DSXchange

It should take about a minute per job to do manually, so you could do all 200 in about half a day. This is almost certainly less time than you would spend faffing around creating scripts.

You also need to specify whether this is a straight line distance or a great circle (that is, over the surface of the earth, which is curved). There are web services out there that can perform this calculation.

Perhaps also choose a sample for your initial analysis, and select the option not to profile columns over 500 characters.

As they put it in the training class for IA: "don't try to boil the ocean".

My solution would use two jobs - one to re-order the columns into the standard format, and the other to process that file. The first job would read the file as a single VarChar column, including reading the heading row as data (do not check the "first line is column headings check box), and sto...

Do the files have column headings? (If yes, that will make the task a lot easier).

That said, I don't believe there's going to be an easy "one job" solution. You will almost certainly be up for doing some coding, either a routine or a Build stage.

Do the files have column headings? (If yes, that will make the task a lot easier).

You might also like to check whether the file genuinely has DOS-style line terminators, as mentioned in the schema file.

Designed that way because not every database supports a TRUNCATE statement.

It will when your string isn't a valid timestamp.

You can improve the performance of the posted Convert() function by initializing a stage variable to the result of the concatenated Char() functions, and not further deriving the stage variable. That way, you aren't re-evaluating the same thing for every row. Your expression then becomes Convert(svC...

You will use explicit conversion, because there is no implicit conversion of string to timestamp.

Use a test utility such as SOAPui to test the invocation, and to report what comes back in the fault report. The fault information can also be captured in the Web Service Transformer stage, but using SOAPui eliminates anything in DataStage from being the culprit.

You could try nested functions, which sometimes work and sometimes doesn't in the Modify stage. Otherwise you could use two Modify stages; these cannot be adjacent, but you can put a Copy stage between them. And, yes, do note that to include microseconds your string will need to be at least 26 chara...

A specification that specifies that the output is a string must also specify the size of that string (for Char data type) or the maximum size of the string (for VarChar data type).

Can you please post all the specifications in your Modify stage?

Also, please verify that the metadata for any column mentioned in a KEEP specification is identical on input and output link?