topic using parseDate for string entry type in Talend Studio

using parseDate for string entry type

Anonymous — Tue, 03 May 2016 16:06:47 GMT

Hello,
I have a column (string) containing dates in the format yyyyMMdd or 0 when the date is not indicated. I am trying to convert the string to date using this expression

getData.name.equals("0") ? null:(TalendDate.parseDate("yyyyMMdd",getData.name))

but I have this error

Exception in component tMSSqlOutput_1
java.sql.SQLException: Only dates in the AD era are accepted.

When I tried with tlogrow to see if the conversion worked after t_map, I have found that:
if I have 20060330 it returns 30-03-2006 ,but if it's 0 it returns 30-11-0002 , so it is not taking in account the first condition in my statement.
Do you have any Idea how to correct this?

Thank you

Re: using parseDate for string entry type

Anonymous — Tue, 03 May 2016 16:20:28 GMT

I do not think that it is a value of "0" that is yielding 30-11-0002 as it ought to throw an exception before hitting an SQL exception.
I also cannot see anything wrong with your "0" test, if that is what you're value is.

Re: using parseDate for string entry type

Anonymous — Wed, 04 May 2016 09:03:48 GMT

Howerver it seems the case. Here is an example of a line from my source file and the output from the tlogrow component

example.zip.zip

Re: using parseDate for string entry type

Anonymous — Wed, 04 May 2016 09:30:24 GMT

I don't think you're providing enough information.
Your tLogRow output is showing both String and Date Format values, so it can't be of Date type.
Your data is coming from Excel. Maybe something is going wrong before tMap. Try dropping in an additional tLogRow to see?
A screenshot of your tMap may also help.

Re: using parseDate for string entry type

Anonymous — Wed, 04 May 2016 11:31:26 GMT

Exactly! putting tlogrow before tmap showed me the error!
The "0" in my excel file is shown as "00000000" in the tlogrow
Now my expression is right and it's working

row2.name("00000000") ? null:(TalendDate.parseDate("yyyyMMdd",row2.name))

Thanks a lot for your help!