Have you/do you have access to any data profiling tools, such as Information Server Information Analyzer or InfoSphere Discovery (there are also a whole host of non ibm and open source options). These sorts of tools should be used during the onboarding/analysis of new data sources to identify the unique patterns (e.g. pattern YYYY-MM-DD occurs 70% of the time and DD-MMM-YYYY occurs 20% etc) as opposed to values found in the data so that transformation requirements can be identified upfront.
Else if you want to test in DataStage a precheck something like
this post might allow you to trap the data and perform some additional transformations.