Right now date normalization only works for a very limited set of matching formats.
This should be extended to include time formats, both standalone in a separate column and combined with a date in a single column.
This significantly increases the complexity.
It will likely require:
- Creating formats via combinations of independent date and time formats
- Creation of equivalence classes (several formats should be considered the same)
- Matching specific formats based on input string lengths (limits combinatorial explosion)
Right now date normalization only works for a very limited set of matching formats.
This should be extended to include time formats, both standalone in a separate column and combined with a date in a single column.
This significantly increases the complexity.
It will likely require: