Commit 0a14389c 0a14389c9e72e597698089b7929d194b2e5a40ab by 沈秋雨

feat(etl): 实现三平台 PostgreSQL 写入函数

1 parent ab7f4ed4
import uuid
# ─── QQ Music ────────────────────────────────────────────────────────────────
def upsert_qq_singers(cur, singers: list[dict]) -> None:
if not singers:
return
sql = """
INSERT INTO crawler_qqmusic_singers
(id, mid, name, avatar, sex, area, "index", intro, home_url, created_at, updated_at)
VALUES (%s, %s, %s, %s, %s::singer_sex, %s::singer_area, %s::singer_index, %s, %s, NOW(), NOW())
ON CONFLICT (id) DO NOTHING
"""
rows = [(
s['id'], s['mid'], s['name'], s.get('avatar', ''),
s.get('sex') or 'U', s.get('area') or '其他', s.get('index') or '#',
s.get('intro'), s.get('home_url'),
) for s in singers]
cur.executemany(sql, rows)
def upsert_qq_albums(cur, albums: list[dict]) -> None:
if not albums:
return
sql = """
INSERT INTO crawler_qqmusic_albums
(id, mid, cover, title, intro, type, company_id, company, is_owner, published_at, created_at, updated_at)
VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW(), NOW())
ON CONFLICT (id) DO NOTHING
"""
rows = [(
a['id'], a.get('mid', ''), a.get('cover', ''), a.get('title', ''),
a.get('intro'), a.get('type', ''), a.get('company_id', 0) or 0,
a.get('company', ''), a.get('is_owner', 0) or 0, a.get('published_at'),
) for a in albums]
cur.executemany(sql, rows)
def upsert_qq_songs(cur, songs: list[dict]) -> None:
if not songs:
return
sql = """
INSERT INTO crawler_qqmusic_songs
(id, platform_song_id, mid, album_id, cover, title, name, duration,
lyric, composer_name, lyricist_name, url, lyric_url,
platform_index_url, published_at, singers, status, created_at, updated_at)
VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s::jsonb, 0, NOW(), NOW())
ON CONFLICT (platform_song_id) DO NOTHING
"""
rows = [(
s['song_uuid'], s['platform_song_id'], s['mid'], s.get('album_id'),
s.get('cover', ''), s.get('title', ''), s.get('name', ''), s.get('duration', 0) or 0,
s.get('lyric'), s.get('composer_name'), s.get('lyricist_name'),
s.get('url', ''), s.get('lyric_url'),
s.get('platform_index_url'), s.get('published_at'), s.get('singers_json', '[]'),
) for s in songs]
cur.executemany(sql, rows)
def upsert_qq_singer_songs(cur, pairs: list[tuple]) -> None:
"""pairs: [(singer_id, song_uuid), ...]"""
if not pairs:
return
sql = """
INSERT INTO crawler_qqmusic_singer_songs (id, singer_id, song_id)
VALUES (%s, %s, %s)
ON CONFLICT (singer_id, song_id) DO NOTHING
"""
rows = [(str(uuid.uuid4()), singer_id, song_uuid) for singer_id, song_uuid in pairs]
cur.executemany(sql, rows)
def upsert_qq_singer_albums(cur, pairs: list[tuple]) -> None:
"""pairs: [(singer_id, album_id), ...]"""
if not pairs:
return
sql = """
INSERT INTO crawler_qqmusic_singer_albums (id, singer_id, album_id)
VALUES (%s, %s, %s)
ON CONFLICT (singer_id, album_id) DO NOTHING
"""
rows = [(str(uuid.uuid4()), singer_id, album_id) for singer_id, album_id in pairs]
cur.executemany(sql, rows)
# ─── Kugou ───────────────────────────────────────────────────────────────────
def upsert_kugou_singers(cur, singers: list[dict]) -> None:
if not singers:
return
sql = """
INSERT INTO crawler_kugou_singers
(id, name, avatar, sex, area, "index", intro, home_url, created_at, updated_at)
VALUES (%s, %s, %s, %s::kugou_singer_sex, %s::kugou_singer_area, %s::kugou_singer_index, %s, %s, NOW(), NOW())
ON CONFLICT (id) DO NOTHING
"""
rows = [(
s['id'], s['name'], s.get('avatar', ''),
s.get('sex') or 'U', s.get('area') or '其他', s.get('index') or '#',
s.get('intro'), s.get('home_url', ''),
) for s in singers]
cur.executemany(sql, rows)
def upsert_kugou_albums(cur, albums: list[dict]) -> None:
if not albums:
return
sql = """
INSERT INTO crawler_kugou_albums
(id, cover, title, intro, type, company_id, company, is_owner, published_at, created_at, updated_at)
VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, NOW(), NOW())
ON CONFLICT (id) DO NOTHING
"""
rows = [(
a['id'], a.get('cover', ''), a.get('title', ''), a.get('intro'),
a.get('type', ''), a.get('company_id', 0) or 0,
a.get('company', ''), a.get('is_owner', 0) or 0, a.get('published_at'),
) for a in albums]
cur.executemany(sql, rows)
def upsert_kugou_songs(cur, songs: list[dict]) -> None:
if not songs:
return
sql = """
INSERT INTO crawler_kugou_songs
(id, platform_song_id, hash, album_audio_id, album_id, cover, title, name, duration,
lyric, composer_name, lyricist_name, url, lyric_url,
platform_index_url, published_at, singers, status, created_at, updated_at)
VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s::jsonb, 0, NOW(), NOW())
ON CONFLICT (platform_song_id) DO NOTHING
"""
rows = [(
s['song_uuid'], s['platform_song_id'], s.get('hash', ''),
s.get('album_audio_id', 0) or 0, s.get('album_id'),
s.get('cover', ''), s.get('title', ''), s.get('name', ''), s.get('duration', 0) or 0,
s.get('lyric'), s.get('composer_name'), s.get('lyricist_name'),
s.get('url', ''), s.get('lyric_url'),
s.get('platform_index_url'), s.get('published_at'), s.get('singers_json', '[]'),
) for s in songs]
cur.executemany(sql, rows)
def upsert_kugou_singer_songs(cur, pairs: list[tuple]) -> None:
if not pairs:
return
sql = """
INSERT INTO crawler_kugou_singer_songs (id, singer_id, song_id)
VALUES (%s, %s, %s)
ON CONFLICT (singer_id, song_id) DO NOTHING
"""
cur.executemany(sql, [(str(uuid.uuid4()), s, sg) for s, sg in pairs])
def upsert_kugou_singer_albums(cur, pairs: list[tuple]) -> None:
if not pairs:
return
sql = """
INSERT INTO crawler_kugou_singer_albums (id, singer_id, album_id)
VALUES (%s, %s, %s)
ON CONFLICT (singer_id, album_id) DO NOTHING
"""
cur.executemany(sql, [(str(uuid.uuid4()), s, a) for s, a in pairs])
# ─── Netease ─────────────────────────────────────────────────────────────────
def upsert_netease_singers(cur, singers: list[dict]) -> None:
if not singers:
return
sql = """
INSERT INTO crawler_netease_singers
(id, name, avatar, sex, area, "index", intro, home_url, created_at, updated_at)
VALUES (%s, %s, %s, %s::netease_singer_sex, %s::netease_singer_area, %s::netease_singer_index, %s, %s, NOW(), NOW())
ON CONFLICT (id) DO NOTHING
"""
rows = [(
s['id'], s['name'], s.get('avatar', ''),
s.get('sex') or 'U', s.get('area') or '其他', s.get('index') or '#',
s.get('intro'), s.get('home_url'),
) for s in singers]
cur.executemany(sql, rows)
def upsert_netease_albums(cur, albums: list[dict]) -> None:
if not albums:
return
sql = """
INSERT INTO crawler_netease_albums
(id, cover, title, intro, type, company_id, company, is_owner, published_at, created_at, updated_at)
VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, NOW(), NOW())
ON CONFLICT (id) DO NOTHING
"""
rows = [(
a['id'], a.get('cover', ''), a.get('title', ''), a.get('intro'),
a.get('type', ''), a.get('company_id', 0) or 0,
a.get('company', ''), a.get('is_owner', 0) or 0, a.get('published_at'),
) for a in albums]
cur.executemany(sql, rows)
def upsert_netease_songs(cur, songs: list[dict]) -> None:
if not songs:
return
sql = """
INSERT INTO crawler_netease_songs
(id, platform_song_id, album_id, cover, title, name, duration,
lyric, composer_name, lyricist_name, url, lyric_url,
platform_index_url, published_at, singers, status, created_at, updated_at)
VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s::jsonb, 0, NOW(), NOW())
ON CONFLICT (platform_song_id) DO NOTHING
"""
rows = [(
s['song_uuid'], s['platform_song_id'], s.get('album_id'),
s.get('cover', ''), s.get('title', ''), s.get('name', ''), s.get('duration', 0) or 0,
s.get('lyric'), s.get('composer_name'), s.get('lyricist_name'),
s.get('url', ''), s.get('lyric_url'),
s.get('platform_index_url'), s.get('published_at'), s.get('singers_json', '[]'),
) for s in songs]
cur.executemany(sql, rows)
def upsert_netease_singer_songs(cur, pairs: list[tuple]) -> None:
if not pairs:
return
sql = """
INSERT INTO crawler_netease_singer_songs (id, singer_id, song_id)
VALUES (%s, %s, %s)
ON CONFLICT (singer_id, song_id) DO NOTHING
"""
cur.executemany(sql, [(str(uuid.uuid4()), s, sg) for s, sg in pairs])
def upsert_netease_singer_albums(cur, pairs: list[tuple]) -> None:
if not pairs:
return
sql = """
INSERT INTO crawler_netease_singer_albums (id, singer_id, album_id)
VALUES (%s, %s, %s)
ON CONFLICT (singer_id, album_id) DO NOTHING
"""
cur.executemany(sql, [(str(uuid.uuid4()), s, a) for s, a in pairs])
from unittest.mock import MagicMock, call
from etl_to_crawler.writer import upsert_qq_singers, upsert_qq_singer_songs
def test_upsert_qq_singers_executes_insert():
cur = MagicMock()
singers = [{
'id': 1, 'mid': 'abc', 'name': '歌手', 'avatar': 'https://archive-dev.oss-cn-beijing.aliyuncs.com/a.jpg',
'sex': 'M', 'area': '华语', 'index': 'G', 'intro': None, 'home_url': None
}]
upsert_qq_singers(cur, singers)
assert cur.executemany.called
args = cur.executemany.call_args
assert 'crawler_qqmusic_singers' in args[0][0]
assert 'ON CONFLICT' in args[0][0]
def test_upsert_qq_singers_empty_does_nothing():
cur = MagicMock()
upsert_qq_singers(cur, [])
cur.executemany.assert_not_called()
def test_upsert_qq_singer_songs():
cur = MagicMock()
upsert_qq_singer_songs(cur, [(1, 'uuid-abc'), (2, 'uuid-def')])
assert cur.executemany.called
sql = cur.executemany.call_args[0][0]
assert 'crawler_qqmusic_singer_songs' in sql