1
votes

I'm trying to execute a simple pipeline in azure data lake analytics, but I'm having some trouble with U-SQL. I was wondering if someone can give a helping hand.

My Query:

DECLARE @log_file string = "/datalake/valores.tsv";
DECLARE @summary_file string = "/datalake/output.tsv";

@log = EXTRACT valor string from @log_file USING Extractors.Tsv(); 

@summary = select sum(int.valor) as somavalor from @log;OUTPUT @summary 
TO @summary_file USING Outputters.Tsv();

Error: Erro

Other general doubts: 1. When I deploy a new pipeline to ADF sometimes it doesn't appear in the activity window and sometime it does. I didn't get the logic. (I'm using the OneTime pipeline mode) 2. There is a better way to create new pipeline (other than manipulate raw Json files?) 3.There is any U-SQL parser? What is the easiest way to teste my query?

Thanks a lot.

1
Please use formatting tools to properly edit and format your question/answer. Codes within sentences are to be formatted as code Very Important words to be bold , lesser important onces Italic Also use lists if necessaryMorse

1 Answers

1
votes

U-SQL is case-sensitive so your U-SQL should look more like this:

DECLARE @log_file string = "/datalake/valores.tsv";
DECLARE @summary_file string = "/datalake/output.tsv";

@log =
    EXTRACT valor int
    FROM @log_file
    USING Extractors.Tsv();

@summary =
    SELECT SUM(valor) AS somavalor
    FROM @log;

OUTPUT @summary 
TO @summary_file USING Outputters.Tsv();

I have assumed your input file has only a single column of type int.

Use Visual Studio U-SQL projects, VS Code U-SQL add-in to ensure you write valid U-SQL. You can also submit U-SQL jobs via the portal.