import os
import sys
import json
import subprocess
import time

# Force stdout to be UTF-8 for Windows terminals
if hasattr(sys.stdout, 'reconfigure'):
    sys.stdout.reconfigure(encoding='utf-8')

NOW = time.time()

def get_git_timestamp(filepath):
    # Normalize paths to use forward slashes for git compatibility
    git_path = filepath.replace('\\', '/')
    
    # 1. Check if the file exists on disk.
    if not os.path.exists(filepath):
        return 0
        
    # 2. Check for local modifications (staged or unstaged)
    try:
        status_proc = subprocess.run(
            ['git', 'status', '--porcelain', '--', git_path],
            capture_output=True, text=True, check=True
        )
        if status_proc.stdout.strip():
            # Locally modified files are considered "edited just now"
            return NOW
    except subprocess.SubprocessError:
        pass

    # 3. Get the last commit timestamp
    try:
        log_proc = subprocess.run(
            ['git', 'log', '-1', '--format=%ct', '--', git_path],
            capture_output=True, text=True, check=True
        )
        output = log_proc.stdout.strip()
        if output.isdigit():
            return int(output)
    except subprocess.SubprocessError:
        pass

    # 4. Fallback to OS modification time if file exists but has no git commits yet
    try:
        return os.path.getmtime(filepath)
    except OSError:
        return 0

def check_dirty():
    graph_path = os.path.join(os.path.dirname(__file__), 'graph.json')
    if not os.path.exists(graph_path):
        print(f"Error: graph.json not found at {graph_path}")
        sys.exit(1)
        
    with open(graph_path, 'r', encoding='utf-8') as f:
        graph = json.load(f)
        
    files_dict = graph.get('files', {})
    timestamps = {}
    
    # Calculate timestamps for all files
    for filepath in files_dict.keys():
        timestamps[filepath] = get_git_timestamp(filepath)
        
    dirty_files = {}
    
    # Check dependencies
    for filepath, info in files_dict.items():
        depends_on = info.get('depends_on', [])
        file_ts = timestamps[filepath]
        
        # If the file itself doesn't exist, it's not "dirty" in the sense of being out of date,
        # but it is missing. However, to keep it clean, if it doesn't exist, we can skip or flag it.
        # Let's say if a file exists, we check if its dependencies are newer.
        if file_ts == 0:
            # File is missing on disk
            continue
            
        file_dirty_deps = []
        for dep in depends_on:
            dep_ts = get_git_timestamp(dep)
            if dep_ts > file_ts:
                file_dirty_deps.append((dep, dep_ts, file_ts))
                
        if file_dirty_deps:
            dirty_files[filepath] = file_dirty_deps

    if dirty_files:
        print("\n[DIRTY] 发现以下下游文件落后于其上游依赖：")
        for filepath, deps in dirty_files.items():
            print(f"\n* {filepath}")
            for dep, dep_ts, file_ts in deps:
                dep_time = time.strftime('%Y-%m-%d %H:%M:%S', time.localtime(dep_ts))
                file_time = time.strftime('%Y-%m-%d %H:%M:%S', time.localtime(file_ts))
                print(f"  [依赖于] {dep}")
                print(f"  (上游最后提交: {dep_time} > 下游最后提交: {file_time})")
        print("\n请跟进并更新上述下游文件，重新 commit 即可刷新基线并自动消脏。\n")
        return False
    else:
        print("\n[CLEAN] 所有文件均已与依赖同步，无脏文件。\n")
        return True

if __name__ == '__main__':
    success = check_dirty()
    sys.exit(0 if success else 1)