eh_logic.py

import os
import shutil
import subprocess
import gradio as gr
import json
from utils import BASE_DIR, PDF_FINAL_DIR, sanitize_filename, manual_merge_pdf

# 统一的 User-Agent
USER_AGENT = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36"

def extract_title_from_json(json_data):
    """
    辅助函数：从复杂的 JSON 结构（可能是列表嵌套）中递归查找 title
    """
    if isinstance(json_data, dict):
        # 优先找 title，没有找 gallery_title
        if 'title' in json_data: return json_data['title']
        if 'gallery_title' in json_data: return json_data['gallery_title']
        # 如果当前层没有，递归找 value
        for v in json_data.values():
            res = extract_title_from_json(v)
            if res: return res
    elif isinstance(json_data, list):
        # 如果是列表，遍历每个元素找
        for v in json_data:
            res = extract_title_from_json(v)
            if res: return res
    return None

def get_gallery_title(eh_url, cookie_file_path):
    """
    获取本子标题 (修复版：支持多行 JSON 解析)
    """
    try:
        cmd = [
            "gallery-dl",
            "--cookies", cookie_file_path,
            "--user-agent", USER_AGENT,
            "--dump-json",
            "--range", "1",
            eh_url
        ]
        
        result = subprocess.run(cmd, capture_output=True, text=True, encoding='utf-8', errors='ignore')
        
        if result.returncode == 0:
            # 🔴 关键修复：直接解析整个 stdout 字符串，而不是一行行读
            try:
                data = json.loads(result.stdout)
                title = extract_title_from_json(data)
                if title: return str(title)
            except json.JSONDecodeError:
                # 如果整体解析失败，再尝试处理混合数据流的情况（虽然现在应该不需要了）
                pass
                
    except Exception as e:
        print(f"[-] 获取标题失败: {e}")
    
    return None

def debug_eh_info(eh_url, cookies_str):
    """
    调试工具函数 (修复版)
    """
    debug_report = []
    debug_report.append(f"🔍 [Debug] 开始分析链接: {eh_url}")
    
    eh_base_dir = os.path.join(BASE_DIR, "eh_debug_temp")
    os.makedirs(eh_base_dir, exist_ok=True)
    
    cookie_file_path = os.path.join(BASE_DIR, "eh_cookies.txt")
    with open(cookie_file_path, "w", encoding="utf-8") as f:
        f.write(cookies_str)
        
    debug_report.append(f"🍪 [Debug] Cookies 已写入")

    cmd = [
        "gallery-dl",
        "--cookies", cookie_file_path,
        "--user-agent", USER_AGENT,
        "--dump-json",
        "--range", "1",
        "--verbose",
        eh_url
    ]
    
    try:
        process = subprocess.run(
            cmd,
            capture_output=True,
            text=True,
            encoding='utf-8',
            errors='ignore'
        )
        
        debug_report.append(f"🏷️ [Return Code]: {process.returncode}")
        
        # 尝试解析
        debug_report.append("\n🧐 [Debug] 尝试解析 JSON:")
        try:
            # 🔴 关键修复：和上面一样，整体解析
            data = json.loads(process.stdout)
            title = extract_title_from_json(data)
            if title:
                debug_report.append(f"✅ 成功解析! 标题是: {title}")
            else:
                debug_report.append("⚠️ JSON 有效，但没找到 title 字段。")
        except Exception as e:
             debug_report.append(f"❌ JSON 解析失败: {e}")
             debug_report.append("可能是返回的数据不是纯 JSON，以下是原始输出:")
             debug_report.append(process.stdout[:1000] + "...")

    except Exception as e:
        debug_report.append(f"💥 执行出错: {e}")

    return "\n".join(debug_report)

def run_eh_download(eh_url, cookies_str, quality, progress=gr.Progress()):
    """
    E-Hentai 下载主逻辑 (标题修复版)
    """
    print(f"[-] Debug:收到任务 URL: {eh_url}, 画质: {quality}")

    if not eh_url:
        yield None, "❌ 请输入画廊链接", "参数缺失"
        return
    
    progress(0, desc="正在初始化...")
    
    eh_base_dir = os.path.join(BASE_DIR, "eh_temp")
    if os.path.exists(eh_base_dir):
        try: shutil.rmtree(eh_base_dir)
        except: pass
    os.makedirs(eh_base_dir, exist_ok=True)
    
    # 1. 写入 Cookies
    cookie_file_path = os.path.join(BASE_DIR, "eh_cookies.txt")
    with open(cookie_file_path, "w", encoding="utf-8") as f:
        f.write(cookies_str)

    # 2. 获取标题
    progress(0.1, desc="正在解析标题...")
    real_title = get_gallery_title(eh_url, cookie_file_path)
    
    if real_title:
        clean_title = sanitize_filename(real_title)
        yield None, f"📖 识别到标题: {real_title}", "标题获取成功"
    else:
        clean_title = "Unknown_Gallery"
        yield None, "⚠️ 标题解析失败，将使用默认名称", "标题未知"

    # 3. 开始下载
    cmd = [
        "gallery-dl",
        "--verbose", 
        "--user-agent", USER_AGENT,
        "--directory", eh_base_dir,
        "--cookies", cookie_file_path,
        eh_url
    ]

    yield None, f"🚀 正在启动下载...", "启动中"
    progress(0.2, desc=f"正在下载: {clean_title[:15]}...")

    try:
        process = subprocess.Popen(
            cmd, 
            stdout=subprocess.PIPE, 
            stderr=subprocess.STDOUT,
            text=True,
            bufsize=1 
        )
        
        log_buffer = [] 
        img_count = 0
        
        for line in process.stdout:
            line = line.strip()
            if not line: continue
            
            log_buffer.append(line)
            display_text = "\n".join(log_buffer[-20:]) 
            
            if "#" in line and "http" in line:
                img_count += 1
                progress(None, desc=f"下载中 ({img_count}张)...")
            
            if "403 Forbidden" in line:
                 progress(None, desc="❌ 403 拒绝访问")

            yield None, display_text, f"已下载: {img_count}"
        
        process.wait()

        if process.returncode != 0:
            yield None, f"❌ 下载失败 (Code {process.returncode})", "失败"
            return

    except Exception as e:
        yield None, f"❌ 调用出错: {e}", "系统错误"
        return

    # 4. 寻找图片
    progress(0.9, desc="打包中...")
    target_img_dir = None
    for root, dirs, files in os.walk(eh_base_dir):
        if any(f.lower().endswith(('.jpg', '.png', '.jpeg', '.webp')) for f in files):
            target_img_dir = root
            break
    
    if not target_img_dir:
        yield None, "❌ 未找到图片", "空目录"
        return

    # 5. 合并 PDF
    if len(clean_title) > 100: clean_title = clean_title[:100]
    quality_tag = "Original" if quality >= 100 else f"Q{quality}"
    final_pdf_name = f"[EH][{quality_tag}] {clean_title}.pdf"
    final_pdf_path = os.path.join(PDF_FINAL_DIR, final_pdf_name)

    yield None, f"🔨 正在生成: {final_pdf_name}...", "合并中"
    
    try:
        manual_merge_pdf(target_img_dir, final_pdf_path, quality=quality)
    except Exception as e:
        yield None, f"❌ 合并失败: {e}", "合并失败"
        return

    progress(1.0, desc="完成！")
    yield [final_pdf_path], f"✅ 处理完成！\n文件名: {final_pdf_name}", "完成"