#!/usr/bin/env python3
"""
Complete PLC Generation Workflow

Runs the entire pipeline from raw Excel to compiled ACD:
1. PLC Data Generator (raw Excel → DESC_IP_MERGED.xlsx)
2. Routines Generator (DESC_IP_MERGED.xlsx → L5X files)
3. L5X2ACD Compiler (L5X files → ACD files)
"""

from __future__ import annotations

import sys
import argparse
import subprocess
from pathlib import Path
import io
import contextlib
import xml.etree.ElementTree as ET
import re
from datetime import datetime

def get_project_paths():
    """Get standardized paths for all project components."""
    project_root = Path(__file__).parent.parent.resolve()
    
    return {
        'project_root': project_root,
        'data_generator': project_root / "PLC Data Generator",
        'routines_generator': project_root / "Routines Generator", 
        'l5x2acd_compiler': project_root / "L5X2ACD Compiler",
        'io_tree_generator': project_root / "IO Tree Configuration Generator"
    }

def resolve_project_config_files(project_name: str, project_root: Path) -> tuple[Path, Path]:
    """Resolve configuration file paths based on project name.
    
    Args:
        project_name: Project name (e.g., 'MTN6_MCM02', 'SAT9_MCM01', 'CNO8_MCM01')
        project_root: Root directory of the project
        
    Returns:
        tuple: (generator_config_path, zones_config_path)
        
    Raises:
        FileNotFoundError: If required config files don't exist
    """
    # Extract project prefix (MTN6, SAT9, CNO8, etc.)
    import re
    project_match = re.match(r'^([A-Z0-9]+)_', project_name.upper())
    if not project_match:
        raise ValueError(f"Invalid project name format: {project_name}. Expected format: PREFIX_MCMXX")
    
    project_prefix = project_match.group(1)
    
    # Build config file paths
    generator_config = project_root / f"{project_prefix}_generator_config.json"
    zones_config = project_root / f"{project_prefix}_zones.json"
    
    # Validate files exist
    if not generator_config.exists():
        raise FileNotFoundError(f"Generator config not found: {generator_config}")
    if not zones_config.exists():
        raise FileNotFoundError(f"Zones config not found: {zones_config}")
    
    return generator_config, zones_config

def get_available_projects(project_root: Path) -> list[str]:
    """Get list of available project prefixes based on config files.
    
    Args:
        project_root: Root directory of the project
        
    Returns:
        List of available project prefixes (e.g., ['MTN6', 'SAT9', 'CNO8'])
    """
    import re
    projects = set()
    
    # Look for *_generator_config.json files
    for config_file in project_root.glob("*_generator_config.json"):
        match = re.match(r'^([A-Z0-9]+)_generator_config\.json$', config_file.name)
        if match:
            prefix = match.group(1)
            # Check if corresponding zones file exists
            zones_file = project_root / f"{prefix}_zones.json"
            if zones_file.exists():
                projects.add(prefix)
    
    return sorted(list(projects))

def resolve_boilerplate_directory(project_prefix: str, io_tree_dir: Path) -> str:
    """Resolve the boilerplate directory based on project prefix.
    
    Args:
        project_prefix: Project prefix (e.g., 'MTN6', 'SAT9', 'CNO8')
        io_tree_dir: IO Tree Configuration Generator directory
        
    Returns:
        Boilerplate directory name (e.g., 'MTN6_boilerplate')
    """
    boilerplate_dir = f"{project_prefix}_boilerplate"
    boilerplate_path = io_tree_dir / boilerplate_dir
    
    if boilerplate_path.exists():
        return boilerplate_dir
    else:
        # Fall back to default boilerplate directory
        print(f"Warning: Project-specific boilerplate directory not found: {boilerplate_dir}")
        print("         Using default 'boilerplate' directory")
        return "boilerplate"

def _append_log(log_path: Path | None, header: str, body: str) -> None:
    if not log_path:
        return
    try:
        log_path.parent.mkdir(parents=True, exist_ok=True)
        with open(log_path, 'a', encoding='utf-8') as f:
            f.write(f"\n=== {header} ===\n")
            if body:
                f.write(body)
                if not body.endswith("\n"):
                    f.write("\n")
    except Exception:
        pass


def run_plc_data_generator(raw_excel_file: Path, paths: dict, verbose: bool = False, log_file: Path | None = None) -> bool:
    """Run the PLC Data Generator to create DESC_IP_MERGED.xlsx."""
    data_gen_dir = paths['data_generator']
    data_gen_script = data_gen_dir / "main.py"
    
    if not data_gen_script.exists():
        print(f"ERROR: PLC Data Generator not found at {data_gen_script}")
        return False
        
    if not raw_excel_file.exists():
        print(f"ERROR: Raw Excel file not found at {raw_excel_file}")
        return False
    
    try:
        # Run the PLC Data Generator with the Excel file path as argument
        result = subprocess.run([
            sys.executable, 
            str(data_gen_script),
            str(raw_excel_file.resolve())  # Pass the Excel file path as argument
        ], cwd=data_gen_dir, capture_output=True, text=True)
        
        # Check if core processing succeeded by looking for output files
        # Even if there's a permission error at the end, the processing might have completed
        source = data_gen_dir / "DESC_IP_MERGED.xlsx"
        success_indicators = [
            "Processing complete!" in result.stdout,
            "New Excel file created:" in result.stdout,
            source.exists()
        ]
        
        # Consider it successful if the essential files were created, even with permission errors
        if result.returncode == 0 or (any(success_indicators) and "[Errno 1] Operation not permitted" in result.stdout):
            if verbose and result.returncode != 0:
                _append_log(log_file, "Step 1 Warning", "Permission error at end of processing, core processing completed")
            if verbose:
                _append_log(log_file, "Step 1: PLC Data Generator stdout", result.stdout)
                _append_log(log_file, "Step 1: PLC Data Generator stderr", result.stderr)
            
            # Copy DESC_IP_MERGED.xlsx from data generator output (it already has safety sheets)
            dest = paths['routines_generator'] / "DESC_IP_MERGED.xlsx"
            
            if source.exists():
                import shutil
                shutil.copy2(source, dest)
                return True
            else:
                return False
        else:
            if verbose:
                _append_log(log_file, "Step 1 Error", "Data processing failed")
                _append_log(log_file, "Step 1: PLC Data Generator stdout", result.stdout)
                _append_log(log_file, "Step 1: PLC Data Generator stderr", result.stderr)
            return False
            
    except Exception as e:
        if verbose:
            _append_log(log_file, "Step 1 Exception", str(e))
        return False

def run_routines_generator(paths: dict, project_name: str = None, ignore_estop1ok: bool = False, safety_only: bool = False, verbose: bool = False, log_file: Path | None = None, config_path: Path = None) -> bool:
    """Run the Routines Generator.

    When safety_only is True, runs safety-only generation (inputs, outputs, resets,
    estops, zones, estop_check, safety tag map). Otherwise runs the standard
    generator with DPM and other routines.
    
    Args:
        config_path: Path to project-specific generator config file. If None, uses default.
    """
    routines_dir = paths['routines_generator']
    
    try:
        # Use provided config path or fall back to default
        if config_path is None:
            config_path = paths['project_root'] / 'generator_config.json'
        
        excel_path = routines_dir / 'DESC_IP_MERGED.xlsx'
        subcmd = 'safety' if safety_only else 'all'
        # Build args with global flags BEFORE the subcommand
        cmd_args = [
            sys.executable,
            '-m', 'src.unified_cli',
            '--config', str(config_path),
            '--excel-file', str(excel_path),
        ]
        if verbose:
            cmd_args.extend(['--log-level', 'DEBUG'])
            if log_file is not None:
                cmd_args.extend(['--log-file', str(log_file)])
        cmd_args.append(subcmd)
        # Note: routine inclusion/exclusion is driven by config; project_name and ignore-estop1ok are configured in JSON
            
        # Run the unified CLI
        result = subprocess.run(cmd_args, cwd=routines_dir, capture_output=True, text=True)
        
        if verbose:
            _append_log(log_file, "Step 2: Routines Generator stdout", result.stdout)
            if result.stderr:
                _append_log(log_file, "Step 2: Routines Generator stderr", result.stderr)
        
        if result.returncode == 0:
            return True
        else:
            return False
            
    except Exception as e:
        if verbose:
            _append_log(log_file, "Step 2 Exception", str(e))
        return False

def run_io_tree_generator(paths: dict, project_name: str, safety_only: bool = False, verbose: bool = False, log_file: Path | None = None, boilerplate_dir: str = None) -> bool:
    """Run the IO Tree Configuration Generator.

    If safety_only is True, skip this step to avoid generating non-safety routines.
    """
    if safety_only:
        return True
    
    io_tree_dir = paths['io_tree_generator']
    enhanced_mcm_script = io_tree_dir / "enhanced_mcm_generator.py"
    # Use the file directly from PLC Data Generator since we're skipping Routines Generator
    desc_ip_file = paths['data_generator'] / "DESC_IP_MERGED.xlsx"
    
    if not enhanced_mcm_script.exists():
        return False
    
    # Zones fully removed: do not attempt to load or pass zones
    zones_json = None
    
    try:
        # Build command arguments
        cmd_args = [
            sys.executable, 
            str(enhanced_mcm_script),
            str(desc_ip_file),
            project_name
        ]
        
        # Add boilerplate directory if specified
        if boilerplate_dir:
            cmd_args.append(boilerplate_dir)
            
        # Run the IO Tree Configuration Generator
        result = subprocess.run(cmd_args, cwd=io_tree_dir, capture_output=True, text=True)
        
        if verbose:
            _append_log(log_file, "Step 3: IO Tree Generator stdout", result.stdout)
            if result.stderr:
                _append_log(log_file, "Step 3: IO Tree Generator stderr", result.stderr)
        
        if result.returncode == 0:
            return True
        else:
            return False
            
    except Exception as e:
        if verbose:
            _append_log(log_file, "Step 3 Exception", str(e))
        return False

def run_l5x_to_acd_compiler(paths: dict, project_name: str, safety_only: bool = False, verbose: bool = False, log_file: Path | None = None) -> bool:
    """Prepare for L5X2ACD Compilation using dynamic compilation manager.

    If safety_only is True, skip this step since a full project L5X wasn't generated.
    """
    if safety_only:
        return True
    
    # Find the generated complete project L5X file
    io_tree_dir = paths['io_tree_generator']
    generated_projects_dir = io_tree_dir / "generated_projects"
    
    if not generated_projects_dir.exists():
        return False
    
    # Look for L5X files that start with the project name
    l5x_files = list(generated_projects_dir.glob(f"{project_name}*.L5X"))
    
    if not l5x_files:
        # Minimal output: rely on caller to report FAIL
        return False
    
    if len(l5x_files) > 1 and verbose:
        print(f"Warning: Multiple L5X files found, using first: {l5x_files[0].name}")
    
    complete_l5x = l5x_files[0]

    # Inject SafetyTagMap from SafetyTagMapping.txt before compilation (if available)
    try:
        mapping_file = paths['routines_generator'] / 'SafetyTagMapping.txt'
        if mapping_file.exists():
            # Always inject silently (minimal logging)
            _inject_safety_tag_map_into_l5x(complete_l5x, mapping_file, False)
    except Exception as e:
        if verbose:
            _append_log(log_file, "Step 4 Warning", f"Failed to inject SafetyTagMap: {e}")
    
    # Use the dynamic compilation manager
    l5x2acd_dir = paths['project_root'] / "L5X2ACD Compiler"
    
    try:
        # Import and use the compilation manager
        import sys
        sys.path.append(str(l5x2acd_dir))
        from compilation_manager import CompilationManager
        
        # Create compilation manager
        manager = CompilationManager(l5x2acd_dir)
        
        # Determine project-specific options (silent)
        project_type = "UNKNOWN"
        options = {}
        if project_name:
            if "MCM01" in project_name.upper():
                project_type = "MCM01"
                options['enable_safety_validation'] = True
            elif "MCM04" in project_name.upper():
                project_type = "MCM04"
                options['enable_feeder_optimization'] = True
        
        # Setup compilation with smart replacement (no wiping)
        # Always run quietly and suppress tool output; caller prints step OK/FAIL
        _buf = io.StringIO()
        with contextlib.redirect_stdout(_buf), contextlib.redirect_stderr(_buf):
            result = manager.setup_compilation(
                source_l5x=complete_l5x,
                project_name=project_name or complete_l5x.stem,
                compilation_options=options,
                wipe_existing=False,  # Use smart replacement instead of wiping
                replace_mode=True
            )
        if verbose:
            _append_log(log_file, "Step 4: L5X2ACD Compiler output", _buf.getvalue())
        
        return True
        
    except Exception as e:
        # Minimal output; let caller handle FAIL display
        return False


def _inject_safety_tag_map_into_l5x(l5x_path: Path, mapping_file: Path, verbose: bool = False) -> None:
    """Inject or replace <SafetyTagMap> inside the existing Controller/SafetyInfo using text edits.

    - Preserves the original XML header exactly
    - Does not create additional SafetyInfo blocks
    - Formats SafetyTagMap on its own line between SafetyInfo open/close tags
    """
    mapping_text = mapping_file.read_text(encoding='utf-8').strip()
    if not mapping_text:
        if verbose:
            print("SafetyTagMapping.txt is empty; skipping injection")
        return

    xml_text = l5x_path.read_text(encoding='utf-8')

    # Find Controller block
    ctrl_match = re.search(r"<Controller\b[\s\S]*?</Controller>", xml_text)
    if not ctrl_match:
        if verbose:
            print("No <Controller> found; skipping injection")
        return
    ctrl_start, ctrl_end = ctrl_match.span()
    ctrl_text = xml_text[ctrl_start:ctrl_end]

    # Locate first SafetyInfo (body or self-closing)
    m_body = re.search(r"<SafetyInfo\b([^>]*)>([\s\S]*?)</SafetyInfo>", ctrl_text)
    m_self = re.search(r"<SafetyInfo\b([^>]*)/>", ctrl_text)

    if not m_body and not m_self:
        if verbose:
            print("No <SafetyInfo> under <Controller>; skipping injection")
        return

    # Determine indentation based on the SafetyInfo line
    first_match = m_body if (m_body and (not m_self or m_body.start() < m_self.start())) else m_self
    safety_start = first_match.start()
    line_start = ctrl_text.rfind('\n', 0, safety_start)
    indent = ctrl_text[line_start+1:safety_start] if line_start != -1 else ''
    map_line = f"\n{indent}  <SafetyTagMap> {mapping_text} </SafetyTagMap>\n"

    def dedup_safety_infos(text: str) -> str:
        seen = False
        def repl(match: re.Match) -> str:
            nonlocal seen
            if seen:
                return ''
            seen = True
            return match.group(0)
        pat = re.compile(r"(<SafetyInfo\b[^>]*/>)|(\n?\s*<SafetyInfo\b[^>]*>[\s\S]*?</SafetyInfo>)")
        return pat.sub(repl, text)

    if m_body and (not m_self or m_body.start() < m_self.start()):
        # Replace or insert SafetyTagMap inside existing body
        attrs = m_body.group(1)
        inner = m_body.group(2)
        # Replace existing map if present
        if re.search(r"<SafetyTagMap>[\s\S]*?</SafetyTagMap>", inner):
            new_inner = re.sub(r"<SafetyTagMap>[\s\S]*?</SafetyTagMap>", map_line.strip('\n'), inner, count=1)
            # Remove any additional maps
            new_inner = re.sub(r"<SafetyTagMap>[\s\S]*?</SafetyTagMap>", '', new_inner)
        else:
            new_inner = map_line + inner
        new_block = f"<SafetyInfo{attrs}>{new_inner}</SafetyInfo>"
        new_ctrl_text = ctrl_text[:m_body.start()] + new_block + ctrl_text[m_body.end():]
    else:
        # Convert self-closing to body with map
        attrs = m_self.group(1)
        new_block = f"<SafetyInfo{attrs}>{map_line}</SafetyInfo>"
        new_ctrl_text = ctrl_text[:m_self.start()] + new_block + ctrl_text[m_self.end():]

    new_ctrl_text = dedup_safety_infos(new_ctrl_text)
    new_xml = xml_text[:ctrl_start] + new_ctrl_text + xml_text[ctrl_end:]
    l5x_path.write_text(new_xml, encoding='utf-8')
    if verbose:
        print("SafetyTagMap injection OK")

def main() -> None:
    """Main entry point for complete workflow."""
    
    parser = argparse.ArgumentParser(description="Complete PLC generation workflow from raw Excel to ACD")
    parser.add_argument('--excel-file', type=str, help='Raw Excel file to process')
    
    # Project selection
    parser.add_argument('--project', help='Project prefix (e.g., MTN6, SAT9, CNO8) - automatically selects config files')
    parser.add_argument('--project-name', help='Project name (e.g., MTN6_MCM02) - used for output naming and compatibility')
    
    parser.add_argument('--ignore-estop1ok', action='store_true', help='Ignore ESTOP1OK tags in safety routines generation')
    parser.add_argument('--safety-only', action='store_true', help='Generate only safety routines and safety checks')
    parser.add_argument('--verbose', action='store_true', help='Write detailed logs for each step to a file (no console spam)')
    parser.add_argument('--list-projects', action='store_true', help='List available projects and exit')
    
    args = parser.parse_args()
    
    # Get project paths
    paths = get_project_paths()
    
    # Handle --list-projects (allow without --excel-file)
    if args.list_projects:
        available_projects = get_available_projects(paths['project_root'])
        print("Available projects:")
        for project in available_projects:
            generator_config = paths['project_root'] / f"{project}_generator_config.json"
            zones_config = paths['project_root'] / f"{project}_zones.json"
            boilerplate_dir = paths['io_tree_generator'] / f"{project}_boilerplate"
            boilerplate_status = "✓" if boilerplate_dir.exists() else "✗"
            print(f"  {project:<6} - Config: {generator_config.name}, Zones: {zones_config.name}, Boilerplate: {boilerplate_status}")
        if not available_projects:
            print("  No projects found. Expected files: PREFIX_generator_config.json and PREFIX_zones.json")
        return
    
    # Resolve project configuration
    generator_config_path = None
    zones_config_path = None
    project_name = args.project_name  # Use provided project name if given
    
    if args.project:
        # Project selection mode - use project prefix to find config files
        try:
            # If no project name provided, derive it from Excel file name
            if not project_name:
                excel_name = args.excel_file.stem
                if 'MCM' in excel_name.upper():
                    # Try to extract MCM info from filename
                    import re
                    mcm_match = re.search(r'(MCM\d+)', excel_name.upper())
                    if mcm_match:
                        project_name = f"{args.project.upper()}_{mcm_match.group(1)}"
                    else:
                        project_name = f"{args.project.upper()}_MCM01"  # Default fallback
                else:
                    project_name = f"{args.project.upper()}_MCM01"  # Default fallback
            
            # Use project prefix to find config files, but use provided or derived project name
            generator_config_path, zones_config_path = resolve_project_config_files(f"{args.project.upper()}_MCM01", paths['project_root'])
            print(f"Using project: {args.project.upper()}")
            print(f"  Generator config: {generator_config_path.name}")
            print(f"  Zones config: {zones_config_path.name}")
            print(f"  Project name: {project_name}")
        except (ValueError, FileNotFoundError) as e:
            print(f"Error: {e}")
            available_projects = get_available_projects(paths['project_root'])
            if available_projects:
                print(f"Available projects: {', '.join(available_projects)}")
            else:
                print("No projects found. Run with --list-projects to see details.")
            sys.exit(1)
    elif args.project_name:
        # Backward compatibility mode using --project-name only
        project_name = args.project_name
        # Try to auto-detect config files based on project name
        try:
            generator_config_path, zones_config_path = resolve_project_config_files(project_name, paths['project_root'])
            print(f"Auto-detected config files for {project_name}:")
            print(f"  Generator config: {generator_config_path.name}")
            print(f"  Zones config: {zones_config_path.name}")
        except (ValueError, FileNotFoundError):
            # Fall back to default config files
            print(f"Using default config files (project-specific configs not found for {project_name})")
            generator_config_path = None  # Will use default in run_routines_generator
    else:
        # No project specified - require at least one
        print("Error: Either --project or --project-name must be specified")
        available_projects = get_available_projects(paths['project_root'])
        if available_projects:
            print(f"Available projects: {', '.join(available_projects)}")
        sys.exit(1)
    
    # Validate excel-file is provided for actual processing (not just listing)
    if not args.excel_file:
        print("Error: --excel-file is required for processing")
        parser.print_help()
        sys.exit(1)
    
    # Normalize the Excel file path to handle Windows-style paths in WSL
    excel_file_path = Path(args.excel_file.replace('\\', '/'))
    if not excel_file_path.is_absolute():
        excel_file_path = paths['project_root'] / excel_file_path
    args.excel_file = excel_file_path
    
    # Setup enhanced logging
    from src.logging_config import setup_logging, get_logger
    ts = datetime.now().strftime('%Y%m%d_%H%M%S')
    log_dir = paths['project_root'] / 'logs'
    base = project_name or 'project'
    log_file = log_dir / f"workflow_{base}_{ts}.log" if args.verbose else None
    
    # Configure logging with location information
    setup_logging(
        level='DEBUG' if args.verbose else 'INFO',
        console_format='human',
        log_file=log_file,
        show_module=True,
        show_location=True,
        use_colors=True
    )
    
    logger = get_logger(__name__)
    logger.info("PLC Generation Workflow started", excel_file=str(args.excel_file), project_name=args.project_name)
    
    print("PLC Generation Workflow")
    
    # Step 1: Process raw Excel data
    print("Step 1: Data processing ...", end=" ")
    logger.info("Starting data processing step")
    try:
        ok = run_plc_data_generator(args.excel_file, paths, verbose=args.verbose, log_file=log_file)
        print("OK" if ok else "FAIL")
        if not ok:
            logger.error("Data processing failed")
            if not args.verbose:
                print("(details suppressed; re-run with --verbose)")
            sys.exit(1)
        logger.info("Data processing completed successfully")
    except Exception as e:
        logger.exception("Data processing step failed with exception", error=str(e))
        print("FAIL")
        sys.exit(1)
    
    # Step 2: Generate L5X programs (Routines Generator)
    print("Step 2: Routine generation ...", end=" ")
    logger.info("Starting routine generation step")
    try:
        ok = run_routines_generator(paths, project_name, args.ignore_estop1ok, args.safety_only, verbose=args.verbose, log_file=log_file, config_path=generator_config_path)
        print("OK" if ok else "FAIL")
        if not ok:
            logger.error("Routine generation failed")
            if not args.verbose:
                print("(details suppressed; re-run with --verbose)")
            sys.exit(1)
        logger.info("Routine generation completed successfully")
    except Exception as e:
        logger.exception("Routine generation step failed with exception", error=str(e))
        print("FAIL")
        sys.exit(1)

    # Step 3: Generate complete project L5X (IO Tree Generator)
    if args.safety_only:
        print("Step 3: IO tree generation ... SKIPPED")
    else:
        print("Step 3: IO tree generation ...", end=" ")
        # Determine boilerplate directory based on project
        boilerplate_dir = None
        if args.project:
            boilerplate_dir = resolve_boilerplate_directory(args.project.upper(), paths['io_tree_generator'])
        ok = run_io_tree_generator(paths, project_name, args.safety_only, verbose=args.verbose, log_file=log_file, boilerplate_dir=boilerplate_dir)
        print("OK" if ok else "FAIL")
        if not ok:
            if not args.verbose:
                print("(details suppressed; re-run with --verbose)")
            sys.exit(1)
    
    # Step 4: Compile L5X to ACD
    if args.safety_only:
        print("Step 4: Prepare compilation ... SKIPPED")
    else:
        print("Step 4: Prepare compilation ...", end=" ")
        ok = run_l5x_to_acd_compiler(paths, project_name, args.safety_only, verbose=args.verbose, log_file=log_file)
        print("OK" if ok else "FAIL")
        if not ok:
            if not args.verbose:
                print("(details suppressed; re-run with --verbose)")
            sys.exit(1)
    
    print("Workflow complete")
    if args.verbose and log_file is not None:
        print(f"Logs: {log_file}")
    if not args.safety_only and project_name:
        print(f"L5X: IO Tree Configuration Generator/generated_projects/{project_name}.L5X")

if __name__ == '__main__':
    main()