feat: 提交热榜评论分析工具 MVP 基线

This commit is contained in:
meijiali
2026-07-03 11:12:12 +08:00
parent 0231be6338
commit 9bad7b6e63
60 changed files with 4058 additions and 0 deletions
+1
View File
@@ -0,0 +1 @@
+117
View File
@@ -0,0 +1,117 @@
import time
from dataclasses import dataclass
from datetime import UTC, datetime
from typing import Any
import httpx
@dataclass
class HotspotData:
source_hot_id: str | None
title: str
rank: int | None = None
heat_value: str | None = None
raw_data: dict[str, Any] | None = None
@dataclass
class ContentItemData:
source_item_id: str
item_type: str
title: str | None = None
summary: str | None = None
url: str | None = None
raw_data: dict[str, Any] | None = None
@dataclass
class CommentData:
source_comment_id: str | None
content: str
author: str | None = None
like_count: int | None = None
comment_time: datetime | None = None
raw_data: dict[str, Any] | None = None
class PlatformAPIError(Exception):
def __init__(self, message: str, *, error_type: str, status_code: int | None = None):
super().__init__(message)
self.error_type = error_type
self.status_code = status_code
class TikHubClient:
def __init__(
self,
*,
base_url: str,
api_key: str,
timeout_seconds: int = 20,
max_retries: int = 3,
http_client: httpx.Client | None = None,
) -> None:
self.base_url = base_url.rstrip("/")
self.api_key = api_key
self.timeout_seconds = timeout_seconds
self.max_retries = max_retries
self._http_client = http_client or httpx.Client(timeout=timeout_seconds)
def get(self, path: str, *, params: dict[str, Any] | None = None) -> dict[str, Any]:
return self._request("GET", path, params=params)
def post(self, path: str, *, json: dict[str, Any] | None = None) -> dict[str, Any]:
return self._request("POST", path, json=json)
def _request(self, method: str, path: str, **kwargs) -> dict[str, Any]:
url = f"{self.base_url}{path}"
headers = {"Authorization": f"Bearer {self.api_key}"} if self.api_key else {}
last_status: int | None = None
for attempt in range(self.max_retries + 1):
try:
response = self._http_client.request(method, url, headers=headers, **kwargs)
except httpx.RequestError as exc:
if attempt >= self.max_retries:
raise PlatformAPIError("External API request failed", error_type="network_error") from exc
time.sleep(2**attempt)
continue
last_status = response.status_code
if response.status_code == 429:
if attempt >= self.max_retries:
raise PlatformAPIError(
"External API rate limited",
error_type="rate_limited",
status_code=response.status_code,
)
time.sleep(2**attempt)
continue
if response.is_error:
raise PlatformAPIError(
f"External API returned HTTP {response.status_code}",
error_type="api_error",
status_code=response.status_code,
)
return response.json()
raise PlatformAPIError("External API request failed", error_type="api_error", status_code=last_status)
def parse_timestamp(value: Any) -> datetime | None:
if value in (None, ""):
return None
try:
number = int(value)
except (TypeError, ValueError):
return None
if number > 10_000_000_000:
number = number // 1000
return datetime.fromtimestamp(number, tz=UTC)
def first_present(data: dict[str, Any], *keys: str) -> Any:
for key in keys:
if key in data and data[key] not in (None, ""):
return data[key]
return None
+99
View File
@@ -0,0 +1,99 @@
from typing import Any
from app.platforms.base import CommentData, ContentItemData, HotspotData, TikHubClient, first_present, parse_timestamp
class DouyinPlatform:
def __init__(self, client: TikHubClient | None) -> None:
self.client = client
def map_hotspots(self, payload: dict[str, Any], *, limit: int) -> list[HotspotData]:
data = payload.get("data", {})
items = data.get("word_list") or data.get("list") or data.get("item_list") or []
result = []
for index, item in enumerate(items[:limit], start=1):
result.append(
HotspotData(
source_hot_id=str(item.get("query_id")) if item.get("query_id") is not None else None,
title=str(first_present(item, "title", "sentence", "word") or ""),
rank=item.get("rank") or index,
heat_value=str(item.get("hot_score")) if item.get("hot_score") is not None else None,
raw_data=item,
)
)
return result
def map_items(self, payload: dict[str, Any], *, limit: int) -> list[ContentItemData]:
data = payload.get("data", [])
raw_items = data.get("business_data") or data.get("data") or data.get("items") or [] if isinstance(data, dict) else data
result = []
for item in raw_items[:limit]:
nested = item.get("data", item) if isinstance(item, dict) else {}
aweme = nested.get("aweme_info", nested) if isinstance(nested, dict) else {}
aweme_id = aweme.get("aweme_id")
if not aweme_id:
continue
result.append(
ContentItemData(
source_item_id=str(aweme_id),
item_type="video",
title=aweme.get("desc"),
summary=aweme.get("desc"),
url=aweme.get("share_url"),
raw_data=aweme,
)
)
return result
def map_comments(self, payload: dict[str, Any], *, limit: int) -> list[CommentData]:
comments = payload.get("comments") or payload.get("data", {}).get("comments") or []
result = []
for comment in comments[:limit]:
content = comment.get("text")
if not content:
continue
result.append(
CommentData(
source_comment_id=first_present(comment, "comment_id", "cid"),
content=str(content),
author=self._author_name(comment),
like_count=comment.get("digg_count"),
comment_time=parse_timestamp(first_present(comment, "create_time", "create_time_str")),
raw_data=comment,
)
)
return result
def fetch_hotspots(self, *, limit: int) -> list[HotspotData]:
payload = self.client.get(
"/api/v1/douyin/creator/fetch_creator_hot_spot_billboard",
params={"billboard_tag": 0, "hot_search_type": 1},
)
return self.map_hotspots(payload, limit=limit)
def search_items_by_hotspot(self, keyword: str, *, limit: int) -> list[ContentItemData]:
payload = self.client.post(
"/api/v1/douyin/search/fetch_video_search_v2",
json={
"keyword": keyword,
"cursor": 0,
"sort_type": "0",
"publish_time": "0",
"filter_duration": "0",
"content_type": "1",
"search_id": "",
"backtrace": "",
},
)
return self.map_items(payload, limit=limit)
def fetch_comments(self, source_item_id: str, *, limit: int) -> list[CommentData]:
payload = self.client.get(
"/api/v1/douyin/app/v3/fetch_video_comments",
params={"aweme_id": source_item_id, "cursor": 0, "count": 20},
)
return self.map_comments(payload, limit=limit)
def _author_name(self, comment: dict[str, Any]) -> str | None:
user = comment.get("user") or {}
return user.get("nickname") or user.get("name") if isinstance(user, dict) else None
+84
View File
@@ -0,0 +1,84 @@
from typing import Any
from app.platforms.base import CommentData, ContentItemData, HotspotData, TikHubClient, first_present, parse_timestamp
class XiaohongshuPlatform:
def __init__(self, client: TikHubClient | None) -> None:
self.client = client
def map_hotspots(self, payload: dict[str, Any], *, limit: int) -> list[HotspotData]:
items = payload.get("data", {}).get("data", {}).get("items", [])
hotspots = []
for index, item in enumerate(items[:limit], start=1):
hotspots.append(
HotspotData(
source_hot_id=str(item.get("id")) if item.get("id") is not None else None,
title=str(item.get("title") or ""),
rank=index,
heat_value=str(item.get("score")) if item.get("score") is not None else None,
raw_data=item,
)
)
return hotspots
def map_items(self, payload: dict[str, Any], *, limit: int) -> list[ContentItemData]:
raw_items = payload.get("data", {}).get("data", {}).get("items", [])
notes = [item.get("note", item) for item in raw_items]
notes.sort(key=lambda note: 0 if int(note.get("comments_count") or 0) > 0 else 1)
result = []
for note in notes[:limit]:
note_id = note.get("id")
if not note_id:
continue
result.append(
ContentItemData(
source_item_id=str(note_id),
item_type="note",
title=note.get("title"),
summary=note.get("desc"),
url=note.get("url"),
raw_data=note,
)
)
return result
def map_comments(self, payload: dict[str, Any], *, limit: int) -> list[CommentData]:
comments = payload.get("data", {}).get("data", {}).get("comments", [])
result = []
for comment in comments[:limit]:
content = first_present(comment, "content", "text")
if not content:
continue
result.append(
CommentData(
source_comment_id=first_present(comment, "comment_id", "id"),
content=str(content),
author=self._author_name(comment),
like_count=comment.get("like_count"),
comment_time=parse_timestamp(first_present(comment, "create_time", "create_time_str")),
raw_data=comment,
)
)
return result
def fetch_hotspots(self, *, limit: int) -> list[HotspotData]:
return self.map_hotspots(self.client.get("/api/v1/xiaohongshu/web_v2/fetch_hot_list"), limit=limit)
def search_items_by_hotspot(self, keyword: str, *, limit: int) -> list[ContentItemData]:
payload = self.client.get(
"/api/v1/xiaohongshu/app_v2/search_notes",
params={"keyword": keyword, "page": 1, "sort": "general", "note_type": 0},
)
return self.map_items(payload, limit=limit)
def fetch_comments(self, source_item_id: str, *, limit: int) -> list[CommentData]:
payload = self.client.get(
"/api/v1/xiaohongshu/app_v2/get_note_comments",
params={"note_id": source_item_id, "cursor": "", "index": 0, "pageArea": "UNFOLDED", "sort_strategy": "latest_v2"},
)
return self.map_comments(payload, limit=limit)
def _author_name(self, comment: dict[str, Any]) -> str | None:
user = comment.get("user_info") or comment.get("user") or {}
return user.get("nickname") or user.get("name") if isinstance(user, dict) else None