Druid Ingest Format问题

 

Druid使用过程中需要对历史数据进行索引,由于历史数据都是hive表形式,分隔符为\001,所以需要druid对ingest的format 任意delimiter进行支持,以下是支持的形式:

"parser" : {
"type" : "hadoopyString",
"parseSpec" : {
"format" : "tsv",
"timestampSpec" : {
"column" : "dt",
"format" : "posix"
},
"dimensionsSpec" : {
"dimensions": ["grade","src_flag","school_name","gender_desc","prov_name","city_name","school_prov"]
},
"delimiter":"\u0001",
"listDelimiter":"\u0002",
"columns": ["dt","uid","grade","src_flag","school_name","gender_desc","prov_name","city_name","school_prov","flag"]
}
},

指定解析格式是tsv,delimiter为\u0001,listDelimiter是multi时候使用,目前没有使用,定义为\u0002,启动任务即可。