#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
全站文案深度分析与 AI 味/啰嗦度检测工具
"""
import os
import re

BASE_DIR = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))

TEMPLATES = [
    'addons/official/view/web/official_home/index.html',
    'addons/official/view/web/official_home/index_e.html',
    'addons/official/view/web/official_home/single.html',
    'addons/official/view/web/official_home/single_e.html',
    'addons/official/view/web/official_product/service_page.html',
    'addons/official/view/web/official_product/service_software.html',
    'addons/official/view/web/official_product/service_app.html',
    'addons/official/view/web/official_product/service_agent.html',
    'addons/official/view/web/official_product/product_page.html',
    'addons/official/view/web/official_product/list_e.html',
    'addons/official/view/web/official_case_show/list.html',
    'addons/official/view/web/official_case_show/list_e.html',
    'addons/official/view/web/official_online_message/list.html',
    'addons/official/view/web/official_online_message/list_e.html',
    'addons/official/view/web/official_common/pc/footer.html',
    'addons/official/view/web/official_common/pc/nav.html',
    'addons/official/view/web/official_common/mobile/footer.html',
    'addons/official/view/web/official_common/mobile/nav.html'
]

# AI 味黑名单词库（空洞、宏大叙事、假大空口号、泛化套话、阿里味/公关腔）
AI_FLAVOR_TERMS = [
    '赋能', '铸造数字生产力', '新纪元', '新篇章', '新范式', '乘风破浪',
    '全方位', '多维度', '深耕', '重器', '基石', '筑牢', '铸牢',
    '大放异彩', '跃迁', '引领', '颠覆', '无可争议', '无懈可击',
    '滴水不漏', '毫无保留', '无论您是', '助力企业全渠道数字化运营',
    '加速企业数字化转型', '助力移动产品极速上架', '准备打造您的专属',
    '全天候守护', '赋能培训', '即开即用，伴随业务快速启航',
    'AI 智能安全库存', 'AI智能安全库存', '开启新一代数字化与智能化升级',
    '工匠精神打造', '深厚一线大型系统经验', '深厚工程积累'
]

# 冗长啰嗦句式模式
VERBOSE_PATTERNS = [
    r'不仅.*?而且.*?',
    r'为了能够更好地.*?',
    r'致力于为.*?提供.*?',
    r'旨在为.*?打造.*?',
]

def analyze():
    print("=================== 全站文案深度分析报告 ===================")
    total_issues = 0
    for rel_path in TEMPLATES:
        abs_path = os.path.join(BASE_DIR, rel_path.replace('/', os.sep))
        if not os.path.exists(abs_path):
            continue
        with open(abs_path, 'r', encoding='utf-8') as f:
            lines = f.readlines()
        
        file_issues = []
        for line_no, line in enumerate(lines, 1):
            text = line.strip()
            # 过滤注释与脚本
            if text.startswith('<!--') and text.endswith('-->'):
                # 检查注释里是否也有未清理的旧口号
                pass
            for term in AI_FLAVOR_TERMS:
                if term in text:
                    file_issues.append((line_no, f"AI味词汇: '{term}'", text))
            
            for vp in VERBOSE_PATTERNS:
                m = re.search(vp, text)
                if m:
                    file_issues.append((line_no, f"啰嗦句式: '{m.group(0)}'", text))
        
        if file_issues:
            print(f"\n[{rel_path}] 发现 {len(file_issues)} 处潜在问题:")
            for lno, itype, ltext in file_issues:
                # 截断过长文字
                display_text = ltext if len(ltext) <= 80 else ltext[:80] + '...'
                print(f"  行 {lno:4d} | {itype:25s} | {display_text}")
            total_issues += len(file_issues)
    
    print(f"\n=================== 扫描完毕：共检出 {total_issues} 处待优化文案 ===================")

if __name__ == '__main__':
    analyze()