import asyncio
import base64
import io
import logging
import wave
from typing import Optional

import requests
from elevenlabs.client import ElevenLabs
from fastapi import FastAPI, File, HTTPException, UploadFile
from fastapi.middleware.cors import CORSMiddleware
from google import genai
from google.genai import types
from PIL import Image
from pydantic import BaseModel
from pydub import AudioSegment

from prompts import (ACCESSIBILITY_PROMPT, NARRATION_PROMPT,
                     NARRATION_SYSTEM_PROMPT)

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)


app = FastAPI(
    title="Accessibility Service API",
    description="API for generating audio narrations and making images accessible",
    version="1.0.0",
)

app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)


class ProcessImageRequest(BaseModel):
    imageUrl: str
    googleApiKey: str
    elevenlabsApiKey: Optional[str] = None


class ProcessImageUploadRequest(BaseModel):
    googleApiKey: str
    elevenlabsApiKey: Optional[str] = None


class ProcessImageResponse(BaseModel):
    accessibleImage: str
    description: str
    narrationURL: Optional[str] = None


def get_google_client(api_key: str) -> genai.Client:
    """Create and return a Google Genai client with the provided API key"""
    try:
        return genai.Client(api_key=api_key)
    except Exception as e:
        raise HTTPException(
            status_code=400, detail=f"Failed to initialize Google client: {str(e)}"
        )


def get_elevenlabs_client(api_key: str) -> ElevenLabs:
    """Create and return an ElevenLabs client with the provided API key"""
    try:
        return ElevenLabs(api_key=api_key)
    except Exception as e:
        raise HTTPException(
            status_code=400, detail=f"Failed to initialize ElevenLabs client: {str(e)}"
        )


def download_image(url: str) -> Image.Image:
    """Download image from URL and return PIL Image object"""
    try:
        image_bytes = requests.get(url).content
        image = types.Part.from_bytes(data=image_bytes, mime_type="image/jpeg")
        return image
    except Exception as e:
        raise HTTPException(
            status_code=400, detail=f"Failed to download image: {str(e)}"
        )


def image_to_base64(image: Image.Image) -> str:
    """Convert PIL Image to base64 data URL"""
    buffer = io.BytesIO()
    image.save(buffer, format="PNG")
    img_bytes = buffer.getvalue()
    img_base64 = base64.b64encode(img_bytes).decode()
    return f"data:image/png;base64,{img_base64}"


async def generate_description(image: Image.Image, google_client: genai.Client) -> str:
    """Generate a text description of the image using a generative AI model."""
    try:
        response = google_client.models.generate_content(
            model="gemini-2.5-flash",
            contents=[image, NARRATION_PROMPT],
            config=types.GenerateContentConfig(
                system_instruction=NARRATION_SYSTEM_PROMPT, temperature=0.1
            ),
        )
        return response.text
    except Exception as e:
        raise HTTPException(
            status_code=500, detail=f"Failed to generate description: {str(e)}"
        )


def audio_to_base64(audio_bytes: bytes, mime: str = "audio/mpeg") -> str:
    """Convert audio bytes to base64 data URL"""
    audio_base64 = base64.b64encode(audio_bytes).decode()
    return f"data:{mime};base64,{audio_base64}"


async def generate_narration_audio(text: str, elevenlabs_client) -> str:
    """Generate audio from ElevenLabs (MP3) and return base64 data URL."""
    try:
        audio = elevenlabs_client.text_to_speech.convert(
            text=text,
            voice_id="XfNU2rGpBa01ckF309OY",
            model_id="eleven_multilingual_v2",
            output_format="mp3_44100_128",
            apply_text_normalization="on",
        )

        audio_bytes = b""
        for chunk in audio:
            audio_bytes += chunk

        return audio_to_base64(audio_bytes, mime="audio/mpeg")

    except Exception as e:
        raise HTTPException(
            status_code=500, detail=f"Failed to generate audio: {str(e)}"
        )


async def generate_narration_audio_gemini(description, google_client) -> str:
    """Generate audio from Gemini (PCM → WAV → MP3) and return base64 data URL."""
    try:
        response = google_client.models.generate_content(
            model="gemini-2.5-flash-preview-tts",
            contents=description,
            config=types.GenerateContentConfig(
                response_modalities=["AUDIO"],
                speech_config=types.SpeechConfig(
                    voice_config=types.VoiceConfig(
                        prebuilt_voice_config=types.PrebuiltVoiceConfig(
                            voice_name="Leda",
                        )
                    )
                ),
            ),
        )

        audio_data = response.candidates[0].content.parts[0].inline_data.data

        if isinstance(audio_data, str):
            audio_data = base64.b64decode(audio_data)

        wav_buffer = io.BytesIO()
        with wave.open(wav_buffer, "wb") as wf:
            wf.setnchannels(1)  # mono
            wf.setsampwidth(2)  # 16-bit PCM
            wf.setframerate(24000)  # 24kHz
            wf.writeframes(audio_data)
        wav_buffer.seek(0)

        audio_segment = AudioSegment.from_wav(wav_buffer)
        mp3_buffer = io.BytesIO()
        audio_segment.export(mp3_buffer, format="mp3")
        mp3_bytes = mp3_buffer.getvalue()

        return audio_to_base64(mp3_bytes, mime="audio/mpeg")

    except Exception as e:
        raise HTTPException(
            status_code=500, detail=f"Failed to generate audio (Gemini): {str(e)}"
        )


async def generate_accessible_image(
    original_image: Image.Image, google_client: genai.Client
) -> str:
    """
    Generate an accessible version of the image using AI image generation.
    """
    try:
        response = google_client.models.generate_content(
            model="gemini-2.5-flash-image-preview",
            contents=[ACCESSIBILITY_PROMPT, original_image],
        )
        first_image = None
        for part in response.candidates[0].content.parts:
            if part.inline_data is not None:
                first_image = Image.open(io.BytesIO(part.inline_data.data))

        if not first_image:
            raise HTTPException(
                status_code=500, detail="Failed to generate accessible image"
            )

        return image_to_base64(first_image)
    except Exception as e:
        raise HTTPException(
            status_code=500, detail=f"Failed to generate accessible image: {str(e)}"
        )


async def _process(
    google_api_key: str, elevenlabs_api_key: Optional[str], original_image: Image.Image
) -> ProcessImageResponse:
    google_client = get_google_client(google_api_key)

    description, accessible_image_base64 = await asyncio.gather(
        generate_description(original_image, google_client),
        generate_accessible_image(original_image, google_client),
    )
    logger.info(f"Generated image and description")

    # Generate narration only if ElevenLabs API key is provided
    narration_url = "None"
    if elevenlabs_api_key:
        try:
            elevenlabs_client = get_elevenlabs_client(elevenlabs_api_key)
            narration_url = await generate_narration_audio(
                description, elevenlabs_client
            )
        except Exception as e:
            logger.warning(f"Warning: Failed to generate audio narration: {str(e)}")
    else:
        narration_url = await generate_narration_audio_gemini(
            description, google_client
        )

    logger.info(f"Generated narration audio")

    return ProcessImageResponse(
        accessibleImage=accessible_image_base64,
        description=description,
        narrationURL=narration_url,
    )


@app.post("/process", response_model=ProcessImageResponse)
async def process_image(request: ProcessImageRequest):
    """
    Process an image URL to generate accessible version, description, and narration.
    Runs all generation tasks in parallel for optimal performance.

    Args:
        request: JSON payload containing imageUrl, googleApiKey, and optional elevenlabsApiKey

    Returns:
        JSON response with accessibleImage, description, and optional narrationURL
    """
    logger.info(f"Processing image from URL: {request.imageUrl}")
    try:
        image = await asyncio.to_thread(download_image, request.imageUrl)

        return await _process(request.googleApiKey, request.elevenlabsApiKey, image)

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


@app.post("/processupload")
async def process_image_upload(
    image: UploadFile = File(...),
    googleApiKey: str = None,
    elevenlabsApiKey: Optional[str] = None,
):
    """
    Process an uploaded image file to generate accessible version, description, and narration.

    Args:
        image: Uploaded image file
        googleApiKey: Google API key (required)
        elevenlabsApiKey: ElevenLabs API key (optional)

    Returns:
        JSON response with accessibleImage, description, and optional narrationURL
    """
    try:
        if not googleApiKey:
            raise HTTPException(status_code=400, detail="googleApiKey is required")

        if not image.content_type.startswith("image/"):
            raise HTTPException(status_code=400, detail="File must be an image")

        image_data = await image.read()

        try:
            pil_image = Image.open(io.BytesIO(image_data))
            pil_image.verify()  # Verify it's a valid image
        except Exception as e:
            raise HTTPException(status_code=400, detail=f"Invalid image file: {str(e)}")

        pil_image = Image.open(io.BytesIO(image_data))

        buffer = io.BytesIO()
        pil_image.save(buffer, format="PNG")
        buffer.seek(0)
        original_image = types.Part.from_bytes(
            data=buffer.getvalue(), mime_type="image/png"
        )

        return await _process(googleApiKey, elevenlabsApiKey, original_image)

    except HTTPException:
        # Re-raise HTTP exceptions
        raise
    except Exception as e:
        # Catch any other errors
        raise HTTPException(status_code=500, detail=f"Error processing image: {str(e)}")


@app.get("/health")
async def health_check():
    """Health check endpoint"""
    return {"status": "healthy", "message": "Accessibility Service API is running"}