Bug in tSampleRow Range when using context variables

[Anonymous]

I am writing data into a DB2 database. Sadly this is slow because the ODBC driver.

For test purposes, I am sampling the data when the context "test" is used, else I take all of them.

To show the issue, I created a minimal working example:

It works flawlessly, when the "Range" field in "tSampleRow_1" component is set to: "1,5,10..20" (the pre-filled example).

What I tried and is not working is the following: "1.." + context.db2_table_HCWH2H1P_sampleSize

In the "test" context, the context variable is set to 100, and in "production" to 99999999.

The job is not compiling. The source shows:

String[] rangetSampleRow_1 = "1.." + context.db2_table_HCWH2H1P_sampleSize.split(",");

with the error:

Type mismatch: Cannot convert from String to String[]

My guess is, that "Talend Open Studio for Big Data" is taking the value in the "Range" field, and puts a '.split(",")' at the end.

What happens is, that the context variable gets split rather than the whole string. It ends up with a "String" and this mismatches the "String[]" variable type of "rangetSampleRow_1".

An easy solution is, to put some brackets around the "Range" field, and append the '.split(",")' after the closing bracket.

For my case, I worked around the problem by putting the whole string "1..100" into the context variable, rather than only the upper bound "100".

Additional information:

- Product: Talend Open Studio for Big Data

- Version: 6.4.1

- OS: Win10

- Java: JDK 1.8

[TRF · Accepted solution]

Hi,

Use this syntax:

("1.." + context.db2_table_HCWH2H1P_sampleSize)

Hope this helps.

[Anonymous]

This works fine. But I still think that it is a bug in Talend.

[TRF]

Not really. This is due to the way TOS concatenates its part when generationg the Java code.
However don't forget to mark the solution as accepted.