This is currently my config file
{
"config": {
"haltOnError": false
},
"source": {
"file": {
"path": "/home/user1/temp/real_user/user3.csv"
}
},
"extractor": {
"csv": {
"columns": ["id", "name", "token", "username", "password", "created", "updated", "enabled", "is_admin", "is_banned", "userAvatar"],
"columnsOnFirstLine": true
},
"field": {
"fieldName": "created",
"expression": "created.asDateTime()"
}
},
"transformers": [{
"vertex": {
"class": "user"
}
}],
"loader": {
"orientdb": {
"dbURL": "plocal:/home/user1/orientdb/real_user",
"dbAutoCreateProperties": true,
"dbType": "graph",
"classes": [{
"name": "user",
"extends": "V"
}],
"indexes": [{
"class": "user",
"fields": ["id:long"],
"type": "UNIQUE"
}]
}
}
}and my csv currently looks like this
6,Olivia Ong,2jkjkl54k5jklj5k4j5k4jkkkjjkj,\N,\N,2013-11-15 16:36:33,2013-11-15 16:36:33,1,0,\N,\N
7,Matthew,32kj4h3kjh44hjk3hk43hkhhkjhasd,\N,\N,2013-11-18 17:29:13,2013-11-15 16:36:33,1,0,\N,\N
I still wonder when I execute the ETL, orientdb wont recognize my datetime as datetime.
I tried putting datatype in column fields "created:datetime", but it ended up not showing any data.
I wonder what is the proper solution for this case.
from next version, 2.2.8, you will be able to define different default pattern for date and datetime: CSV extractor documentation
Note that when you define the columns, you need to specify the column's type:
"columns": ["id:string", "created:date", "updated:datetime"],
You can use the snapshot jar of 2.2.8 of ETL module with 2.2.7 without any problem: