How to get hold of Warnings in Pandera as objects instead of string ?

[nrjchnd2]

I am new to Pandera and using it to run some schema validations on my dataframe. I want to use a mix of warnings and errors. The error part is what is working seamlessly. What I am doing is to catch the rows in the original dataframe which failed validations using index column of failure_cases in SchemaErrors and send them back to users as a dataframe.

Below works for errors.

    except pa.errors.SchemaErrors as e:
        # run some custom logic logic using the index of the row which failed validation
        print(e.failure_cases.groupby('index')

I want to do the same with Warnings as well but I am not able to find an easy way to get a handle on index of the validation failed rows in case of warnings and I get warnings as type = str.

So for below code, I get

with warnings.catch_warnings(record=True) as caught_warnings:
        dynamic_schema.validate(df, lazy=True, inplace=False)
        if caught_warnings:
           print(type((warning.message.args[0])) # print str

output (below is one single string)

<Schema Column(name=series_value_date, type=DataType(str))> failed element-wise validator 0:
<Check validate: last saved date check failed>
failure cases:
   index                 failure_case
0      0  2024-03-26T00:00:00.0000000
1      1  2024-03-24T00:00:00.0000000
2      2  2024-03-23T00:00:00.0000000
3      3  2024-03-22T00:00:00.0000000

If somehow, I can access index column in above string output, this will fulfill my use case. Is there a way to do this ?