import os
import glob
import time
import threading
from concurrent.futures import ThreadPoolExecutor, as_completed
from openai import OpenAI

api_key = input("Enter SiliconFlow API Key (sk-xxxx): ").strip()

client = OpenAI(
    api_key=api_key,
    base_url="https://api.siliconflow.cn/v1"
)

MODEL_NAME = "Qwen/Qwen2.5-72B-Instruct"
book_dir = "./docs"
output_dir = "./life-skill"
os.makedirs(os.path.join(output_dir, "chapters"), exist_ok=True)

md_files = sorted(glob.glob(os.path.join(book_dir, "**", "*.md"), recursive=True))
if not md_files:
    md_files = sorted(glob.glob("./*.md"))

total = len(md_files)
print(f"Detected {total} files to process...")

lock = threading.Lock()
finished = 0


def is_rate_limited(e):
    status = getattr(e, "status_code", None)
    if status is None:
        status = getattr(e, "status", None)
    if status is None and getattr(e, "response", None) is not None:
        status = getattr(e.response, "status_code", None)
    code = getattr(e, "code", None)
    if code is None:
        code = getattr(e, "error", None)
        code = getattr(code, "code", None) if code is not None else None
    msg = str(getattr(e, "message", "") or getattr(e, "error", "") or e)
    lowered = msg.lower()
    if status not in (None, 200):
        if int(status) == 429:
            return True
    if code is not None and "rate_limit" in str(code).lower():
        return True
    if "429" in msg:
        return True
    if "rate limit" in lowered or "ratelimit" in lowered:
        return True
    return False


def create_completion_with_retry(prompt, max_retries=5):
    delay = 3
    for attempt in range(max_retries + 1):
        try:
            resp = client.chat.completions.create(
                model=MODEL_NAME,
                messages=[{"role": "user", "content": prompt}],
                temperature=0.3
            )
            return resp.choices[0].message.content
        except Exception as e:
            if attempt >= max_retries or not is_rate_limited(e):
                raise
            with lock:
                print(
                    f"Rate limited, retrying in {delay}s "
                    f"(attempt {attempt + 1}/{max_retries}) ..."
                )
            time.sleep(delay)
            delay = delay * 2
    raise RuntimeError("unreachable")


def process_chapter(file_path):
    global finished
    file_name = os.path.basename(file_path)
    if file_name in ["README.md", "SUMMARY.md"]:
        return None
    out_file = os.path.join(output_dir, "chapters", file_name)
    if os.path.exists(out_file) and os.path.getsize(out_file) > 0:
        with lock:
            finished += 1
            print(f"[{finished}/{total}] Skipped (already exists): {file_name}")
        return out_file
    try:
        with open(file_path, "r", encoding="utf-8") as f:
            content = f.read()
    except Exception:
        return None

    prompt = (
        "你是一个顶级的知识架构工程师。请将以下《高性价比人生指南》章节内容精炼为适合 AI Agent 调用的技能参考模块。\n"
        "要求：\n"
        "1. 提取出核心原则、硬核决策卡尺、避免踩坑清单、核心执行步骤。\n"
        "2. 保留具体的量化标准（如成本、具体法条或医保操作）。\n"
        "3. 语言精炼，按 Markdown 输出，去掉冗余闲聊。\n\n"
        f"章节原始内容：\n{content[:15000]}"
    )

    try:
        result_text = create_completion_with_retry(prompt)
        with open(out_file, "w", encoding="utf-8") as f:
            f.write(result_text)
        time.sleep(0.5)
        with lock:
            finished += 1
            print(f"[{finished}/{total}] Finished: {file_name}")
        return out_file
    except Exception as e:
        with lock:
            print(f"Error processing {file_name}: {e}")
        return None


completed_files = []
with ThreadPoolExecutor(max_workers=2) as executor:
    futures = {executor.submit(process_chapter, fp): fp for fp in md_files}
    for future in as_completed(futures):
        result = future.result()
        if result:
            completed_files.append(os.path.basename(result))

print("Generating master index SKILL.md ...")
summaries = []
if os.path.isdir(os.path.join(output_dir, "chapters")):
    for file_name in sorted(os.listdir(os.path.join(output_dir, "chapters"))):
        chap_path = os.path.join(output_dir, "chapters", file_name)
        if os.path.isfile(chap_path) and os.path.getsize(chap_path) > 0:
            summaries.append(f"- [{file_name}](chapters/{file_name})")

skill_content = (
    "---\n"
    "name: high-cost-performance-life-guide\n"
    "description: Evidence-based high-ROI life decision rules covering health, law, finance, career, and housing\n"
    "---\n\n"
    "# 高性价比人生指南技能包\n\n"
    "模块工序索引：\n\n" + "\n".join(summaries)
)

with open(os.path.join(output_dir, "SKILL.md"), "w", encoding="utf-8") as f:
    f.write(skill_content)

print("All tasks finished successfully! Output saved in ./life-skill")