@whyshu/argent-engine 0.0.0-stage → 0.6.0

package/engine/scripts/wenyao_login.py

@@ -0,0 +1,145 @@
+#!/usr/bin/env python3
+"""wenyao_login.py — 上海文鳐门户「人工过验证 + 保存登录态 + 抓入口」通用脚本（2026-09-10）
+
+为什么需要人工：门户登录使用**阿里云行为验证码（滑块）**。合成鼠标事件过不了（实测 2 次失败），
+且反复尝试会触发风控 ⇒ 采用业界通行的**人工过一次 + 复用登录态**：
+  1. 本脚本打开可见 Chrome（专用持久 profile），自动预填手机号/密码；
+  2. **你**把滑块拖到最右边并点「登录」（脚本等 `--wait` 秒，默认 1800）；
+  3. 脚本检测登录成功 → 保存 `storage_state`（600）+ 持久 profile（之后自动化复用，无需再验证）；
+  4. 顺带抓取登录后的**菜单/入口链接**，尤其是「模板/下载」类入口（用于找《上海文鳐发票模版》）。
+
+只读侦察：**不导入订单、不导出文件、不改对方任何业务数据**。
+
+用法：
+  ~/.dsh/argent/venv/bin/python3 scripts/wenyao_login.py                 # 等 30 分钟
+  ~/.dsh/argent/venv/bin/python3 scripts/wenyao_login.py --wait 600
+  ~/.dsh/argent/venv/bin/python3 scripts/wenyao_login.py --check         # 只检查登录态是否还有效
+凭据：`~/.dsh/argent/wenyao.env`（WENYAO_URL/WENYAO_PHONE/WENYAO_PASSWORD，600 权限，不进仓库）
+产物：`~/.dsh/argent/wenyao-storage.json`(600) / `~/.dsh/argent/wenyao-probe/*.png|json`
+"""
+import argparse
+import json
+import os
+import sys
+import time
+from pathlib import Path
+from playwright.sync_api import sync_playwright
+
+HOME = Path.home()
+# 数据目录：**唯一判据**在 `scripts/lib/argent_datadir.py`（三级链，与 JS 内核同语义）
+sys.path.insert(0, str(Path(__file__).resolve().parent / 'lib'))
+from argent_datadir import resolve_data_dir  # noqa: E402  —— 勿在此复制判据
+DATA = Path(resolve_data_dir()['dir'])
+OUT = DATA / "wenyao-probe"
+PROFILE = str(DATA / "wenyao-profile")
+CHROME = os.environ.get("WENYAO_CHROME", "/Applications/Google Chrome.app/Contents/MacOS/Google Chrome")
+STORAGE = DATA / "wenyao-storage.json"
+
+
+def load_env():
+    env = {}
+    f = DATA / "wenyao.env"
+    if f.exists():
+        for line in f.read_text(encoding="utf-8").splitlines():
+            line = line.strip()
+            if line and not line.startswith("#") and "=" in line:
+                k, v = line.split("=", 1)
+                env[k.strip()] = v.strip()
+    env.setdefault("WENYAO_URL", "https://my.ffl-tech.com")
+    return env
+
+
+def body(page):
+    try:
+        return page.evaluate("() => (document.body.innerText||'').replace(/\\s+/g,' ')")
+    except Exception:
+        return ""
+
+
+def logged_in(page, login_url):
+    t = body(page)
+    if "请按住滑块" in t or "拖动到最右" in t:
+        return False
+    if "账号密码登录" in t and "短信验证登录" in t:
+        return False
+    return page.url.rstrip("/") != login_url.rstrip("/") or any(
+        k in t for k in ("退出登录", "个人中心", "导入订单", "订单管理", "我的订单", "头程"))
+
+
+def capture(page, tag):
+    OUT.mkdir(parents=True, exist_ok=True)
+    try:
+        page.screenshot(path=str(OUT / f"{tag}.png"), full_page=True)
+    except Exception:
+        pass
+    t = body(page)
+    links = []
+    try:
+        links = page.evaluate("""() => Array.from(document.querySelectorAll('a,button,li,span,div'))
+            .map(e => ({t:(e.innerText||'').trim(), h:e.getAttribute && (e.getAttribute('href')||'')}))
+            .filter(x => x.t && x.t.length < 30 && !/^[\\d\\s]+$/.test(x.t))
+            .filter((v,i,a) => a.findIndex(y => y.t===v.t)===i).slice(0,150)""")
+    except Exception:
+        pass
+    (OUT / f"{tag}.json").write_text(json.dumps({"url": page.url, "title": page.title(), "text": t[:4000], "entries": links},
+                                               ensure_ascii=False, indent=1), encoding="utf-8")
+    interesting = [x for x in links if any(k in (x.get("t") or "") for k in ("模板", "模版", "下载", "导入", "导出", "订单", "发票"))]
+    return {"tag": tag, "url": page.url, "entries": len(links), "interesting": interesting[:20], "text_head": t[:200]}
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--wait", type=int, default=1800, help="人工过验证的等待秒数（默认 1800）")
+    ap.add_argument("--check", action="store_true", help="只检查登录态是否仍有效")
+    a = ap.parse_args()
+    env = load_env()
+    url = env["WENYAO_URL"]
+    with sync_playwright() as p:
+        ctx = p.chromium.launch_persistent_context(
+            user_data_dir=PROFILE, executable_path=CHROME, headless=False,
+            viewport={"width": 1440, "height": 900}, args=["--no-first-run", "--no-default-browser-check"])
+        page = ctx.pages[0] if ctx.pages else ctx.new_page()
+        page.goto(url, wait_until="domcontentloaded", timeout=45000)
+        page.wait_for_timeout(4000)
+        if logged_in(page, url):
+            print(json.dumps({"stage": "already-logged-in", **capture(page, "40-check")}, ensure_ascii=False))
+            ctx.storage_state(path=str(STORAGE)); os.chmod(STORAGE, 0o600)
+            ctx.close()
+            return 0
+        if a.check:
+            print(json.dumps({"stage": "not-logged-in", "url": page.url}, ensure_ascii=False))
+            ctx.close()
+            return 1
+        for el in page.query_selector_all("input"):
+            try:
+                if not el.is_visible():
+                    continue
+                t = (el.get_attribute("type") or "text").lower()
+                if t == "password":
+                    el.fill(env.get("WENYAO_PASSWORD", ""))
+                elif t in ("text", "tel", "number", ""):
+                    el.fill(env.get("WENYAO_PHONE", ""))
+            except Exception:
+                continue
+        print(json.dumps({"stage": "wait-human", "seconds": a.wait,
+                          "todo": "在打开的 Chrome 窗口里：把滑块拖到最右边 → 点「登录」"}, ensure_ascii=False), flush=True)
+        deadline = time.time() + a.wait
+        ok = False
+        while time.time() < deadline:
+            page.wait_for_timeout(3000)
+            if logged_in(page, url):
+                ok = True
+                break
+        print(json.dumps({"stage": "human-result", "logged_in": ok}, ensure_ascii=False), flush=True)
+        info = capture(page, "41-after-login" if ok else "41-still-login")
+        print(json.dumps({"stage": "capture", **info}, ensure_ascii=False), flush=True)
+        if ok:
+            ctx.storage_state(path=str(STORAGE)); os.chmod(STORAGE, 0o600)
+            print(json.dumps({"stage": "storage-saved", "path": str(STORAGE)}, ensure_ascii=False), flush=True)
+        page.wait_for_timeout(1000)
+        ctx.close()
+    return 0 if ok else 1
+
+
+if __name__ == "__main__":
+    sys.exit(main())