import requests, sys, json, lzma, os, time
from bs4 import BeautifulSoup # pip install beautifulsoup4
from multiprocessing.pool import Pool

DOMAIN = "https://thichtruyen.vn"

t = open("cats.txt", "rt").read()
cats = []
for idx, x in enumerate(t.strip().split("\n\n")):
    u, t = x.split("\n")
    u = DOMAIN + u
    # print(u, t)
    cats.append({"title": t, "url": u, "_id": idx})


def crawl(cat):
    cat_id = cat["_id"]
    cat_url = cat["url"]
    cat_file = f"cats/{cat_id}.jsonl"

    crawled_pages = set()
    if os.path.exists(cat_file):
        with open(cat_file, "rt") as fin:
            for line in fin:
                crawled_pages.add(json.loads(line)["page"])

    for page in range(1, 9999):
        if page in crawled_pages: continue
        page_url = f"{cat_url}?page={page}"; print(page_url)
        r = requests.get(page_url)
        soup = BeautifulSoup(r.content, "html.parser")

        stories = []
        for item in soup.select("#tab-over .view-category-item-infor"):
            a = item.select("a")[0]
            url = DOMAIN + a["href"]
            title = a.select("h3")[0].text.strip()
            print(url, title)
            stories.append({
                "title": title,
                "url": url,
                "cat": cat["title"],
                "page": page,
            })

        with open(cat_file, "at") as fout:
            for story in stories:
                fout.write(json.dumps(story, ensure_ascii=False) + "\n")

        if len(stories) == 0: break
        time.sleep(1) # chậm thôi :)

# crawl(cats[0])
with Pool() as pool:
    for _ in pool.imap_unordered(crawl, cats):
        False