data_schema.py•1.65 kB
import json
from typing import Any
from mcp import types
from pydantic import ConfigDict
from .model import Data
class DataSchemaInputSchema(Data):
model_config = ConfigDict(
validate_assignment=True,
frozen=True,
extra="forbid",
arbitrary_types_allowed=True,
)
@staticmethod
def input_schema() -> dict:
return {
"type": "object",
"properties": {
"input_data_file_path": {
"type": "string",
"description": "Path to the input data file",
},
},
}
@staticmethod
def from_schema(input_data_file_path: str) -> "DataSchemaInputSchema":
data = Data.from_file(input_data_file_path)
return DataSchemaInputSchema(df=data.df)
@staticmethod
def from_args(arguments: dict[str, Any]) -> "DataSchemaInputSchema":
input_data_file_path = arguments["input_data_file_path"]
return DataSchemaInputSchema.from_schema(input_data_file_path=input_data_file_path)
async def handle_data_schema(
arguments: dict[str, Any],
) -> list[types.TextContent | types.ImageContent | types.EmbeddedResource]:
data_schema_input = DataSchemaInputSchema.from_args(arguments)
schema = data_schema_input.df.schema
schema_dict = {col: str(dtype) for col, dtype in schema.items()}
return [
types.TextContent(
type="text",
text=json.dumps(
{
"description": "Data schema of the input data",
"schema": schema_dict,
}
),
)
]