Skip to content

Commit

Permalink
update doc
Browse files Browse the repository at this point in the history
  • Loading branch information
liuwenli committed Dec 6, 2018
1 parent 1003c2e commit e72800d
Showing 1 changed file with 9 additions and 3 deletions.
12 changes: 9 additions & 3 deletions docs/user-guide.md
Original file line number Diff line number Diff line change
Expand Up @@ -294,7 +294,9 @@ Lookup SQL具体可参考Spark Flow Transformation的Lookup SQL章节

Flink SQL 用于处理 Source Namespace 数据,from 后面直接接表名即可。Flink支持window,UDF和UDAF操作

####### Window


###### Window

process time处理方式中window中相应的字段名称为processing_time。例:SELECT name, SUM(key) as keysum from ums GROUP BY TUMBLE(processing_time, INTERVAL '1' HOUR), name;

Expand All @@ -310,7 +312,9 @@ event time处理方式中window中相应的字段名称为ums_ts_。例:SELECT

在Transformation Config中可对这三个参数进行配置,配置格式为json。例如:{"min_idle_state_retention_time":"10","max_idle_state_retention_time":"20","preserve_message_flag":"true"}

###### #UDF


###### UDF

Wormhole Flink UDF支持普通的java程序,而不需要按照Flink官方文档的格式实现UDF。UDF名称大小写敏感。UDF相应的字段需要使用as指定新字段的名称。例如:

Expand All @@ -324,7 +328,9 @@ Java程序:
使用UDF的Flink SQL:

select intvalue, fInt(intvalue) as fint from mytable;
###### #UDAF


###### UDAF

(1)使用UDAF需要进行以下操作

Expand Down

0 comments on commit e72800d

Please sign in to comment.