Load an explicit schema from an external metadata.csv file or a json file for reading csv's into dataframe

[AnandNair]

Hi,

I have a metadata csv file which contains column name, and datatype such as

Colm1: INT

Colm2: String.

I can also get the same in a json format as shown:

I can store this on ADLS. How can I convert this into a schema like: "Myschema" that I can then pass during spark.read.format("csv") method while reading the datafile for the same metadata? When I infer schema for the datafile csv for multiple incremental files , I get clashes while writing into delta such as

"Failed to merge fields 'Colm1' and 'Colm1'. Failed to merge incompatible data types IntegerType and StringType

Any pointers/notes would be appreciated.

Thanks!