149 lines
5.6 KiB
Python
149 lines
5.6 KiB
Python
import os
|
||
import requests
|
||
import json
|
||
|
||
try:
|
||
import markdown
|
||
except ImportError:
|
||
raise ImportError("错误: 'markdown' 库未安装。请运行: pip install markdown")
|
||
|
||
# --- 1. 辅助函数与模块常量加载 ---
|
||
|
||
def _load_asset_file(filename: str) -> str:
|
||
"""从 assets 文件夹加载一个文本文件的内容。"""
|
||
# 获取当前文件 (analyzer.py) 的绝对路径
|
||
current_dir = os.path.dirname(os.path.abspath(__file__))
|
||
# 构造 assets 文件的完整路径
|
||
file_path = os.path.join(current_dir, 'assets', filename)
|
||
try:
|
||
with open(file_path, 'r', encoding='utf-8') as f:
|
||
return f.read()
|
||
except FileNotFoundError:
|
||
print(f"CRITICAL ERROR: Asset file not found at '{file_path}'")
|
||
raise
|
||
except Exception as e:
|
||
print(f"CRITICAL ERROR: Failed to read asset file '{file_path}'. Error: {e}")
|
||
raise
|
||
|
||
# 在模块加载时，只执行一次文件读取操作
|
||
try:
|
||
FULL_SYSTEM_PROMPT = _load_asset_file('main_prompt.md')
|
||
HTML_TEMPLATE = _load_asset_file('report_template.html')
|
||
except Exception:
|
||
# 如果加载失败，程序应中止
|
||
exit(1)
|
||
|
||
|
||
# --- 2. 内部功能函数 ---
|
||
|
||
def _convert_markdown_to_html(markdown_text: str, title: str) -> str:
|
||
"""将Markdown文本转换为带样式的完整HTML页面。"""
|
||
html_body = markdown.markdown(markdown_text, extensions=['fenced_code', 'tables'])
|
||
html_body_styled = html_body.replace('<table>', '<table class="table table-striped table-bordered table-hover">')
|
||
|
||
# 使用从文件加载的模板，并替换占位符
|
||
return HTML_TEMPLATE.format(title=title, html_body_styled=html_body_styled)
|
||
|
||
def _analyze_chat_file(filepath: str, api_key: str, model_name: str, site_url: str, site_name: str) -> str | None:
|
||
"""读取单个聊天文件并调用OpenRouter API进行分析。"""
|
||
print(f"-> 正在分析新文件: {os.path.basename(filepath)}...")
|
||
|
||
try:
|
||
with open(filepath, 'r', encoding='utf-8') as file:
|
||
chat_content = file.read()
|
||
except FileNotFoundError:
|
||
print(f" 错误: 文件未找到 {filepath}")
|
||
return None
|
||
|
||
if len(chat_content.strip()) < 50:
|
||
print(f" 警告: 文件 {os.path.basename(filepath)} 内容过短，已跳过。")
|
||
return "NO_CONTENT"
|
||
|
||
api_payload = {
|
||
"model": model_name,
|
||
"messages": [
|
||
{"role": "system", "content": FULL_SYSTEM_PROMPT},
|
||
{"role": "user", "content": f"请根据系统提示中的需求，分析以下聊天记录：\\n\\n---\\n{chat_content}\\n---"}
|
||
],
|
||
"temperature": 0.7, "top_p": 1, "max_tokens": 8192
|
||
}
|
||
|
||
try:
|
||
response = requests.post(
|
||
url="https://openrouter.ai/api/v1/chat/completions",
|
||
headers={
|
||
"Authorization": f"Bearer {api_key}",
|
||
"HTTP-Referer": site_url,
|
||
"X-Title": site_name,
|
||
},
|
||
data=json.dumps(api_payload),
|
||
timeout=180
|
||
)
|
||
|
||
if response.status_code == 200:
|
||
print(f" 成功接收到来自 {model_name} 的分析结果。")
|
||
analysis_result = response.json()
|
||
return analysis_result['choices'][0]['message']['content']
|
||
else:
|
||
print(f" API请求失败，状态码: {response.status_code}, 错误: {response.text}")
|
||
return None
|
||
except requests.exceptions.RequestException as e:
|
||
print(f" API请求异常: {e}")
|
||
return None
|
||
|
||
# --- 3. 主调用函数 (无变动) ---
|
||
|
||
def run_analysis_on_new_files(input_folder: str, output_folder: str, config: dict) -> int:
|
||
"""
|
||
扫描输入文件夹，对新出现的聊天记录进行分析并生成HTML报告。
|
||
"""
|
||
if not os.path.exists(input_folder):
|
||
print(f"错误：找不到输入文件夹 '{input_folder}'。")
|
||
return 0
|
||
|
||
if not os.path.exists(output_folder):
|
||
os.makedirs(output_folder)
|
||
print(f"已创建输出文件夹: {output_folder}")
|
||
|
||
newly_analyzed_count = 0
|
||
all_log_files = sorted([f for f in os.listdir(input_folder) if f.endswith(".txt")])
|
||
|
||
if not all_log_files:
|
||
return 0
|
||
|
||
latest_log_filename = all_log_files[-1]
|
||
|
||
for filename in all_log_files:
|
||
base_name = os.path.splitext(filename)[0]
|
||
input_path = os.path.join(input_folder, filename)
|
||
output_filename = f"{base_name}_analysis.html"
|
||
output_path = os.path.join(output_folder, output_filename)
|
||
|
||
is_latest_file = (filename == latest_log_filename)
|
||
|
||
if os.path.exists(output_path) and not is_latest_file:
|
||
continue
|
||
|
||
if os.path.exists(output_path) and is_latest_file:
|
||
print(f"-> 发现最新日期的报告已存在，将强制重新分析 '{filename}'...")
|
||
|
||
analysis_markdown = _analyze_chat_file(
|
||
filepath=input_path,
|
||
api_key=config.get("OPENROUTER_API_KEY"),
|
||
model_name=config.get("MODEL_NAME"),
|
||
site_url=config.get("YOUR_SITE_URL"),
|
||
site_name=config.get("YOUR_SITE_NAME")
|
||
)
|
||
|
||
if analysis_markdown and analysis_markdown != "NO_CONTENT":
|
||
report_title = f"玩家社群分析报告 - {base_name}"
|
||
html_content = _convert_markdown_to_html(analysis_markdown, title=report_title)
|
||
|
||
with open(output_path, 'w', encoding='utf-8') as f:
|
||
f.write(html_content)
|
||
|
||
print(f" ✔ HTML分析报告已成功保存至: {output_path}\\n")
|
||
newly_analyzed_count += 1
|
||
|
||
return newly_analyzed_count
|