"""
SDR Status Tracker - FastAPI Backend
Fetches data from Google Sheets and serves the dashboard
Supports Excel file fallback when Google Sheet access is unavailable
"""
import os
import json
import hmac
import asyncio
from datetime import datetime
from fastapi import FastAPI, HTTPException, Request
from fastapi.staticfiles import StaticFiles
from fastapi.responses import FileResponse, JSONResponse, StreamingResponse
from google.oauth2 import service_account
from googleapiclient.discovery import build
from googleapiclient.errors import HttpError

# Excel file support (optional, for local testing only)
try:
    from excel_parser import read_excel_file, get_excel_file_path
    EXCEL_SUPPORT = True
except ImportError:
    EXCEL_SUPPORT = False
    read_excel_file = None
    get_excel_file_path = None

app = FastAPI(title="SDR Status Tracker")

# Configuration via environment variables (fallback for backward compatibility)
SHEET_ID = os.environ.get("GOOGLE_SHEET_ID", "1af6-2KsRqeTQxdw5KVRp2WCrM6RT7HIcl70m-GgGZB4")
SHEET_GID = os.environ.get("GOOGLE_SHEET_GID", "1864606926")  # Tab ID (configured via env var for current month)

# Month configuration file path (relative to this script's directory)
MONTHS_CONFIG_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), "months_config.json")

# Month configuration (loaded at startup)
_months_config = {"months": [], "default_month": None, "loaded_at": None}

def load_months_config(force_reload=False):
    """
    Load month configuration from JSON file.
    Falls back to env var for backward compatibility if file not found.
    """
    global _months_config

    if _months_config["months"] and not force_reload:
        return _months_config

    try:
        with open(MONTHS_CONFIG_FILE, "r") as f:
            config = json.load(f)

        _months_config["months"] = config.get("months", [])
        _months_config["default_month"] = config.get("default_month")
        _months_config["loaded_at"] = datetime.now().isoformat()
        print(f"Months config loaded from {MONTHS_CONFIG_FILE}: {len(_months_config['months'])} months")

    except FileNotFoundError:
        print(f"Months config file {MONTHS_CONFIG_FILE} not found, using env var fallback")
        # Fallback: create single month from env var
        _months_config["months"] = [{
            "id": "default",
            "label": "Current Month",
            "sheet_id": SHEET_ID,
            "tab_name": "DAILY - for SDR to add data🌟"
        }]
        _months_config["default_month"] = "default"
        _months_config["loaded_at"] = datetime.now().isoformat()

    except Exception as e:
        print(f"Error loading months config: {e}, using env var fallback")
        _months_config["months"] = [{
            "id": "default",
            "label": "Current Month",
            "sheet_id": SHEET_ID,
            "tab_name": "DAILY - for SDR to add data🌟"
        }]
        _months_config["default_month"] = "default"
        _months_config["loaded_at"] = datetime.now().isoformat()

    return _months_config


def get_month_config(month_id: str = None):
    """Get configuration for a specific month. Returns None if not found."""
    config = load_months_config()
    if not month_id:
        month_id = config["default_month"]

    for month in config["months"]:
        if month["id"] == month_id:
            return month
    return None


# Load months config at startup
load_months_config()

# Cache - per-month with webhook invalidation
# Structure: {"2026-01": {"data": [...], "timestamp": datetime}, ...}
_cache = {}
CACHE_TTL = 3600  # 1 hour - webhook will invalidate on actual changes

# Webhook secret for cache invalidation (optional security)
WEBHOOK_SECRET = os.environ.get("WEBHOOK_SECRET", "")

# SSE clients - set of asyncio.Queue objects for connected browsers
_sse_clients: set[asyncio.Queue] = set()


def get_sheets_service():
    """Create Google Sheets API service using service account credentials."""
    creds_json = os.environ.get("GOOGLE_CREDENTIALS")
    if not creds_json:
        raise HTTPException(status_code=500, detail="GOOGLE_CREDENTIALS not configured")

    try:
        creds_dict = json.loads(creds_json)
        credentials = service_account.Credentials.from_service_account_info(
            creds_dict,
            scopes=["https://www.googleapis.com/auth/spreadsheets.readonly"]
        )
        return build("sheets", "v4", credentials=credentials)
    except Exception as e:
        raise HTTPException(status_code=500, detail=f"Failed to initialize Sheets API: {str(e)}")


# Known activity types for block detection
# Note: SQL variants (like "SQL (offer sent)") also terminate blocks
KNOWN_ACTIVITIES = {
    'calls', 'emails', 'linkedin', 'prospects',
    'prospects (activated)', 'discovery', 'sql', 'sql (offer sent)'
}

# Column configuration file path
CONFIG_FILE = "column_config.json"

# Default week configuration (fallback if config file not found)
# Format: (slot_index, daily_start, daily_end, target_col, percentage_col)
# slot_index is 0-based; actual week = start_week + slot_index
DEFAULT_WEEK_CONFIGS = [
    (0, 3, 7, 8, 9),      # Slot 0: daily D-H (3-7), target I (8), pct J (9)
    (1, 10, 14, 15, 16),  # Slot 1: daily K-O (10-14), target P (15), pct Q (16)
    (2, 17, 21, 22, 23),  # Slot 2: daily R-V (17-21), target W (22), pct X (23)
    (3, 24, 28, 30, 31),  # Slot 3: daily Y-AC (24-28), target AE (30), pct AF (31) - extra empty col
]

# Default monthly column configuration (fallback)
DEFAULT_MONTHLY_CONFIG = {"target_col": 32, "actual_col": 33, "pct_col": 34}

# Global column configuration (loaded at startup, can be reloaded)
_column_config = {"weeks": None, "monthly": None, "loaded_at": None}


def load_column_config(force_reload=False):
    """
    Load column configuration from JSON file.
    Falls back to hardcoded defaults if file not found or invalid.
    """
    global _column_config

    if _column_config["weeks"] is not None and not force_reload:
        return _column_config

    try:
        with open(CONFIG_FILE, "r") as f:
            config = json.load(f)

        # Convert weeks config to tuple format
        weeks = []
        for w in config.get("weeks", []):
            weeks.append((
                w["week_num"],
                w["daily_start"],
                w["daily_end"],
                w["target_col"],
                w["pct_col"]
            ))

        monthly = config.get("monthly", DEFAULT_MONTHLY_CONFIG)

        _column_config["weeks"] = weeks if weeks else DEFAULT_WEEK_CONFIGS
        _column_config["monthly"] = monthly
        _column_config["loaded_at"] = datetime.now().isoformat()
        print(f"Column config loaded from {CONFIG_FILE}: {len(weeks)} weeks")

    except FileNotFoundError:
        print(f"Config file {CONFIG_FILE} not found, using defaults")
        _column_config["weeks"] = DEFAULT_WEEK_CONFIGS
        _column_config["monthly"] = DEFAULT_MONTHLY_CONFIG
        _column_config["loaded_at"] = datetime.now().isoformat()

    except Exception as e:
        print(f"Error loading config: {e}, using defaults")
        _column_config["weeks"] = DEFAULT_WEEK_CONFIGS
        _column_config["monthly"] = DEFAULT_MONTHLY_CONFIG
        _column_config["loaded_at"] = datetime.now().isoformat()

    return _column_config


def get_week_configs():
    """Get the current week configurations."""
    config = load_column_config()
    return config["weeks"]


def get_monthly_config():
    """Get the current monthly column configuration."""
    config = load_column_config()
    return config["monthly"]


# Load config at module initialization
load_column_config()


def get_activity(row):
    """Extract activity type from row (column C, index 2)."""
    if len(row) > 2 and row[2]:
        return row[2].strip()
    return None


def scan_block_for_names(block):
    """
    Scan all rows in a block to find Case and GS names.
    Due to merged cells, names may appear on any row within the block.
    """
    case_name = None
    gs_name = None

    for row_idx, row in block:
        if len(row) > 0 and row[0] and row[0].strip():
            case_name = row[0].strip()
        if len(row) > 1 and row[1] and row[1].strip():
            gs_name = row[1].strip()

    return case_name, gs_name


def group_rows_into_blocks(values):
    """
    Group data rows into blocks. Each block ends with an SQL row.
    Returns list of blocks, where each block is a list of (row_idx, row) tuples.
    """
    blocks = []
    current_block = []

    for row_idx, row in enumerate(values):
        if row_idx < 4:  # Skip header rows (rows 1-4)
            continue
        if not row or len(row) < 3:
            continue

        activity = get_activity(row)
        if not activity:
            continue

        activity_lower = activity.lower()
        # Check if activity is known, or is an SQL variant
        is_known = activity_lower in KNOWN_ACTIVITIES or activity_lower.startswith('sql')
        if not is_known:
            continue

        current_block.append((row_idx, row))

        # SQL or SQL variants terminate the block
        if activity_lower.startswith('sql'):  # Block complete
            blocks.append(current_block)
            current_block = []

    # Don't lose incomplete blocks (blocks without SQL at the end)
    if current_block:
        blocks.append(current_block)

    return blocks


def process_activity_row(row, case_name, gs_name, case_data, start_week=2):
    """Process a single activity row and add data to case_data dict."""
    activity = get_activity(row)
    if not activity:
        return

    # Map activity types to our data structure
    activity_lower = activity.lower()
    activity_map = {
        "calls": "calls",
        "emails": "emails",
        "linkedin": "linkedin",
        "prospects": "prospects",
        "prospects (activated)": "prospects",
        "discovery": "discovery",
        "sql": "sql",
        "sql (offer sent)": "sql"
    }

    activity_key = activity_map.get(activity_lower)
    # Handle any other SQL variants not in the map
    if not activity_key and activity_lower.startswith('sql'):
        activity_key = 'sql'
    if not activity_key:
        return

    # Create key for this case+gs combination
    key = f"{case_name}|{gs_name}"

    # Calculate actual week numbers based on start_week
    week_configs = get_week_configs()
    week_numbers = [start_week + slot for slot, _, _, _, _ in week_configs]

    if key not in case_data:
        case_data[key] = {
            "case": case_name,
            "gs": gs_name,
            "weeks": {w: {} for w in week_numbers},
            "monthlyTotal": {"sql": 0, "sqlTarget": 0, "activity": 0, "activityTarget": 0, "sqlPctList": [], "activityPctList": []}
        }

    # Extract weekly data: actual (sum of daily), target, and percentage from sheet
    for slot, daily_start, daily_end, target_col, pct_col in week_configs:
        week_num = start_week + slot
        actual = sum_daily(row, daily_start, daily_end)
        target = safe_int(row, target_col)
        percentage = extract_percentage(row, pct_col)

        if week_num not in case_data[key]["weeks"]:
            case_data[key]["weeks"][week_num] = {}

        case_data[key]["weeks"][week_num][activity_key] = actual
        case_data[key]["weeks"][week_num][f"{activity_key}Target"] = target
        # Store percentage for debugging/validation (frontend calculates its own)
        if percentage is not None:
            case_data[key]["weeks"][week_num][f"{activity_key}Pct"] = percentage

    # Get monthly target, actual, and percentage from config
    monthly_config = get_monthly_config()
    monthly_target = safe_int(row, monthly_config["target_col"])
    monthly_actual = safe_int(row, monthly_config["actual_col"])
    monthly_pct = extract_percentage(row, monthly_config.get("pct_col"))

    # Update monthly totals
    # Discovery is aggregated with SQL for monthly view (both are pipeline metrics)
    if activity_key == "sql":
        case_data[key]["monthlyTotal"]["sql"] = monthly_actual
        case_data[key]["monthlyTotal"]["sqlTarget"] = monthly_target
        if monthly_pct is not None:
            case_data[key]["monthlyTotal"]["sqlPctList"].append(monthly_pct)
    elif activity_key == "discovery":
        # Discovery adds to SQL totals for monthly (pipeline metric)
        case_data[key]["monthlyTotal"]["sql"] += monthly_actual
        case_data[key]["monthlyTotal"]["sqlTarget"] += monthly_target
        if monthly_pct is not None:
            case_data[key]["monthlyTotal"]["sqlPctList"].append(monthly_pct)
    elif activity_key in ["calls", "emails", "linkedin", "prospects"]:
        # Activity includes outreach activities only
        case_data[key]["monthlyTotal"]["activity"] += monthly_actual
        case_data[key]["monthlyTotal"]["activityTarget"] += monthly_target
        if monthly_pct is not None:
            case_data[key]["monthlyTotal"]["activityPctList"].append(monthly_pct)


def parse_sheet_data(values, start_week=2):
    """
    Parse the DAILY sheet data into the format expected by the dashboard.

    Uses block-based parsing to handle Google Sheets merged cells correctly.
    Each SDR/Case has a variable number of activity rows (2-6), with SQL always
    being the last row of each block. Case/GS names may appear on ANY row within
    a block due to merged cell behavior.

    DAILY sheet structure (0-indexed columns):
    - Column A (0): Case name (merged - may be empty)
    - Column B (1): GS/SDR name (merged - may be empty)
    - Column C (2): Activity type (Calls, Emails, LinkedIn, Prospects, SQL)

    Each week block = 5 daily columns + 1 target column + 1 percentage column.
    Column positions are the same each month, but week numbers vary:
    - Slot 0: Cols 3-7 (daily), Col 8 (target), Col 9 (percentage)
    - Slot 1: Cols 10-14 (daily), Col 15 (target), Col 16 (percentage)
    - Slot 2: Cols 17-21 (daily), Col 22 (target), Col 23 (percentage)
    - Slot 3: Cols 24-28 (daily), Col 30 (target), Col 31 (percentage)

    Monthly totals:
    - Column AG (32): Monthly TARGET
    - Column AH (33): Monthly ACTUAL
    - Column AI (34): Monthly PERCENTAGE

    Args:
        values: Raw sheet data
        start_week: The first week number for this month (e.g., 2 for Jan, 6 for Feb)
    """
    if not values or len(values) < 5:
        return []

    # Phase 1: Group rows into blocks (SQL terminates each block)
    blocks = group_rows_into_blocks(values)

    # Phase 2: For each block, find Case/GS and attribute all rows
    case_data = {}
    for block in blocks:
        case_name, gs_name = scan_block_for_names(block)
        if not case_name or not gs_name:
            continue  # Skip blocks without proper attribution

        for row_idx, row in block:
            process_activity_row(row, case_name, gs_name, case_data, start_week)

    # Convert to list format
    cases = []
    for idx, (key, data) in enumerate(case_data.items()):
        cases.append({
            "id": idx + 1,
            "case": data["case"],
            "gs": data["gs"],
            "weeks": data["weeks"],
            "monthlyTotal": data["monthlyTotal"]
        })

    return cases


def safe_int(row, idx):
    """Safely extract an integer from a row."""
    if idx is None or idx >= len(row):
        return 0
    try:
        val = row[idx]
        if val == "" or val is None:
            return 0
        return int(float(val))
    except (ValueError, TypeError):
        return 0


def sum_daily(row, start_col, end_col):
    """Sum daily values from start_col to end_col (inclusive)."""
    total = 0
    for i in range(start_col, end_col + 1):
        total += safe_int(row, i)
    return total


def extract_percentage(row, col):
    """Extract percentage value from a cell (e.g., '267%' -> 267)."""
    if col is None or col >= len(row):
        return None
    try:
        val = row[col]
        if val == "" or val is None:
            return None
        # Handle percentage strings like "267%" or "50%"
        if isinstance(val, str):
            val = val.replace('%', '').strip()
        return int(float(val))
    except (ValueError, TypeError):
        return None


@app.get("/api/data")
async def get_data(month: str = None, source: str = None):
    """
    Fetch data from Google Sheets or local Excel file.

    Args:
        month: Month ID (e.g., "2026-02"). Uses default if not specified.
        source: Data source override. "file" forces local Excel file,
                "api" forces Google Sheets API. Auto-detects if not specified.
    """
    global _cache

    # Get month configuration
    months_config = load_months_config()
    if not month:
        month = months_config["default_month"]

    month_config = get_month_config(month)
    if not month_config:
        raise HTTPException(status_code=400, detail=f"Unknown month: {month}")

    # Check if month config specifies a file source
    config_source = month_config.get("source", "api")
    if source:
        config_source = source  # Override with query param

    # Check per-month cache
    now = datetime.now()
    cache_key = f"{month}:{config_source}"
    if cache_key in _cache:
        cache_entry = _cache[cache_key]
        if cache_entry["data"] and cache_entry["timestamp"]:
            age = (now - cache_entry["timestamp"]).total_seconds()
            if age < CACHE_TTL:
                return JSONResponse(content={
                    "cases": cache_entry["data"],
                    "cached": True,
                    "month": month,
                    "month_label": month_config["label"],
                    "source": cache_entry.get("source", "api")
                })

    # Try to load data from the specified source
    values = None
    actual_source = None
    error_messages = []

    # If source is "file" or config specifies file, try file first
    if config_source == "file":
        if not EXCEL_SUPPORT:
            error_messages.append("Excel support not available (excel_parser module not installed)")
        else:
            file_path = month_config.get("file_path") or get_excel_file_path(month)
            if file_path:
                try:
                    tab_name = month_config.get("tab_name", "DAILY - for SDR to add data🌟")
                    values = read_excel_file(file_path, tab_name)
                    actual_source = "file"
                    print(f"Loaded {len(values)} rows from Excel file: {file_path}")
                except Exception as e:
                    error_messages.append(f"Excel file error: {str(e)}")
            else:
                error_messages.append(f"No Excel file found for month {month}")

    # If source is "api" or file failed, try Google Sheets API
    if values is None and config_source != "file":
        try:
            service = get_sheets_service()
            sheet_id = month_config["sheet_id"]
            sheet_name = month_config.get("tab_name", "DAILY - for SDR to add data🌟")
            result = service.spreadsheets().values().get(
                spreadsheetId=sheet_id,
                range=f"'{sheet_name}'!A:AI"
            ).execute()
            values = result.get("values", [])
            actual_source = "api"
        except HttpError as e:
            error_messages.append(f"Google Sheets API error: {str(e)}")
            # If API fails with permission error, try file fallback
            if EXCEL_SUPPORT and ("403" in str(e) or "permission" in str(e).lower()):
                file_path = month_config.get("file_path") or get_excel_file_path(month)
                if file_path:
                    try:
                        tab_name = month_config.get("tab_name", "DAILY - for SDR to add data🌟")
                        values = read_excel_file(file_path, tab_name)
                        actual_source = "file"
                        print(f"API permission denied, fallback to Excel: {file_path}")
                    except Exception as file_e:
                        error_messages.append(f"Fallback Excel error: {str(file_e)}")
        except Exception as e:
            error_messages.append(f"Error: {str(e)}")

    # If we still have no data, raise an error
    if values is None:
        raise HTTPException(
            status_code=500,
            detail=f"Failed to load data. Errors: {'; '.join(error_messages)}"
        )

    # Parse the data with month-specific start_week
    start_week = month_config.get("start_week", 2)
    cases = parse_sheet_data(values, start_week)

    # Update cache
    _cache[cache_key] = {"data": cases, "timestamp": now, "source": actual_source}

    return JSONResponse(content={
        "cases": cases,
        "cached": False,
        "month": month,
        "month_label": month_config["label"],
        "source": actual_source
    })


@app.get("/api/config")
async def get_config():
    """Return current configuration (sheet ID, etc.)."""
    return JSONResponse(content={
        "sheetId": SHEET_ID,
        "sheetGid": SHEET_GID,
        "cacheTtl": CACHE_TTL
    })


@app.get("/api/months")
async def get_months():
    """Return list of available months for the dropdown selector."""
    config = load_months_config()
    return JSONResponse(content={
        "months": [
            {"id": m["id"], "label": m["label"], "start_week": m.get("start_week", 2)}
            for m in config["months"]
        ],
        "default_month": config["default_month"]
    })


@app.post("/api/invalidate-cache")
async def invalidate_cache(request: Request, month: str = None):
    """
    Webhook endpoint to invalidate the cache when sheet data changes.
    Called by Google Apps Script onEdit trigger or frontend refresh button.
    If month is specified, only that month's cache is cleared.
    If month is not specified, all months' caches are cleared.
    """
    global _cache

    print(f"[WEBHOOK] Cache invalidation requested, month={month}")

    # Optional: verify webhook secret if configured
    if WEBHOOK_SECRET:
        auth_header = request.headers.get("X-Webhook-Secret", "")
        if not hmac.compare_digest(auth_header, WEBHOOK_SECRET):
            print(f"[WEBHOOK] Invalid webhook secret, rejecting request")
            raise HTTPException(status_code=401, detail="Invalid webhook secret")

    # Clear the cache (specific month or all)
    if month:
        # Clear ALL sources for this month (api, file, etc.)
        # Cache keys are stored as "{month}:{source}" (e.g., "2026-02:api")
        keys_to_clear = [k for k in _cache.keys() if k.startswith(f"{month}:")]
        for key in keys_to_clear:
            del _cache[key]
        cleared = keys_to_clear
    else:
        cleared = list(_cache.keys())
        _cache.clear()

    # Notify all connected SSE clients to refresh
    for queue in _sse_clients.copy():
        try:
            queue.put_nowait("refresh")
        except asyncio.QueueFull:
            pass  # Skip if queue is full

    print(f"[WEBHOOK] Cleared cache keys: {cleared}, notified {len(_sse_clients)} SSE clients")

    return JSONResponse(content={
        "success": True,
        "message": "Cache invalidated",
        "months_cleared": cleared,
        "clients_notified": len(_sse_clients),
        "timestamp": datetime.now().isoformat()
    })


async def sse_event_generator(queue: asyncio.Queue):
    """Generator that yields SSE events for a connected client."""
    try:
        while True:
            # Wait for an event (with timeout to send keepalive)
            try:
                event = await asyncio.wait_for(queue.get(), timeout=30.0)
                yield f"data: {event}\n\n"
            except asyncio.TimeoutError:
                # Send keepalive comment to prevent connection timeout
                yield ": keepalive\n\n"
    except asyncio.CancelledError:
        pass


@app.get("/api/events")
async def sse_events():
    """
    Server-Sent Events endpoint for real-time updates.
    Browsers connect here to receive instant refresh notifications.
    """
    queue: asyncio.Queue = asyncio.Queue(maxsize=10)
    _sse_clients.add(queue)

    async def event_stream():
        try:
            async for event in sse_event_generator(queue):
                yield event
        finally:
            _sse_clients.discard(queue)

    return StreamingResponse(
        event_stream(),
        media_type="text/event-stream",
        headers={
            "Cache-Control": "no-cache",
            "Connection": "keep-alive",
            "X-Accel-Buffering": "no",  # Disable nginx buffering
        }
    )


@app.get("/api/cache-status")
async def cache_status(month: str = None):
    """Check current cache status for a specific month or all months."""
    now = datetime.now()

    if month:
        # Check specific month
        if month in _cache and _cache[month].get("timestamp"):
            age = (now - _cache[month]["timestamp"]).total_seconds()
            return JSONResponse(content={
                "cached": True,
                "month": month,
                "age_seconds": age,
                "ttl_seconds": CACHE_TTL,
                "expires_in": max(0, CACHE_TTL - age)
            })
        return JSONResponse(content={"cached": False, "month": month})

    # Return status for all cached months
    status = {}
    for m, entry in _cache.items():
        if entry.get("timestamp"):
            age = (now - entry["timestamp"]).total_seconds()
            status[m] = {
                "cached": True,
                "age_seconds": age,
                "expires_in": max(0, CACHE_TTL - age)
            }
    return JSONResponse(content={"months": status, "ttl_seconds": CACHE_TTL})


@app.post("/api/reload-config")
async def reload_config(request: Request):
    """
    Reload column configuration and months configuration.
    Also invalidates all data caches to force a fresh fetch with new config.
    """
    global _cache

    # Optional: verify webhook secret if configured
    if WEBHOOK_SECRET:
        auth_header = request.headers.get("X-Webhook-Secret", "")
        if not hmac.compare_digest(auth_header, WEBHOOK_SECRET):
            raise HTTPException(status_code=401, detail="Invalid webhook secret")

    # Reload configurations
    column_config = load_column_config(force_reload=True)
    months_config = load_months_config(force_reload=True)

    # Invalidate all data caches
    _cache.clear()

    return JSONResponse(content={
        "success": True,
        "message": "Configs reloaded, all caches invalidated",
        "column_config_loaded_at": column_config["loaded_at"],
        "months_config_loaded_at": months_config["loaded_at"],
        "weeks_count": len(column_config["weeks"]),
        "months_count": len(months_config["months"]),
        "timestamp": datetime.now().isoformat()
    })


@app.get("/api/column-config")
async def get_column_config():
    """Return current column configuration."""
    config = load_column_config()
    return JSONResponse(content={
        "weeks": [
            {
                "week_num": w[0],
                "daily_start": w[1],
                "daily_end": w[2],
                "target_col": w[3],
                "pct_col": w[4]
            }
            for w in config["weeks"]
        ],
        "monthly": config["monthly"],
        "loaded_at": config["loaded_at"]
    })


@app.get("/api/debug")
async def debug_data(month: str = None):
    """Debug endpoint to see raw sheet data structure and parsed blocks."""
    try:
        # Get month configuration
        months_config = load_months_config()
        if not month:
            month = months_config["default_month"]

        month_config = get_month_config(month)
        if not month_config:
            raise HTTPException(status_code=400, detail=f"Unknown month: {month}")

        service = get_sheets_service()
        sheet_id = month_config["sheet_id"]
        sheet_name = month_config.get("tab_name", "DAILY - for SDR to add data🌟")
        result = service.spreadsheets().values().get(
            spreadsheetId=sheet_id,
            range=f"'{sheet_name}'!A1:AI150"
        ).execute()
        values = result.get("values", [])

        # Parse blocks using the new algorithm
        blocks = group_rows_into_blocks(values)

        # Build block summary
        block_summaries = []
        unattributed_blocks = []

        for block_idx, block in enumerate(blocks):
            case_name, gs_name = scan_block_for_names(block)
            row_range = f"{block[0][0]+1}-{block[-1][0]+1}" if block else "empty"
            activities = [get_activity(row) for _, row in block]

            summary = {
                "block_num": block_idx + 1,
                "row_range": row_range,
                "case": case_name,
                "gs": gs_name,
                "activities": activities,
                "row_count": len(block)
            }

            if case_name and gs_name:
                block_summaries.append(summary)
            else:
                unattributed_blocks.append(summary)

        # Find Bannerflow/Ardit's rows for specific debugging
        ardit_rows = []
        for i, row in enumerate(values):
            if len(row) > 1 and ("ardit" in str(row).lower() or "bannerflow" in str(row).lower()):
                ardit_rows.append({"row_num": i+1, "data": row[:10]})  # Limit columns for readability

        return JSONResponse(content={
            "total_rows": len(values),
            "total_blocks": len(blocks),
            "attributed_blocks": len(block_summaries),
            "unattributed_blocks_count": len(unattributed_blocks),
            "headers_row_1": values[0] if values else [],
            "headers_row_4": values[3] if len(values) > 3 else [],
            "blocks": block_summaries,
            "unattributed_blocks": unattributed_blocks,
            "ardit_rows": ardit_rows
        })
    except Exception as e:
        import traceback
        return JSONResponse(content={"error": str(e), "traceback": traceback.format_exc()})


# Serve static files
app.mount("/static", StaticFiles(directory="static"), name="static")


@app.get("/")
async def root():
    """Serve the main dashboard."""
    return FileResponse("static/index.html")


if __name__ == "__main__":
    import uvicorn
    uvicorn.run(app, host="0.0.0.0", port=7860)