routes/predict.py

import os
import time
import shutil
from pathlib import Path
from fastapi import APIRouter, UploadFile, File, HTTPException, Depends, Body
from fastapi.responses import FileResponse
from auth import get_current_user
from services.sentence_transformer_service import SentenceTransformerService, sentence_transformer_service
from data_lib.input_name_data import InputNameData
from data_lib.base_name_data import COL_NAME_SENTENCE
from mapping_lib.subject_mapper import SubjectMapper
from mapping_lib.name_mapper import NameMapper
from config import UPLOAD_DIR, OUTPUT_DIR
from models import (
    EmbeddingRequest,
    PredictRawRequest,
    PredictRawResponse,
    PredictRecord,
    PredictResult,
)
import pandas as pd

router = APIRouter()

@router.post("/predict")
async def predict(
    current_user=Depends(get_current_user),
    file: UploadFile = File(...),
    sentence_service: SentenceTransformerService = Depends(lambda: sentence_transformer_service)
):
    """
    Process an input CSV file and return standardized names (requires authentication)
    """
    if not file.filename.endswith(".csv"):
        raise HTTPException(status_code=400, detail="Only CSV files are supported")

    # Save uploaded file
    timestamp = int(time.time())
    input_file_path = os.path.join(UPLOAD_DIR, f"input_{timestamp}_{current_user.username}.csv")
    output_file_path = os.path.join(OUTPUT_DIR, f"output_{timestamp}_{current_user.username}.csv")

    try:
        with open(input_file_path, "wb") as buffer:
            shutil.copyfileobj(file.file, buffer)
    finally:
        file.file.close()

    try:
        # Process input data
        start_time = time.time()
        try:
            inputData = InputNameData()
            inputData.load_data_from_csv(input_file_path)
        except Exception as e:
            print(f"Error processing input data: {e}")
            raise HTTPException(status_code=500, detail=str(e))
        try:
            subject_mapper = SubjectMapper(sentence_service.sentenceTransformerHelper, sentence_service.dic_standard_subject)
            dic_subject_map = subject_mapper.map_standard_subjects(inputData.dataframe)
            inputData.dic_standard_subject = dic_subject_map
            inputData.process_data()
        except Exception as e:
            print(f"Error processing input data: {e}")
            raise HTTPException(status_code=500, detail=str(e))
        # Map standard names
        try:
            nameMapper = NameMapper(
                sentence_service.sentenceTransformerHelper,
                sentence_service.standardNameMapData,
                top_count=3
            )
            df_predicted = nameMapper.predict(inputData)
        except Exception as e:
            print(f"Error mapping standard names: {e}")
            raise HTTPException(status_code=500, detail=str(e))

        # Create output dataframe and save to CSV
        # column_to_keep = ['ファイル名', 'シート名', '行', '科目', '中科目', '分類', '名称', '摘要', '備考']
        column_to_keep = ['シート名', '行', '科目', '中科目', '分類', '名称', '摘要', '備考']
        output_df = inputData.dataframe[column_to_keep].copy()
        output_df.reset_index(drop=False, inplace=True)
        output_df.loc[:, "出力_科目"] = df_predicted["標準科目"]
        output_df.loc[:, "出力_項目名"] = df_predicted["標準項目名"]
        output_df.loc[:, "出力_確率度"] = df_predicted["基準名称類似度"]

        # Save with utf_8_sig encoding for Japanese Excel compatibility
        output_df.to_csv(output_file_path, index=False, encoding="utf_8_sig")
        end_time = time.time()
        execution_time = end_time - start_time
        print(f"Execution time: {execution_time} seconds")
        return FileResponse(
            path=output_file_path,
            filename=f"output_{Path(file.filename).stem}.csv",
            media_type="text/csv",
            headers={
                "Content-Disposition": f'attachment; filename="output_{Path(file.filename).stem}.csv"',
                "Content-Type": "application/x-www-form-urlencoded",
            },
        )

    except Exception as e:
        print(f"Error processing file: {e}")
        raise HTTPException(status_code=500, detail=str(e))


@router.post("/embeddings")
async def create_embeddings(
    request: EmbeddingRequest,
    current_user=Depends(get_current_user),
    sentence_service: SentenceTransformerService = Depends(
        lambda: sentence_transformer_service
    ),
):
    """
    Create embeddings for a list of input sentences (requires authentication)
    """
    try:
        start_time = time.time()
        embeddings = sentence_service.sentenceTransformerHelper.create_embeddings(
            request.sentences
        )
        end_time = time.time()
        execution_time = end_time - start_time
        print(f"Execution time: {execution_time} seconds")
        # Convert numpy array to list for JSON serialization
        embeddings_list = embeddings.tolist()
        return {"embeddings": embeddings_list}
    except Exception as e:
        print(f"Error creating embeddings: {e}")
        raise HTTPException(status_code=500, detail=str(e))


@router.post("/predict-raw", response_model=PredictRawResponse)
async def predict_raw(
    request: PredictRawRequest,
    current_user=Depends(get_current_user),
    sentence_service: SentenceTransformerService = Depends(
        lambda: sentence_transformer_service
    ),
):
    """
    Process raw input records and return standardized names (requires authentication)
    """
    try:
        # Convert input records to DataFrame
        records_dict = {
            "科目": [],
            "中科目": [],
            "分類": [],
            "名称": [],
            "摘要": [],
            "備考": [],
            "シート名": [],  # Required by BaseNameData but not used
            "行": [],  # Required by BaseNameData but not used
        }

        for record in request.records:
            records_dict["科目"].append(record.subject)
            records_dict["中科目"].append(record.sub_subject)
            records_dict["分類"].append(record.name_category)
            records_dict["名称"].append(record.name)
            records_dict["摘要"].append(record.abstract or "")
            records_dict["備考"].append(record.memo or "")
            records_dict["シート名"].append("")  # Placeholder
            records_dict["行"].append("")  # Placeholder

        df = pd.DataFrame(records_dict)

        # Process input data
        try:
            inputData = InputNameData(sentence_service.dic_standard_subject)
            # Use _add_raw_data instead of direct assignment
            inputData._add_raw_data(df)
            inputData.process_data(sentence_service.sentenceTransformerHelper)
        except Exception as e:
            print(f"Error processing input data: {e}")
            raise HTTPException(status_code=500, detail=str(e))

        # Map standard names
        try:
            nameMapper = NameMapper(
                sentence_service.sentenceTransformerHelper,
                sentence_service.standardNameMapData,
                top_count=3,
            )
            df_predicted = nameMapper.predict(inputData)
        except Exception as e:
            print(f"Error mapping standard names: {e}")
            raise HTTPException(status_code=500, detail=str(e))

        # Convert results to response format
        results = []
        for _, row in df_predicted.iterrows():
            result = PredictResult(
                standard_subject=row["標準科目"],
                standard_name=row["標準項目名"],
                anchor_name=row["基準名称"],
                similarity_score=float(row["基準名称類似度"]),
            )
            results.append(result)

        return PredictRawResponse(results=results)

    except Exception as e:
        print(f"Error processing records: {e}")
        raise HTTPException(status_code=500, detail=str(e))