refactor(lyric): 优化歌词转存逻辑及更新冲突处理

沈秋雨
Commit 443980dc ... 443980dc1b929d6398b24b3e4159819a7cc502e2 authored 2026-07-13 14:38:40 +0800 by 沈秋雨
Showing 4 changed files with 136 additions and 42 deletions
etl_to_crawler/runner.py
etl_to_crawler/writer.py
tests/test_runner.py
tests/test_writer.py
--- a/etl_to_crawler/runner.py
View file @443980d
+++ b/etl_to_crawler/runner.py
View file @443980d
@@ -100,28 +100,55 @@ def _safe_transfer_audio(url, oss_key, bucket, base_url) -> tuple[str, str]:
        return '', ''  # 不将未转存的音频外链写入 crawler
-def _safe_upload_lyric(
+def _safe_prepare_lyric(
    platform: str,
    unique_id: str,
    spider_lyric: str | None,
    hk_lyrics_url: str | None,
    bucket,
    base_url: str,
-) -> str:
+) -> tuple[str, str]:
-    """优先将音眼歌词 URL 转存到 crawler OSS；spider 歌词文本仅作回退。"""
+    """返回实际写入的 (lyric_url, lyric_text)，音眼歌词优先、spider 回退。"""
+    candidates: list[tuple[str, str]] = []
    if str(hk_lyrics_url or '').strip():
        try:
            hk_lyric = download_text_url(hk_lyrics_url, base_url)
-            uploaded_url = upload_plain_lyric_to_bucket(platform, unique_id, hk_lyric, bucket, base_url)
+            if str(hk_lyric or '').strip():
+                candidates.append(('HK', hk_lyric))
+        except Exception as e:
+            log.warning('HK lyric download failed for %s/%s: %s', platform, unique_id, e)
+    spider_text = spider_lyric or ''
+    if str(spider_text).strip() and all(text != spider_text for _, text in candidates):
+        candidates.append(('spider', spider_text))
+    for source, lyric_text in candidates:
+        try:
+            uploaded_url = upload_plain_lyric_to_bucket(
+                platform, unique_id, lyric_text, bucket, base_url,
+            )
            if uploaded_url:
-                return uploaded_url
+                return uploaded_url, lyric_text
        except Exception as e:
-            log.warning('HK lyric transfer failed for %s/%s: %s', platform, unique_id, e)
+            log.warning('%s lyric upload failed for %s/%s: %s', source, platform, unique_id, e)
-    try:
-        return upload_plain_lyric_to_bucket(platform, unique_id, spider_lyric or '', bucket, base_url)
+    # records2 允许 URL 转存失败时入库；若文本已成功获取，仍尽量保留文本。
-    except Exception as e:
+    return '', candidates[0][1] if candidates else ''
-        log.warning("Lyric upload failed for %s/%s: %s", platform, unique_id, e)
-        return ''
+def _safe_upload_lyric(
+    platform: str,
+    unique_id: str,
+    spider_lyric: str | None,
+    hk_lyrics_url: str | None,
+    bucket,
+    base_url: str,
+) -> str:
+    """兼容仅需要歌词 URL 的调用方。"""
+    lyric_url, _ = _safe_prepare_lyric(
+        platform, unique_id, spider_lyric, hk_lyrics_url, bucket, base_url,
+    )
+    return lyric_url
 def _required_import_assets() -> tuple[str, ...]:
@@ -134,7 +161,9 @@ def _required_import_assets() -> tuple[str, ...]:
 def _require_primary_assets(assets: dict, required: tuple[str, ...] | None = None) -> None:
    """校验当前导入支线要求的资源，允许非必需资源以空值入库。"""
    audio_url, _ = assets.get('audio', ('', ''))
-    values = {'audio': audio_url, 'cover': assets.get('cover'), 'lyric': assets.get('lyric')}
+    lyric_asset = assets.get('lyric')
+    lyric_url = lyric_asset[0] if isinstance(lyric_asset, tuple) else lyric_asset
+    values = {'audio': audio_url, 'cover': assets.get('cover'), 'lyric': lyric_url}
    required = required or ('audio', 'cover', 'lyric')
    missing = [name for name in required if not values.get(name)]
    if missing:
@@ -330,7 +359,7 @@ def _prepare_qq_payload(hk_row: dict, pr: dict, bucket, base_url, sp: dict, sing
            build_oss_key('qq', 'cover', str(song_id_int) + '.jpg'),
            bucket, base_url,
        ),
-        'lyric': lambda: _safe_upload_lyric('qq', mid, raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
+        'lyric': lambda: _safe_prepare_lyric('qq', mid, raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
    }
    if sp.get('album_id') and sp.get('album_cover'):
        tasks['album_cover'] = lambda: _safe_transfer(
@@ -348,7 +377,7 @@ def _prepare_qq_payload(hk_row: dict, pr: dict, bucket, base_url, sp: dict, sing
    _require_primary_assets(assets, _required_import_assets())
    audio_url, audio_md5 = assets['audio']
    cover_url = assets['cover']
-    lyric_url = assets['lyric']
+    lyric_url, lyric_text = assets['lyric']
    album_cover = assets.get('album_cover') or cover_url
    singer_rows = [{
@@ -422,7 +451,7 @@ def _prepare_qq_payload(hk_row: dict, pr: dict, bucket, base_url, sp: dict, sing
        'version': version,
        'name': hk_row['name'],
        'duration': sp.get('duration') or hk_row.get('song_time') or 0,
-        'lyric': ensure_newlines(raw_lyric),
+        'lyric': ensure_newlines(lyric_text),
        'composer_name': sp.get('composer_name') or hk_row.get('composer'),
        'lyricist_name': sp.get('lyricist_name') or hk_row.get('lyricist'),
        'url': audio_url,
@@ -460,7 +489,7 @@ def _prepare_kugou_payload(hk_row: dict, pr: dict, bucket, base_url, sp: dict, s
            build_oss_key('kugou', 'cover', str(song_id) + '.jpg'),
            bucket, base_url,
        ),
-        'lyric': lambda: _safe_upload_lyric('kugou', str(song_id), raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
+        'lyric': lambda: _safe_prepare_lyric('kugou', str(song_id), raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
    }
    if sp.get('album_id') and sp.get('album_cover'):
        tasks['album_cover'] = lambda: _safe_transfer(
@@ -478,7 +507,7 @@ def _prepare_kugou_payload(hk_row: dict, pr: dict, bucket, base_url, sp: dict, s
    _require_primary_assets(assets, _required_import_assets())
    audio_url, audio_md5 = assets['audio']
    cover_url = assets['cover']
-    lyric_url = assets['lyric']
+    lyric_url, lyric_text = assets['lyric']
    album_cover = assets.get('album_cover') or cover_url
    singer_rows = [{
@@ -535,7 +564,7 @@ def _prepare_kugou_payload(hk_row: dict, pr: dict, bucket, base_url, sp: dict, s
        'version': version,
        'name': hk_row['name'],
        'duration': sp.get('duration') or hk_row.get('song_time') or 0,
-        'lyric': ensure_newlines(raw_lyric),
+        'lyric': ensure_newlines(lyric_text),
        'composer_name': sp.get('composer_name') or hk_row.get('composer'),
        'lyricist_name': sp.get('lyricist_name') or hk_row.get('lyricist'),
        'url': audio_url,
@@ -572,7 +601,7 @@ def _prepare_netease_payload(hk_row: dict, pr: dict, bucket, base_url, sp: dict,
            build_oss_key('netease', 'cover', str(song_id) + '.jpg'),
            bucket, base_url,
        ),
-        'lyric': lambda: _safe_upload_lyric('netease', str(song_id), raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
+        'lyric': lambda: _safe_prepare_lyric('netease', str(song_id), raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
    }
    if sp.get('album_id') and sp.get('album_cover'):
        tasks['album_cover'] = lambda: _safe_transfer(
@@ -590,7 +619,7 @@ def _prepare_netease_payload(hk_row: dict, pr: dict, bucket, base_url, sp: dict,
    _require_primary_assets(assets, _required_import_assets())
    audio_url, audio_md5 = assets['audio']
    cover_url = assets['cover']
-    lyric_url = assets['lyric']
+    lyric_url, lyric_text = assets['lyric']
    album_cover = assets.get('album_cover') or cover_url
    singer_rows = [{
@@ -659,7 +688,7 @@ def _prepare_netease_payload(hk_row: dict, pr: dict, bucket, base_url, sp: dict,
        'version': version,
        'name': hk_row['name'],
        'duration': sp.get('duration') or hk_row.get('song_time') or 0,
-        'lyric': ensure_newlines(raw_lyric),
+        'lyric': ensure_newlines(lyric_text),
        'composer_name': sp.get('composer_name') or hk_row.get('composer'),
        'lyricist_name': sp.get('lyricist_name') or hk_row.get('lyricist'),
        'url': audio_url,
@@ -870,7 +899,7 @@ def _process_qq(
            build_oss_key('qq', 'cover', str(song_id_int) + '.jpg'),
            bucket, base_url
        ),
-        'lyric': lambda: _safe_upload_lyric('qq', mid, raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
+        'lyric': lambda: _safe_prepare_lyric('qq', mid, raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
    }
    if sp.get('album_id') and sp.get('album_cover'):
        tasks['album_cover'] = lambda: _safe_transfer(
@@ -890,7 +919,7 @@ def _process_qq(
    _require_primary_assets(assets)
    audio_url, audio_md5 = assets['audio']
    cover_url = assets['cover']
-    lyric_url = assets['lyric']
+    lyric_url, lyric_text = assets['lyric']
    album_cover = assets.get('album_cover') or cover_url
    # 歌手头像转移 + 写入 singers
@@ -967,7 +996,7 @@ def _process_qq(
        'version': version,
        'name': hk_row['name'],
        'duration': sp.get('duration') or hk_row.get('song_time') or 0,
-        'lyric': ensure_newlines(raw_lyric),
+        'lyric': ensure_newlines(lyric_text),
        'composer_name': sp.get('composer_name') or hk_row.get('composer'),
        'lyricist_name': sp.get('lyricist_name') or hk_row.get('lyricist'),
        'url': audio_url,
@@ -1026,7 +1055,7 @@ def _process_kugou(
            build_oss_key('kugou', 'cover', str(song_id) + '.jpg'),
            bucket, base_url
        ),
-        'lyric': lambda: _safe_upload_lyric('kugou', str(song_id), raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
+        'lyric': lambda: _safe_prepare_lyric('kugou', str(song_id), raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
    }
    if sp.get('album_id') and sp.get('album_cover'):
        tasks['album_cover'] = lambda: _safe_transfer(
@@ -1046,7 +1075,7 @@ def _process_kugou(
    _require_primary_assets(assets)
    audio_url, audio_md5 = assets['audio']
    cover_url = assets['cover']
-    lyric_url = assets['lyric']
+    lyric_url, lyric_text = assets['lyric']
    album_cover = assets.get('album_cover') or cover_url
    singer_rows = []
@@ -1103,7 +1132,7 @@ def _process_kugou(
        'version': version,
        'name': hk_row['name'],
        'duration': sp.get('duration') or hk_row.get('song_time') or 0,
-        'lyric': ensure_newlines(raw_lyric),
+        'lyric': ensure_newlines(lyric_text),
        'composer_name': sp.get('composer_name') or hk_row.get('composer'),
        'lyricist_name': sp.get('lyricist_name') or hk_row.get('lyricist'),
        'url': audio_url,
@@ -1161,7 +1190,7 @@ def _process_netease(
            build_oss_key('netease', 'cover', str(song_id) + '.jpg'),
            bucket, base_url
        ),
-        'lyric': lambda: _safe_upload_lyric('netease', str(song_id), raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
+        'lyric': lambda: _safe_prepare_lyric('netease', str(song_id), raw_lyric, hk_row.get('lyrics_url'), bucket, base_url),
    }
    if sp.get('album_id') and sp.get('album_cover'):
        tasks['album_cover'] = lambda: _safe_transfer(
@@ -1181,7 +1210,7 @@ def _process_netease(
    _require_primary_assets(assets)
    audio_url, audio_md5 = assets['audio']
    cover_url = assets['cover']
-    lyric_url = assets['lyric']
+    lyric_url, lyric_text = assets['lyric']
    album_cover = assets.get('album_cover') or cover_url
    singer_rows = []
@@ -1250,7 +1279,7 @@ def _process_netease(
        'version': version,
        'name': hk_row['name'],
        'duration': sp.get('duration') or hk_row.get('song_time') or 0,
-        'lyric': ensure_newlines(raw_lyric),
+        'lyric': ensure_newlines(lyric_text),
        'composer_name': sp.get('composer_name') or hk_row.get('composer'),
        'lyricist_name': sp.get('lyricist_name') or hk_row.get('lyricist'),
        'url': audio_url,
--- a/etl_to_crawler/writer.py
View file @443980d
+++ b/etl_to_crawler/writer.py
View file @443980d
@@ -442,7 +442,14 @@ def upsert_qq_songs(cur, songs: list[dict]) -> None:
             platform_index_url, published_at, album, singers, status, created_at, updated_at,
             version, audio_md5, provider_name, crawler_source_data)
        VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s::json, %s::jsonb, 0, NOW(), NOW(), %s, %s, %s, %s::json)
-        ON CONFLICT (platform_song_id) DO NOTHING
+        ON CONFLICT (platform_song_id) DO UPDATE SET
+            cover = CASE WHEN BTRIM(COALESCE(crawler_qqmusic_songs.cover, '')) = ''
+                         THEN EXCLUDED.cover ELSE crawler_qqmusic_songs.cover END,
+            lyric = CASE WHEN BTRIM(COALESCE(crawler_qqmusic_songs.lyric, '')) = ''
+                         THEN EXCLUDED.lyric ELSE crawler_qqmusic_songs.lyric END,
+            lyric_url = CASE WHEN BTRIM(COALESCE(crawler_qqmusic_songs.lyric_url, '')) = ''
+                             THEN EXCLUDED.lyric_url ELSE crawler_qqmusic_songs.lyric_url END,
+            updated_at = NOW()
    """
    rows = [(
        s['song_uuid'], s['platform_song_id'], s['mid'], s.get('album_id'),
@@ -534,7 +541,14 @@ def upsert_kugou_songs(cur, songs: list[dict]) -> None:
             platform_index_url, published_at, singers, status, created_at, updated_at,
             version, audio_md5, provider_name, crawler_source_data)
        VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s::jsonb, 0, NOW(), NOW(), %s, %s, %s, %s::json)
-        ON CONFLICT (platform_song_id) DO NOTHING
+        ON CONFLICT (platform_song_id) DO UPDATE SET
+            cover = CASE WHEN BTRIM(COALESCE(crawler_kugou_songs.cover, '')) = ''
+                         THEN EXCLUDED.cover ELSE crawler_kugou_songs.cover END,
+            lyric = CASE WHEN BTRIM(COALESCE(crawler_kugou_songs.lyric, '')) = ''
+                         THEN EXCLUDED.lyric ELSE crawler_kugou_songs.lyric END,
+            lyric_url = CASE WHEN BTRIM(COALESCE(crawler_kugou_songs.lyric_url, '')) = ''
+                             THEN EXCLUDED.lyric_url ELSE crawler_kugou_songs.lyric_url END,
+            updated_at = NOW()
    """
    rows = [(
        s['song_uuid'], s['platform_song_id'], s.get('hash', ''),
@@ -623,7 +637,14 @@ def upsert_netease_songs(cur, songs: list[dict]) -> None:
             platform_index_url, published_at, album, singers, status, created_at, updated_at,
             version, audio_md5, provider_name, crawler_source_data)
        VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s::json, %s::jsonb, 0, NOW(), NOW(), %s, %s, %s, %s::json)
-        ON CONFLICT (platform_song_id) DO NOTHING
+        ON CONFLICT (platform_song_id) DO UPDATE SET
+            cover = CASE WHEN BTRIM(COALESCE(crawler_netease_songs.cover, '')) = ''
+                         THEN EXCLUDED.cover ELSE crawler_netease_songs.cover END,
+            lyric = CASE WHEN BTRIM(COALESCE(crawler_netease_songs.lyric, '')) = ''
+                         THEN EXCLUDED.lyric ELSE crawler_netease_songs.lyric END,
+            lyric_url = CASE WHEN BTRIM(COALESCE(crawler_netease_songs.lyric_url, '')) = ''
+                             THEN EXCLUDED.lyric_url ELSE crawler_netease_songs.lyric_url END,
+            updated_at = NOW()
    """
    rows = [(
        s['song_uuid'], s['platform_song_id'], s.get('album_id'),
--- a/tests/test_runner.py
View file @443980d
+++ b/tests/test_runner.py
View file @443980d
@@ -105,32 +105,35 @@ def test_records2_import_allows_empty_cover_and_lyric(monkeypatch):
    }, runner._required_import_assets())
-def test_lyric_upload_prefers_hk_url_and_forces_oss_transfer(monkeypatch):
+def test_lyric_prepare_prefers_hk_text_for_both_url_and_text_column(monkeypatch):
    download = MagicMock(return_value='[00:01.00]HK lyric')
    monkeypatch.setattr(runner, 'download_text_url', download)
    upload = MagicMock(return_value='https://archive.example/crawler/lyric/qq/mid.txt')
    monkeypatch.setattr(runner, 'upload_plain_lyric_to_bucket', upload)
    bucket = object()
-    result = runner._safe_upload_lyric(
+    result = runner._safe_prepare_lyric(
        'qq', 'mid', 'spider lyric', 'https://source.example/lyric.txt', bucket, 'https://archive.example',
    )
-    assert result == 'https://archive.example/crawler/lyric/qq/mid.txt'
+    assert result == (
+        'https://archive.example/crawler/lyric/qq/mid.txt',
+        '[00:01.00]HK lyric',
+    )
    download.assert_called_once_with('https://source.example/lyric.txt', 'https://archive.example')
    upload.assert_called_once_with(
        'qq', 'mid', '[00:01.00]HK lyric', bucket, 'https://archive.example',
    )
-def test_lyric_upload_falls_back_to_spider_text_when_hk_transfer_fails(monkeypatch):
+def test_lyric_prepare_falls_back_to_spider_text_when_hk_download_fails(monkeypatch):
    monkeypatch.setattr(runner, 'download_text_url', MagicMock(side_effect=TimeoutError('timeout')))
    upload = MagicMock(return_value='https://archive.example/crawler/lyric/qq/mid.txt')
    monkeypatch.setattr(runner, 'upload_plain_lyric_to_bucket', upload)
-    assert runner._safe_upload_lyric(
+    assert runner._safe_prepare_lyric(
        'qq', 'mid', 'spider lyric', 'https://source.example/lyric.txt', object(), 'https://archive.example',
-    ) == 'https://archive.example/crawler/lyric/qq/mid.txt'
+    ) == ('https://archive.example/crawler/lyric/qq/mid.txt', 'spider lyric')
    assert upload.call_args.args[2] == 'spider lyric'
@@ -519,7 +522,10 @@ def test_process_qq_builds_album_json_for_song_insert(monkeypatch):
    monkeypatch.setattr(runner, 'fetch_qq_singers', lambda conn, song_ids: {})
    monkeypatch.setattr(runner, '_safe_transfer', lambda url, oss_key, bucket, base_url: url)
    monkeypatch.setattr(runner, '_safe_transfer_audio', lambda url, oss_key, bucket, base_url: (url, 'audio-md5'))
-    monkeypatch.setattr(runner, '_safe_upload_lyric', lambda *args: 'https://bucket.example.com/lyric.txt')
+    monkeypatch.setattr(
+        runner, '_safe_prepare_lyric',
+        lambda *args: ('https://bucket.example.com/lyric.txt', '[00:01.00]HK歌词'),
+    )
    monkeypatch.setattr(runner, 'upsert_qq_singers', lambda cur, singers: None)
    monkeypatch.setattr(runner, 'upsert_qq_albums', lambda cur, albums: None)
    monkeypatch.setattr(runner, 'upsert_qq_songs', lambda cur, songs: inserted_songs.extend(songs))
@@ -550,6 +556,8 @@ def test_process_qq_builds_album_json_for_song_insert(monkeypatch):
    assert '"title": "专辑"' in inserted_songs[0]['album_json']
    assert inserted_songs[0]['title'] == '化风行万里'
    assert inserted_songs[0]['version'] == 'DJ默涵版'
+    assert inserted_songs[0]['lyric'] == '[00:01.00]HK歌词'
+    assert inserted_songs[0]['lyric_url'] == 'https://bucket.example.com/lyric.txt'
 def test_process_netease_builds_album_json_for_song_insert(monkeypatch):
@@ -582,7 +590,10 @@ def test_process_netease_builds_album_json_for_song_insert(monkeypatch):
    monkeypatch.setattr(runner, 'fetch_netease_singers', lambda conn, song_ids: {})
    monkeypatch.setattr(runner, '_safe_transfer', lambda url, oss_key, bucket, base_url: url)
    monkeypatch.setattr(runner, '_safe_transfer_audio', lambda url, oss_key, bucket, base_url: (url, 'audio-md5'))
-    monkeypatch.setattr(runner, '_safe_upload_lyric', lambda *args: 'https://bucket.example.com/lyric.txt')
+    monkeypatch.setattr(
+        runner, '_safe_prepare_lyric',
+        lambda *args: ('https://bucket.example.com/lyric.txt', '[00:01.00]HK歌词'),
+    )
    monkeypatch.setattr(runner, 'upsert_netease_singers', lambda cur, singers: None)
    monkeypatch.setattr(runner, 'upsert_netease_albums', lambda cur, albums: None)
    monkeypatch.setattr(runner, 'upsert_netease_songs', lambda cur, songs: inserted_songs.extend(songs))
@@ -625,7 +636,10 @@ def test_process_netease_uses_prefetched_song_and_singers(monkeypatch):
    monkeypatch.setattr(runner, 'fetch_netease_singers', fetch_singers)
    monkeypatch.setattr(runner, '_safe_transfer', lambda url, oss_key, bucket, base_url: url)
    monkeypatch.setattr(runner, '_safe_transfer_audio', lambda url, oss_key, bucket, base_url: (url, 'audio-md5'))
-    monkeypatch.setattr(runner, '_safe_upload_lyric', lambda *args: 'https://bucket.example.com/lyric.txt')
+    monkeypatch.setattr(
+        runner, '_safe_prepare_lyric',
+        lambda *args: ('https://bucket.example.com/lyric.txt', '[00:01.00]HK歌词'),
+    )
    monkeypatch.setattr(runner, 'upsert_netease_singers', lambda cur, singers: None)
    monkeypatch.setattr(runner, 'upsert_netease_albums', lambda cur, albums: None)
    monkeypatch.setattr(runner, 'upsert_netease_songs', lambda cur, songs: inserted_songs.extend(songs))
--- a/tests/test_writer.py
View file @443980d
+++ b/tests/test_writer.py
View file @443980d
 from unittest.mock import MagicMock, call
+import pytest
 import etl_to_crawler.writer as writer
 from etl_to_crawler.writer import (
    fetch_pending_yinyan_song_records,
@@ -327,6 +329,34 @@ def test_upsert_kugou_songs_writes_provider_and_source_data():
    assert rows[0][-4:] == ('DJ默涵版', 'md5-200', 'yinyan', '{"id": 200}')
+@pytest.mark.parametrize(
+    ('upsert', 'table_name', 'extra'),
+    [
+        (upsert_qq_songs, 'crawler_qqmusic_songs', {'mid': 'song-mid'}),
+        (upsert_kugou_songs, 'crawler_kugou_songs', {'hash': 'hash'}),
+        (upsert_netease_songs, 'crawler_netease_songs', {}),
+    ],
+)
+def test_song_conflict_only_fills_empty_cover_and_lyric_fields(upsert, table_name, extra):
+    cur = MagicMock()
+    upsert(cur, [{
+        'song_uuid': 'song-uuid',
+        'platform_song_id': 200,
+        'name': '歌名',
+        'cover': 'https://example.com/cover.jpg',
+        'lyric': '歌词文本',
+        'lyric_url': 'https://example.com/lyric.txt',
+        **extra,
+    }])
+    sql = cur.executemany.call_args.args[0]
+    assert 'ON CONFLICT (platform_song_id) DO UPDATE SET' in sql
+    assert f"BTRIM(COALESCE({table_name}.cover, '')) = ''" in sql
+    assert f"BTRIM(COALESCE({table_name}.lyric, '')) = ''" in sql
+    assert f"BTRIM(COALESCE({table_name}.lyric_url, '')) = ''" in sql
+    assert 'THEN EXCLUDED.lyric' in sql
 def test_upsert_kugou_singers_writes_provider_and_source_data():
    cur = MagicMock()
    upsert_kugou_singers(cur, [{