datax 写入oracle,怎么增量替换

在 DataX 中，实现 Oracle 增量替换的方式通常有两种：

1. 使用 `writer.oracle` 的 `replace` 参数

可以在 Oracle Writer 的配置文件中设置 `replace` 参数为 `true`，这样 DataX 就会使用 replace 方式进行增量替换。具体配置如下：

{
  "job": {
    "content": [
      {
        "reader": {
          "name": "mysqlreader",
          "parameter": {
            "username": "root",
            "password": "123456",
            "column": [
              "id",
              "name",
              "age",
              "sex"
            ],
            "splitPk": "id",
            "connection": [
              {
                "jdbcUrl": "jdbc:mysql://localhost:3306/test?useSSL=false",
                "table": [
                  "user"
                ]
              }
            ]
          }
        },
        "writer": {
          "name": "oraclewriter",
          "parameter": {
            "username": "test",
            "password": "test",
            "column": [
              "id",
              "name",
              "age",
              "sex"
            ],
            "preSql": "truncate table user",
            "replace": true,
            "connection": [
              {
                "jdbcUrl": "jdbc:oracle:thin:@localhost:1521:orcl",
                "table": [
                  "user"
                ]
              }
            ]
          }
        }
      }
    ],
    "setting": {
      "speed": {
        "channel": 3
      }
    }
  }
}

通过设置 `replace` 参数为 `true`，Oracle Writer 将使用 `replace` 方式进行增量替换。

2. 在 SQL 语句中使用 `merge` 语句

另一种方式是在 SQL 语句中使用 `merge` 语句实现增量替换。具体实现方法如下：

{
  "job": {
    "content": [
      {
        "reader": {
          "name": "mysqlreader",
          "parameter": {
            "username": "root",
            "password": "123456",
            "column": [
              "id",
              "name",
              "age",
              "sex"
            ],
            "splitPk": "id",
            "connection": [
              {
                "jdbcUrl": "jdbc:mysql://localhost:3306/test?useSSL=false",
                "table": [
                  "user"
                ]
              }
            ]
          }
        },
        "writer": {
          "name": "oraclewriter",
          "parameter": {
            "username": "test",
            "password": "test",
            "column": [
              "id",
              "name",
              "age",
              "sex"
            ],
            "preSql": "truncate table user",
            "connection": [
              {
                "jdbcUrl": "jdbc:oracle:thin:@localhost:1521:orcl",
                "table": [
                  "user"
                ]
              }
            ],
            "writeMode": "batch",
            "postSql": [
              "merge into user t1 using (select ? as id, ? as name, ? as age, ? as sex from dual) t2 on (t1.id = t2.id) when matched then update set t1.name = t2.name, t1.age = t2.age, t1.sex = t2.sex when not matched then insert (t1.id, t1.name, t1.age, t1.sex) values (t2.id, t2.name, t2.age, t2.sex)"
            ]
          }
        }
      }
    ],
    "setting": {
      "speed": {
        "channel": 3
      }
    }
  }
}

在 Oracle Writer 的配置文件中设置 `writeMode` 参数为 `batch`，并在 `postSql` 中使用 `merge` 语句实现增量替换。在 `merge` 语句中，使用 `when matched then update` 来更新已有记录，使用 `when not matched then insert` 来插入新记录。