"""
API Integration Guide - Applying Dynamic Validation to COGNEXA ML Endpoints

This guide shows how to integrate the DynamicRequestValidator from dynamic_api_handler.py
into your FastAPI endpoints to fix 422 validation errors.

PROBLEM: The following endpoints return 422 errors due to strict Pydantic validation:
- /api/personality/analyze-responses
- /api/productivity/forecast  
- /api/personalization/predict-engagement

SOLUTION: Replace Pydantic BaseModel with flexible dynamic validation
"""

# ============================================================================
# STEP 1: Import the Dynamic Validator
# ============================================================================

from fastapi import FastAPI, Body
from fastapi.responses import JSONResponse

app = FastAPI()

from ML.dynamic_api_handler import DynamicRequestValidator, FlexibleRequestHandler

# Helper function stubs (API integration guide only)

def analyze_personality_scores(responses: dict, response_scale_max: int):
    return {"responses": responses, "scale": response_scale_max}


def generate_forecast(user_id: str, historical_data: list, forecast_days: int, personality: dict = None):
    return {"user_id": user_id, "forecast_days": forecast_days, "data_points": len(historical_data)}


def predict_user_engagement(user_id: str, task_id: str, task: dict, user_profile: dict = None):
    return 0.5


def run_prediction(user_id: str, complexity: float, count: int):
    return {"user_id": user_id, "complexity": complexity, "count": count}

# ============================================================================
# STEP 2: Update Personality Endpoint
# ============================================================================

# BEFORE (Strict Pydantic - will fail on "4.5" instead of 4.5):
"""
from pydantic import BaseModel

class PersonalityResponse(BaseModel):
    responses: dict
    response_scale_max: int

@app.post("/api/personality/analyze-responses")
async def analyze_responses(request: PersonalityResponse):
    # If user sends {"q1": "4"} instead of {"q1": 4}, this fails with 422
    ...
"""

# AFTER (Flexible Dynamic Validation):
from fastapi import Body
from fastapi.responses import JSONResponse

@app.post("/api/personality/analyze-responses")
async def analyze_responses(request: dict = Body(...)):
    """
    Flexible endpoint that coerces types and handles missing fields.
    
    Accepts:
    - {"q1": 4, "q2": "3.5", ...} ✅ Coerces string to float
    - {"q1": 4} ✅ Fills missing fields with defaults
    - {"q1": "abc"} ✅ Returns helpful error message instead of 422
    """
    
    # Validate and coerce the request
    try:
        validated = DynamicRequestValidator.validate_personality_responses(request)
        
        # Proceed with analysis
        result = analyze_personality_scores(
            responses=validated['responses'],
            response_scale_max=validated['response_scale_max']
        )
        
        return JSONResponse(
            status_code=200,
            content={"status": "success", "analysis": result}
        )
    
    except ValueError as e:
        return JSONResponse(
            status_code=400,  # Bad Request (not 422)
            content={"error": str(e)}
        )


# ============================================================================
# STEP 3: Update Productivity Forecast Endpoint
# ============================================================================

# BEFORE:
"""
class ProductivityForecast(BaseModel):
    user_id: str
    forecast_days: int
    historical_data: list

@app.post("/api/productivity/forecast")
async def forecast_productivity(request: ProductivityForecast):
    # Strict validation fails on type mismatches
    ...
"""

# AFTER:
@app.post("/api/productivity/forecast")
async def forecast_productivity(request: dict = Body(...)):
    """
    Flexible forecast endpoint with intelligent type coercion.
    
    Accepts:
    - Different data formats
    - Missing optional fields (uses defaults)
    - String numbers ("7" → 7)
    """
    
    try:
        validated = DynamicRequestValidator.validate_productivity_forecast(request)
        
        # Proceed with forecasting
        forecast = generate_forecast(
            user_id=validated['user_id'],
            historical_data=validated['historical_data'],
            forecast_days=validated['forecast_days'],
            personality=validated.get('personality')
        )
        
        return JSONResponse(
            status_code=200,
            content={"status": "success", "forecast": forecast}
        )
    
    except ValueError as e:
        return JSONResponse(
            status_code=400,
            content={"error": str(e)}
        )


# ============================================================================
# STEP 4: Update Engagement Prediction Endpoint
# ============================================================================

# BEFORE:
"""
class EngagementPrediction(BaseModel):
    user_id: str
    task_id: str
    task: dict

@app.post("/api/personalization/predict-engagement")
async def predict_engagement(request: EngagementPrediction):
    # Fails on any type mismatch
    ...
"""

# AFTER:
@app.post("/api/personalization/predict-engagement")
async def predict_engagement(request: dict = Body(...)):
    """
    Flexible engagement prediction with flexible input handling.
    
    Accepts:
    - Various input formats
    - Flexible data structures
    - Intelligent field coercion
    """
    
    try:
        validated = DynamicRequestValidator.validate_engagement_prediction(request)
        
        # Proceed with prediction
        engagement_score = predict_user_engagement(
            user_id=validated['user_id'],
            task_id=validated['task_id'],
            task=validated['task'],
            user_profile=validated.get('user_profile', {})
        )
        
        return JSONResponse(
            status_code=200,
            content={
                "status": "success",
                "engagement_prediction": engagement_score
            }
        )
    
    except ValueError as e:
        return JSONResponse(
            status_code=400,
            content={"error": str(e)}
        )


# ============================================================================
# STEP 5: Alternative - Use FlexibleRequestHandler for Complex Validation
# ============================================================================

@app.post("/api/complex/predict")
async def complex_prediction(request: dict = Body(...)):
    """
    For more complex validation logic, use FlexibleRequestHandler.
    """
    
    handler = FlexibleRequestHandler()
    
    try:
        # Coerce individual fields
        user_id = handler.coerce_value(request.get('user_id'), str, required=True)
        complexity = handler.coerce_value(request.get('complexity'), float)  # "4.5" → 4.5
        count = handler.coerce_value(request.get('count'), int, default=0)   # "5" → 5
        
        # Proceed with prediction
        result = run_prediction(
            user_id=user_id,
            complexity=complexity,
            count=count
        )
        
        return {"status": "success", "result": result}
    
    except (ValueError, TypeError) as e:
        return JSONResponse(
            status_code=400,
            content={"error": f"Invalid input: {str(e)}"}
        )


# ============================================================================
# STEP 6: Migration Checklist
# ============================================================================

"""
CHECKLIST FOR MIGRATING ENDPOINTS:

1. ✅ Identify problematic endpoints (search for Pydantic BaseModel usage)
   - /api/personality/analyze-responses
   - /api/productivity/forecast
   - /api/personalization/predict-engagement
   
2. ✅ Replace endpoint signature:
   OLD: def endpoint(request: PydanticModel):
   NEW: def endpoint(request: dict = Body(...)):
   
3. ✅ Add DynamicRequestValidator call:
   validated = DynamicRequestValidator.validate_*(request)
   
4. ✅ Update field access:
   OLD: request.field_name
   NEW: validated['field_name']
   
5. ✅ Add error handling:
   try:
       # validation & processing
   except ValueError as e:
       return JSONResponse(status_code=400, content={"error": str(e)})
   
6. ✅ Test with various input types:
   - {"field": 5}          # integer
   - {"field": "5"}        # string number
   - {"field": 5.5}        # float
   - {"field": "5.5"}      # string float
   - {}                    # missing (should use defaults)
   - {"field": "invalid"}  # invalid (should return 400, not 422)
   
7. ✅ Verify 422 errors → 400 errors:
   - Invalid input now returns 400 Bad Request
   - Type coercion handles edge cases
   - Helpful error messages provided
   
8. ✅ Test all endpoints after migration:
   python test_api_endpoints.py
"""


# ============================================================================
# STEP 7: Example Requests (cURL)
# ============================================================================

"""
# Test Personality Analysis Endpoint
curl -X POST http://localhost:8000/api/personality/analyze-responses \\
  -H "Content-Type: application/json" \\
  -d '{
    "responses": {
      "q1": 4,
      "q2": "3.5",
      "q3": 5
    },
    "response_scale_max": 5
  }'

# Expected Response:
{
  "status": "success",
  "analysis": { ... }
}


# Test Productivity Forecast Endpoint
curl -X POST http://localhost:8000/api/productivity/forecast \\
  -H "Content-Type: application/json" \\
  -d '{
    "user_id": "user123",
    "forecast_days": "7",
    "historical_data": [
      {"date": "2026-03-20", "completed_tasks": "5"}
    ]
  }'

# Expected Response:
{
  "status": "success",
  "forecast": { ... }
}


# Test Engagement Prediction Endpoint
curl -X POST http://localhost:8000/api/personalization/predict-engagement \\
  -H "Content-Type: application/json" \\
  -d '{
    "user_id": "user123",
    "task_id": "task456",
    "task": {
      "title": "Review notes",
      "category": "WORK",
      "priority": "MEDIUM"
    }
  }'

# Expected Response:
{
  "status": "success",
  "engagement_prediction": 0.78
}
"""


# ============================================================================
# STEP 8: Validation Schemas (Reference)
# ============================================================================

"""
PERSONALITY RESPONSE VALIDATION:
- responses: dict (required)
  - Keys: string (question IDs)
  - Values: int/float (scores, will be coerced)
- response_scale_max: int (default: 5)
  - Will coerce string to int

PRODUCTIVITY FORECAST VALIDATION:
- user_id: str (required)
- historical_data: list (required)
  - List of dicts with productivity data
- forecast_days: int (default: 7)
  - Will coerce string to int
- personality: dict (optional)
  - Big Five personality scores

ENGAGEMENT PREDICTION VALIDATION:
- user_id: str (required)
- task_id: str (required)
- task: dict (required)
  - Must contain task details
- user_profile: dict (optional)
  - Optional user profile data
"""


# ============================================================================
# RESOURCES
# ============================================================================

"""
Files to reference:
- /ML/dynamic_api_handler.py      - Full validation implementation
- /ML/test_api_endpoints.py       - Comprehensive API tests
- /ML/api_patches.md              - This file
- /ML/test_ml_service.py          - Unit tests for validators

Commands:
- Test validators: python test_ml_service.py
- Test endpoints: python test_api_endpoints.py
- Run server: python main.py

Common Issues & Solutions:
1. "422 Unprocessable Entity"
   → Use DynamicRequestValidator instead of strict Pydantic
   
2. "400 Bad Request" with helpful message
   → This is expected for invalid data (better than 422)
   
3. String numbers not coercing
   → Ensure coerce_value() is being called
   → Check type hints in validation method
   
4. Missing fields causing errors
   → Set defaults in validation method
   → Use .get() with defaults in validated dict
"""