#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
兰升生物科技集团股份有限公司 - 公示公告
https://www.lansen-china.com/lsweb/#/News?tab=9

站点: Vue SPA (LSCMS), 数据全走 JSON API
列表: GET /api/channel?parentId=4 -> 频道列表 (id=9 = 公示公告, alias=notice)
      GET /api/article?channel=notice&channelId=9 -> 全量文章 JSON (无服务端分页,
      前端客户端 slice; 每次调用返回全部)
详情: GET /api/article/{id} -> JSON
      title / publishDate / text(HTML正文, 实体编码) / plainText / description
      fileList / downloads / imageList 附件
"""
import sys
import os
import re
import time
import json
import html as html_lib
import urllib.parse
import sqlite3
import urllib.request

try:
    from bs4 import BeautifulSoup
except ImportError:
    BeautifulSoup = None

# ---------------- config ----------------
BASE_URL = "https://www.lansen-china.com"
API_BASE = BASE_URL + "/api"
SITE_NAME = "兰升生物科技集团-公示公告"
GROUP_NAME = "河北"
SCRIPT_NAME = "crawl_lansen_gsgg.py"
DB_PATH = "/mnt/data/search.db"
CHANNEL_ALIAS = "notice"   # 公示公告
CHANNEL_ID = "9"

HEADERS = {
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0 Safari/537.36",
    "Accept": "application/json, text/plain, */*",
}

# crawler_lib 所在目录 (classify_industry 行业分类复用)
for _p in ("/root/gov_crawler", "/root/gov_crawler"):
    if os.path.isdir(_p) and _p not in sys.path:
        sys.path.insert(0, _p)

# default: daily incremental = 1 page
_MAX_PAGES = 1
for i, a in enumerate(sys.argv):
    if a.startswith("--pages="):
        try:
            _MAX_PAGES = int(a.split("=", 1)[1])
        except ValueError:
            pass
    elif a == "--pages" and i + 1 < len(sys.argv):
        try:
            _MAX_PAGES = int(sys.argv[i + 1])
        except ValueError:
            pass

_PAGES = _MAX_PAGES if _MAX_PAGES >= 1 else 1


# ---------------- helpers ----------------
def clean_title(t):
    """unescape entities + strip &middot;&nbsp; prefix + collapse whitespace."""
    if not t:
        return ""
    t = html_lib.unescape(t)
    t = re.sub(r"^\s*[·•]+\s*", "", t)
    t = t.replace("\xa0", " ").replace("\u200b", "").replace("\u200c", "").replace("\u200d", "").replace("\ufeff", "")
    t = re.sub(r"\s+", " ", t)
    return t.strip()


def api_get(path):
    """GET JSON API with retries."""
    url = API_BASE + path
    last = None
    for i in range(3):
        try:
            req = urllib.request.Request(url, headers=HEADERS)
            with urllib.request.urlopen(req, timeout=30) as r:
                return json.loads(r.read().decode("utf-8"))
        except Exception as e:
            last = e
            time.sleep(1.5 * (i + 1))
    raise last


def html_to_text(content, page_url):
    """Convert article HTML to stored format (same pipeline as gov scripts):
    - keep <table> HTML intact
    - \n\n between <p> blocks
    - attachment <a> links embedded with absolute URL
    """
    if not content:
        return "", 0, []
    content = re.sub(r"<!--.*?-->", "", content, flags=re.S)
    content = re.sub(r"<script[^>]*>.*?</script>", "", content, flags=re.S | re.I)
    content = re.sub(r"<style[^>]*>.*?</style>", "", content, flags=re.S | re.I)
    if BeautifulSoup:
        try:
            _soup = BeautifulSoup(content, "html.parser")
            for tag in _soup.find_all(True):
                for attr in ("style", "class", "lang", "dir", "align", "valign", "width", "height", "border", "cellpadding", "cellspacing"):
                    tag.attrs.pop(attr, None)
            content = str(_soup)
        except Exception:
            pass

    # 1) FIRST protect attachment <a> links
    content2 = content
    attachments = []
    link_protect = {}
    def _link_repl(m):
        href = m.group(1)
        inner = m.group(2)
        txt = re.sub(r"<[^>]+>", "", inner)
        txt = html_lib.unescape(txt).strip()
        txt = re.sub(r"^附件[:：]\s*", "", txt).strip()
        if not txt:
            txt = os.path.basename(href.split("?")[0]) or "附件"
        abs_url = urllib.parse.urljoin(page_url, html_lib.unescape(href))
        attachments.append((abs_url, txt))
        key = f"__LINK__{len(link_protect)}__"
        link_protect[key] = f'<a href="{abs_url}" target="_blank">{txt}</a>'
        return key
    content2 = re.sub(r'<a\s[^>]*href="([^"]+)"[^>]*>(.*?)</a>', _link_repl, content2, flags=re.S | re.I)

    # 2) THEN protect tables
    table_protect = []
    def _tbl_repl(m):
        table_protect.append(m.group(0))
        return f"__TBL__{len(table_protect)-1}__"
    content2 = re.sub(r"<table[^>]*>.*?</table>", _tbl_repl, content2, flags=re.S | re.I)
    has_table = 1 if re.search(r"<table[^>]*>", content, re.I) else 0

    content2 = re.sub(r"</p>", "</p>\n\n", content2, flags=re.I)
    content2 = re.sub(r"<br\s*/?>", "\n", content2, flags=re.I)
    content2 = re.sub(r"</?(?:span|font|o:p|st1?:[a-z]+)\b[^>]*>", "", content2, flags=re.I)

    parts = []
    for block in content2.split("\n\n"):
        b = block.strip()
        if not b:
            continue
        plain = re.sub(r"<[^>]+>", "", b)
        plain = html_lib.unescape(plain)
        plain = plain.replace("\xa0", " ").replace("&nbsp;", " ")
        plain = re.sub(r"[\s\u200b\u200c\u200d\ufeff]+", "", plain)
        if not plain:
            continue
        for k, v in link_protect.items():
            b = b.replace(k, v)
        for i, tbl in enumerate(table_protect):
            b = b.replace(f"__TBL__{i}__", tbl)
        parts.append(b.strip())

    out = "\n\n".join(parts)
    for k, v in link_protect.items():
        out = out.replace(k, v)
    for i, tbl in enumerate(table_protect):
        out = out.replace(f"__TBL__{i}__", tbl)

    out = re.sub(r"__LINK__\d+__", "", out)
    out = re.sub(r"__TBL__\d+__", "", out)

    out = html_lib.unescape(out)
    out = re.sub(r"[ \t]+\n", "\n", out)
    out = re.sub(r"\n{3,}", "\n\n", out)
    out = re.sub(r"[ \t]{2,}", " ", out)

    # dedupe paragraphs
    seen = set()
    final = []
    for p in out.split("\n\n"):
        key = re.sub(r"\s+", "", re.sub(r"<[^>]+>", "", p))
        if not key:
            continue
        if key in seen:
            continue
        seen.add(key)
        final.append(p)
    out = "\n\n".join(final).strip()

    return out, has_table, attachments


def store_record(conn, url, title, content, date, has_table):
    cur = conn.execute(
        "SELECT id FROM gov_raw WHERE page_url=?",
        (url,),
    )
    if cur.fetchone():
        return False
    try:
        industry = "other"
        try:
            from crawler_lib import classify_industry
            industry = classify_industry(title)
        except Exception:
            pass
        cur = conn.execute(
            "INSERT OR IGNORE INTO gov_raw "
            "(page_url, source_url, title, content, publish_date, site_name, group_name, script_name, has_table, industry) "
            "VALUES (?,?,?,?,?,?,?,?,?,?)",
            (url, url, title, content, date, SITE_NAME, GROUP_NAME, SCRIPT_NAME, has_table, industry),
        )
        if cur.rowcount > 0:
            summary = re.sub(r"<[^>]+>", "", content)
            summary = html_lib.unescape(summary)
            summary = re.sub(r"\s+", " ", summary).strip()[:200]
            conn.execute(
                "INSERT OR IGNORE INTO gov_search (rowid, title, site_name, summary) VALUES (?,?,?,?)",
                (cur.lastrowid, title, SITE_NAME, summary),
            )
            return True
        return False
    except sqlite3.OperationalError as e:
        if "locked" in str(e):
            time.sleep(3)
            return store_record(conn, url, title, content, date, has_table)
        raise


def main():
    conn = sqlite3.connect(DB_PATH, timeout=60)
    conn.execute("PRAGMA busy_timeout=60000")
    conn.execute("PRAGMA journal_mode=WAL")

    new_count = 0
    skip_count = 0

    # 列表: 全量文章 (无服务端分页)
    try:
        articles = api_get(f"/article?channel={CHANNEL_ALIAS}&channelId={CHANNEL_ID}")
    except Exception as e:
        print(f"新增: 0 (list error: {e})")
        return
    if not isinstance(articles, list) or not articles:
        print("新增: 0 (empty list)")
        return

    # 排序: publishDate 降序, 取前 N 页等价于前 N*pageSize 条 (pageSize=10)
    articles.sort(key=lambda a: a.get("publishDate", ""), reverse=True)
    limit = _PAGES * 10
    for art in articles[:limit]:
        art_id = art.get("id")
        title = clean_title(art.get("title", ""))
        pub_date = (art.get("publishDate") or "")[:10]
        url = f"{BASE_URL}/article/{art_id}"
        # 详情
        try:
            detail = api_get(f"/article/{art_id}")
        except Exception as e:
            print(f"  [ERR] {art_id}: {e}")
            skip_count += 1
            continue
        content_html = detail.get("text") or ""
        # 附件 fileList
        file_list = detail.get("fileList") or []
        if file_list:
            for f in file_list:
                fname = f.get("name") or f.get("fileName") or ""
                furl = f.get("url") or f.get("fileUrl") or ""
                if furl:
                    if not furl.startswith(("http://", "https://")):
                        furl = urllib.parse.urljoin(BASE_URL, furl)
                    content_html += f'\n<p><a href="{furl}" target="_blank">{html_lib.escape(fname or os.path.basename(furl))}</a></p>'
        content, has_table, attachments = html_to_text(content_html, url)
        if not content:
            # 兜底 description
            desc = detail.get("description") or ""
            if desc:
                content = "<p>" + html_lib.escape(desc).replace("\n", "</p>\n<p>") + "</p>"
                content, has_table, attachments = html_to_text(content, url)
        if not content:
            skip_count += 1
            continue
        if store_record(conn, url, title, content, pub_date, has_table):
            new_count += 1
        else:
            skip_count += 1

    conn.commit()
    conn.close()
    print(f"新增: {new_count}")


if __name__ == "__main__":
    main()