Masking Date columns error out due to "Unparseable Date" string
Article ID: 367440
Updated On:
Products
Issue/Introduction
We have a requirement to obfuscate Oracle tables, which have a billion records, but due to improper data, the job fails with the following error:
Error parsing date: €%-€%-
We are exploring if there are any options in FDM to continue the masking job though if any error occurs, like “continue on failure”. I have tried below options in FDM but no luck.
- CONTINUEONFAILURE = Y
- INTERRUPTIONDISABLED=Y
We are using the ADDDAYS function on a column containing dates (YYYY-MM-DD), but due to bad data in some of the rows for this column, the job continues to fail. Are there any options in FDM that would help skip the rows that contain bad data, and mask the remaining rows in the table?
Environment
FastDataMasker 4.10.1007.0
Cause
I believe the error is getting thrown because the string is not in the proper format (YYYY-MM-DD), and mainly because the € and % characters are not included in the masking alphabet for the ADDDAYS function.
Resolution
If you are working with date strings, you could try using the DOB or DOD Function, along with the BADDATESTRING option, but if the original string contains a character that we cannot parse, it will still cause a failure.
I was able to get the job to run using the FORMATFPE1EXT function, and setting:
- Keep Format = true
- Replace character (parm8) = 0
However, the masked results are not valid dates, which is not ideal.
| Original DOB_OBF | Masked DOB_OBF |
| 1970-02-14 | 1788-35-02 |
| €%°€%° | 0%°0%° |
| 2008-10-31 | 2495-03-45 |
| 1969-12-25 | 1679-07-07 |
| 1976-07-04 | 1080-99-50 |
Using a WHERE cause to exclude the rows containing the invalid characters is your best option.
Additional Information
For more information regarding Masking Functions and Masking options, see:
Feedback
