validate schema from a file

[Hvuj]

Hi.
is it possible to do schema validation from a script, yaml or json file?

some thing like this:

from io import StringIO
import pandas as pd
from pathlib import Path
import pandera as pa


class Schema(pa.DataFrameModel):
        columns_a: Series[pa.String] = pa.Field(some_logic)
        ... some more columns
    

try:
    data = Schema.to_schema().validate(some_data_frame)
except pa.errors.SchemaErrors as exc:
    data_lines = exc.data.strip().split('\n')[:-1]
    cleaned_data_str = "\n".join(data_lines)
    df = pd.read_csv(StringIO(cleaned_data_str),sep="\s+")
    schema = pa.infer_schema(df)
    folder_path = Path.cwd() / "dataframe_schemas"
    folder_path.mkdir(parents=True, exist_ok=True)

   ---- here i would like to use that file as the new reference to the fixed schema 
(imagine here there is a functionality to adjust the columns and checks and auto fix) 
e.g. use the schema.to_script(file_path)

not sure how to properly use the infer_schema in my case

[Hvuj]

Upon further investigation - currently i see its possible by using only DataFrameSchema from_yaml or from_json, e.g.

pa. DataFrameSchema.from_json()
pa. DataFrameSchema.from_yaml()

while DataFrameModel can only use to_yaml, to_schema()

there is also a possibility of using from_frictionless_schema to read third party schemas which has a schema parameter:

schema (Union[str, Path, Dict, Schema])

essentially what i want to do is possible but requires writing to a specific format and from there added functionality to auto update schemas.

a bit unsatisfactory as DataFrameModel should also have this ability to adjust fields.

current approach - fix schemas, convert to string representation of DataFrameModel and write to yaml and from there to schema validation.