data_count.py•1.67 kB
import json
from typing import Any
from mcp import types
from pydantic import ConfigDict
from .model import Data
class DataCountInputSchema(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) -> "DataCountInputSchema":
data = Data.from_file(input_data_file_path)
return DataCountInputSchema(df=data.df)
@staticmethod
def from_args(arguments: dict[str, Any]) -> "DataCountInputSchema":
input_data_file_path = arguments["input_data_file_path"]
return DataCountInputSchema.from_schema(input_data_file_path=input_data_file_path)
async def handle_data_count(
arguments: dict[str, Any],
) -> list[types.TextContent | types.ImageContent | types.EmbeddedResource]:
data_count_input = DataCountInputSchema.from_args(arguments)
count_df = data_count_input.df.count()
# Convert the DataFrame to a dictionary format
count_dict = {
"description": "Number of non-null elements for each column",
"counts": {col: int(val) for col, val in zip(count_df.columns, count_df.row(0))},
}
return [
types.TextContent(
type="text",
text=json.dumps(count_dict),
)
]