Forum Discussion

ragera4's avatar
ragera4
New Contributor II
8 years ago

Mapping Spanish characters with standard English characters

Hello,

We have a requirement where we’re expected to map Spanish characters with standard English characters. The following table illustrates the exact mapping requirement -

|Character|Replacement|

À, Á, á, à, Â, â, Ä, ä, Ã, ã, å, Å → A
Æ, æ → AE
Ç, ç → C
É, é, È, è, Ê, ê, Ë, ë → E
Í, í, Ì, ì, Î, î, Ï, ï → I
Ñ, ñ → N
Ó, ó, Ò, ò, Ô, ô, Ö, ö, Õ, õ, Ø, ø → O
Œ,œ → OE
ß → SZ
Ú, ú, Ù, ù, Û, û, Ü, ü → U
Ý, ý, Ÿ, ÿ → Y

We developed a pipeline and uploaded a .csv file containing a few Spanish characters in the file reader snap, but apparently, SnapLogic is converting those special characters into a question (?) mark.

Is there an efficient way to handle such situations,

Please let us know.

Thanks in advance.

transform

1 Reply

  • stephenknilans's avatar
    stephenknilans
    Contributor
    8 years ago

    I actually tested OUR snaplogic instance with several languages. Ironically, I didn’t test spanish, since I figured the following was good enough, but I tested:
    English
    German
    Hindi
    Traditional Chinese
    and Russian

    It worked fine with every one. BTW I used the source binary text and ZIP connections, and target binary text, and looked inside the preview run.

    Unicode, unfortunately, is TOUCHY, and is complicated even MORE by different environments. If the source, target, network, and system snaplogic(or any program) is on, you will have problems. ALSO, the code you use has to be a valid subset of UTF8 that is supported, and the system must read it as binary, or UTF8 text.

    One problem I have often had, for example, was with the language setting on the system with a driver used to access Oracle. That is an Oracle issue, and supposedly it was fixed with 12c(IIRC). So a LOT of people may have that issue and not know it. Of course, if the language setting is set right on oracle, etc… Even most of the older oracle instances work fine.

    BTW Windows also has some problems here.

    OH, and variants of a questionmark are often intentionally placed there for unicode translation errors. So it means either something didn’t handle unicode, or didn’t use the right codepage. As I said, this can mean anything from source to target.

    Try isolating the error, copying it to a small file, and going source to target. Use the snaplogic preview mode to see WHERE it fails, and go from there. If you can’t see a problem THERE, it is likely between the target driver, and the target device.