cfc/运营文案/小红书发布/_process_scripts/fetch_comments_api.py

"""用页面cookie调评论API获取每篇笔记的评论"""
import json, time, websocket, urllib.request

tabs = json.loads(urllib.request.urlopen("http://127.0.0.1:9222/json").read())
ws_url = None
for t in tabs:
    if "creator.xiaohongshu.com" in t.get("url", ""):
        ws_url = t["webSocketDebuggerUrl"]
        break

ws = websocket.create_connection(ws_url, timeout=30)
msg_id = 0

def send_cmd(method, params=None):
    global msg_id
    msg_id += 1
    ws.send(json.dumps({"id": msg_id, "method": method, "params": params or {}}))
    while True:
        resp = json.loads(ws.recv())
        if resp.get("id") == msg_id:
            return resp.get("result", {})

send_cmd("Page.navigate", {"url": "https://creator.xiaohongshu.com/new/note-manager"})
time.sleep(5)

# 拦截评论API
send_cmd("Network.enable")

# 导航到坦白局笔记详情（触发评论API加载）
# 先尝试直接调评论API
note_ids = [
    ("坦白局", "6a82ca500000000025004062"),
    ("D27教育博主", "6a81a53d000000000600738e"),
]

all_comments = {}
for name, nid in note_ids:
    js = f"""
    (async function() {{
        try {{
            // 尝试多种评论API端点
            var urls = [
                '/api/galaxy/creator/note/comment?noteId={nid}&cursor=&num=20',
                '/api/galaxy/v2/creator/note/comment?noteId={nid}&cursor=&num=20',
                '/api/galaxy/creator/comment/list?noteId={nid}&cursor=&num=20',
                '/api/galaxy/creator/note/comments?noteId={nid}&cursor=&num=20'
            ];
            var results = [];
            for (var i = 0; i < urls.length; i++) {{
                try {{
                    var r = await fetch(urls[i], {{credentials: 'include'}});
                    var text = await r.text();
                    if (text.indexOf('"success":true') >= 0 || text.indexOf('"code":0') >= 0) {{
                        results.push({{url: urls[i], data: text.substring(0, 5000)}});
                    }}
                }} catch(e) {{}}
            }}
            if (results.length === 0) {{
                // 也试试edith域名
                var edith_urls = [
                    'https://edith.xiaohongshu.com/api/sns/web/comment/page?note_id={nid}&cursor=&top_comment_id=&image_formats=jpg,webp,avif&xsec_token=YBieUL8uRjxi1dmXwcv1fHXDs3jL1VSMg8urb7O07P_uc='
                ];
                for (var j = 0; j < edith_urls.length; j++) {{
                    try {{
                        var r2 = await fetch(edith_urls[j], {{credentials: 'include'}});
                        var t2 = await r2.text();
                        results.push({{url: edith_urls[j], data: t2.substring(0, 5000)}});
                    }} catch(e) {{}}
                }}
            }}
            return JSON.stringify(results);
        }} catch(e) {{ return 'ERR:' + e.message; }}
    }})()
    """
    ws.send(json.dumps({"id": 300 + note_ids.index((name, nid)), "method": "Runtime.evaluate", "params": {"expression": js, "returnByValue": True, "awaitPromise": True}}))
    
for i in range(len(note_ids)):
    for _ in range(30):
        resp = json.loads(ws.recv())
        rid = resp.get("id", 0)
        if rid >= 300 and rid < 300 + len(note_ids):
            val = resp.get("result", {}).get("result", {}).get("value", "")
            name = note_ids[rid - 300][0]
            with open(rf"C:\code\cfc\运营文案\小红书发布\_process_scripts\comments_{name}.txt", "w", encoding="utf-8") as f:
                f.write(val)
            print(f"Saved comments for {name}")
            break

ws.close()