Validating datagrid values
This kind of input validation is done automatically and cannot be customized.The other type of input validation can be used to reject any unacceptable data, for example a zero value in a field that must be greater than or equal to one, or an inappropriate string. The example below uses the Column Changing event because the unacceptable value is disallowed for the "Product" column in particular.EG: The Most Important thing is to disable the Save button when the form is invalid.For this we had a static property in our view model named Errors.I am looking at various hybrid approaches including using the Value mapper, filter and validator.I may also have to bite the bullet and learn some java scripting. At least I feel like I'm looking at the same thing as everyone else and I'm not just mis-specifying something. I am having the exact problem with a SQL generated table input or fix file inputs. We don't want to pull the notes, fix one field, return them and then have to get them back again for the next bad field.I can roll the fail streams from multiple validator blocks back together but it would be more elegant if I didn't have too.
There are several dates in various formats like yyyy/mm/dd hh:mm:ss.sss; numeric fields with upper and lower bounds and then there are some that need values from a list.Nick For some of your "constraints" why not use Filter Rows step?To be honest, I know it's not elegant, but having multiple steps make it easier on a later time to open the transformation and understand at a glance what each step is doing. Hi Mick, Because it's a data cleaning exercise there's something wrong with at least one variable in almost every row.The attraction of the data validator is that I can get a single row output of all the errors.
It's medical data so fixing it normally involves pulling a set of patient's notes and verifying the value.
I would probably use 2 Data validation Steps, it would make it easier to read the transformation and it should not slow it down too much.