awssdk.services.s3.model.InputSerialization;
import software.amazon.awssdk.services.s3.model.JSONInput;
import software.amazon.awssdk.services.s3.model.JSONOutput;
import software.amazon.awssdk.services.s3.model.JSONType;
import software.amazon.awssdk.services.s3.model.ObjectIdentifier;
import software.amazon.awssdk.services.s3.model.OutputSerialization;
import software.amazon.awssdk.services.s3.model.Progress;
import software.amazon.awssdk.services.s3.model.PutObjectResponse;
import software.amazon.awssdk.services.s3.model.SelectObjectContentRequest;
import software.amazon.awssdk.services.s3.model.SelectObjectContentResponseHandler;
import software.amazon.awssdk.services.s3.model.Stats;

import java.io.IOException;
import java.net.URL;
import java.util.ArrayList;
import java.util.List;
import java.util.UUID;
import java.util.concurrent.CompletableFuture;

public class SelectObjectContentExample {
    static final Logger logger = LoggerFactory.getLogger(SelectObjectContentExample.class);
    static final String BUCKET_NAME = "select-object-content-" + UUID.randomUUID();
    static final S3AsyncClient s3AsyncClient = S3AsyncClient.create();
    static String FILE_CSV = "csv";
    static String FILE_JSON = "json";
    static String URL_CSV = "https://raw.githubusercontent.com/mledoze/countries/master/dist/countries.csv";
    static String URL_JSON = "https://raw.githubusercontent.com/mledoze/countries/master/dist/countries.json";

    public static void main(String[] args) {
        SelectObjectContentExample selectObjectContentExample = new SelectObjectContentExample();
        try {
            SelectObjectContentExample.setUp();
            selectObjectContentExample.runSelectObjectContentMethodForJSON();
            selectObjectContentExample.runSelectObjectContentMethodForCSV();
        } catch (SdkException e) {
            logger.error(e.getMessage(), e);
            System.exit(1);
        } finally {
            SelectObjectContentExample.tearDown();
        }
    }

    EventStreamInfo runSelectObjectContentMethodForJSON() {
        // Set up request parameters.
        final String queryExpression = "select * from s3object[*][*] c where c.area < 350000";
        final String fileType = FILE_JSON;

        InputSerialization inputSerialization = InputSerialization.builder()
                .json(JSONInput.builder().type(JSONType.DOCUMENT).build())
                .compressionType(CompressionType.NONE)
                .build();

        OutputSerialization outputSerialization = OutputSerialization.builder()
                .json(JSONOutput.builder().recordDelimiter(null).build())
                .build();

        // Build the SelectObjectContentRequest.
        SelectObjectContentRequest select = SelectObjectContentRequest.builder()
                .bucket(BUCKET_NAME)
                .key(FILE_JSON)
                .expression(queryExpression)
                .expressionType(ExpressionType.SQL)
                .inputSerialization(inputSerialization)
                .outputSerialization(outputSerialization)
                .build();

        EventStreamInfo eventStreamInfo = new EventStreamInfo();
        // Call the selectObjectContent method with the request and a response handler.
        // Supply an EventStreamInfo object to the response handler to gather records and information from the response.
        s3AsyncClient.selectObjectContent(select, buildResponseHandler(eventStreamInfo)).join();

        // Log out information gathered while processing the response stream.
        long recordCount = eventStreamInfo.getRecords().stream().mapToInt(record ->
                record.split("\n").length
        ).sum();
        logger.info("Total records {}: {}", fileType, recordCount);
        logger.info("Visitor onRecords for fileType {} called {} times", fileType, eventStreamInfo.getCountOnRecordsCalled());
        logger.info("Visitor onStats for fileType {}, {}", fileType, eventStreamInfo.getStats());
        logger.info("Visitor onContinuations for fileType {}, {}", fileType, eventStreamInfo.getCountContinuationEvents());
        return eventStreamInfo;
    }

    static SelectObjectContentResponseHandler buildResponseHandler(EventStreamInfo eventStreamInfo) {
        // Use a Visitor to process the response stream. This visitor logs information and gathers details while processing.
        final SelectObjectContentResponseHandler.Visitor visitor = SelectObjectContentResponseHandler.Visitor.builder()
                .onRecords(r -> {
                    logger.info("Record event received.");
                    eventStreamInfo.addRecord(r.payload().asUtf8String());
                    eventStreamInfo.incrementOnRecordsCalled();
                })
                .onCont(ce -> {
                    logger.info("Continuation event received.");
                    eventStreamInfo.incrementContinuationEvents();
                })
                .onProgress(pe -> {
                    Progress progress = pe.details();
                    logger.info("Progress event received:\n bytesScanned:{}\nbytesProcessed: {}\nbytesReturned:{}",
                            progress.bytesScanned(),
                            progress.bytesProcessed(),
                            progress.bytesReturned());
                })
                .onEnd(ee -> logger.info("End event received."))
                .onStats(se -> {
                    logger.info("Stats event received.");
                    eventStreamInfo.addStats(se.details());
                })
                .build();

        // Build the SelectObjectContentResponseHandler with the visitor that processes the stream.
        return SelectObjectContentResponseHandler.builder()
                .subscriber(visitor).build();
    }

    // The EventStreamInfo class is used to store information gathered while processing the response stream.
    static class EventStreamInfo {
        private final List<String> records = new ArrayList<>();
        private Integer countOnRecordsCalled = 0;
        private Integer countContinuationEvents = 0;
        private Stats stats;

        void incrementOnRecordsCalled() {
            countOnRecordsCalled++;
        }

        void incrementContinuationEvents() {
            countContinuationEvents++;
        }

        void addRecord(String record) {
            records.add(record);
        }

        void addStats(Stats stats) {
            this.stats = stats;
        }

        public List<String> getRecords() {
            return records;
        }

        public Integer getCountOnRecordsCalled() {
            return countOnRecordsCalled;
        }

        public Integer getCountContinuationEvents() {
            return countContinuationEvents;
        }

        public Stats getStats() {
            return stats;
        }
    }