When reading a text file (.txt, .dat, .csv, etc) a warning is produced stating an unexpected or invalid character was found.
The following warning is generated during text import.
>Warning. Command name: EXECUTE
>(2279) An unexpected/invalid character was encountered in case 1, at or
>around position 5. This character was skipped.
The invalid characters are essentially skipped. This should not result in a mis-read file, but it is a good idea to manually inspect the file for damage.
Data supplied to Statistics to be read contained data that does not map to legitimate character values in the current locale setting.
Resolving the problem
Ensure that Statistics is set to the same locale as the source data file and/or is set to use Unicode (where appropriate). This is done via the SET LOCALE command (please see product documentation on this command for details).
If problem persists despite these settings then it is likely that the data being used as input contains invalid character data and the source of the data should be investigated to determine the origin of the invalid data.