标签：toDataStream Insert DataStream Handling FLINK time table DataTypes event

The following code shows how to use toDataStream for different scenarios.

import org.apache.flink.streaming.api.datastream.DataStream;
import org.apache.flink.table.api.DataTypes;
import org.apache.flink.table.api.Table;
import org.apache.flink.types.Row;
import java.time.Instant;

// POJO with mutable fields
// since no fully assigning constructor is defined, the field order
// is alphabetical [event_time, name, score]
public static class User {

    public String name;

    public Integer score;

    public Instant event_time;
}

tableEnv.executeSql(
    "CREATE TABLE GeneratedTable "
    + "("
    + "  name STRING,"
    + "  score INT,"
    + "  event_time TIMESTAMP_LTZ(3),"
    + "  WATERMARK FOR event_time AS event_time - INTERVAL '10' SECOND"
    + ")"
    + "WITH ('connector'='datagen')");

Table table = tableEnv.from("GeneratedTable");


// === EXAMPLE 1 ===

// use the default conversion to instances of Row

// since `event_time` is a single rowtime attribute, it is inserted into the DataStream
// metadata and watermarks are propagated

DataStream<Row> dataStream = tableEnv.toDataStream(table);


// === EXAMPLE 2 ===

// a data type is extracted from class `User`,
// the planner reorders fields and inserts implicit casts where possible to convert internal
// data structures to the desired structured type

// since `event_time` is a single rowtime attribute, it is inserted into the DataStream
// metadata and watermarks are propagated

DataStream<User> dataStream = tableEnv.toDataStream(table, User.class);

// data types can be extracted reflectively as above or explicitly defined

DataStream<User> dataStream =
    tableEnv.toDataStream(
        table,
        DataTypes.STRUCTURED(
            User.class,
            DataTypes.FIELD("name", DataTypes.STRING()),
            DataTypes.FIELD("score", DataTypes.INT()),
            DataTypes.FIELD("event_time", DataTypes.TIMESTAMP_LTZ(3))));

Note that only non-updating tables are supported by toDataStream. Usually, time-based operations such as windows, interval joins, or the MATCH_RECOGNIZE clause are a good fit for insert-only pipelines next to simple operations like projections and filters.

标签：toDataStream,Insert,DataStream,Handling,FLINK,time,table,DataTypes,event
来源： https://www.cnblogs.com/qiu-hua/p/15204075.html

本站声明： 1. iCode9 技术分享网（下文简称本站）提供的所有内容，仅供技术学习、探讨和分享；
2. 关于本站的所有留言、评论、转载及引用，纯属内容发起人的个人观点，与本站观点和立场无关；
3. 关于本站的所有言论和文字，纯属内容发起人的个人观点，与本站观点和立场无关；
4. 本站文章均是网友提供，不完全保证技术分享内容的完整性、准确性、时效性、风险性和版权归属；如您发现该文章侵犯了您的权益，可联系我们第一时间进行删除；
5. 本站为非盈利性的个人网站，所有内容不会用来进行牟利，也不会利用任何形式的广告来间接获益，纯粹是为了广大技术爱好者提供技术内容和技术思想的分享性交流网站。