NeteaseUrl/tests/test_tagging.py

# -*- coding: utf-8 -*-
"""标签写入（mutagen）与内容嗅探（filetype）的单元测试。

覆盖：
- build_tags 的字段勾选、name/title 别名、LRC 时间标签剔除、曲号/年份缺省
- write_tags 的失败路径（不支持的扩展名 / 文件不存在）
- MP3 与 FLAC 的真实往返（写入后按标签读回）
- _sniff_audio_header 对音频容器放行、对 CDN 错误页/JSON/空文件拒绝
"""
from __future__ import annotations

from typing import Any

import pytest

from music_downloader import MusicInfo
from music_downloader._http import _sniff_audio_header
from music_downloader.tagging import TagSet, build_tags, write_tags

# 最小合法 FLAC：fLaC 魔数 + 一个 34 字节的 STREAMINFO（last-block 标志置位）。
# STREAMINFO 不能全零——mutagen 会拒绝采样率为 0 的流；此处填 44100Hz/2ch/16bit：
#   min_blocksize=4096 max_blocksize=4096 min_framesize=1000 max_framesize=2000
#   随后 8 字节打包 (sample_rate 20bit | channels-1 3bit | bps-1 5bit | total 36bit)
_STREAMINFO = bytes.fromhex(
    '1000' '1000' '0003e8' '0007d0' '0ac442f00000ac44' + '00' * 16)
assert len(_STREAMINFO) == 34
_MINIMAL_FLAC = b'fLaC' + bytes([0x80, 0x00, 0x00, 0x22]) + _STREAMINFO

# 最小 JPEG 头（仅用于标签写入，不参与解码）
_JPEG = b'\xff\xd8\xff\xe0' + b'\x00' * 32


def _info(**kw) -> MusicInfo:
    base: dict[str, Any] = {
        'id': 1, 'name': '歌名', 'artists': '歌手', 'album': '专辑', 'pic_url': '',
        'duration': 200, 'track_number': 3,
        'download_url': 'http://example.invalid/x', 'file_type': 'mp3',
        'file_size': 0, 'quality': 'standard',
        'lyric': '[00:12.34]第一行\n[00:15.00]第二行', 'tlyric': '',
        'publish_year': 2024,
    }
    base.update(kw)
    return MusicInfo(**base)


# ---------- build_tags ----------

def test_build_tags_all_fields():
    tags = build_tags(_info(), None)
    assert tags.title == '歌名'
    assert tags.artist == '歌手'
    assert tags.album_artist == '歌手'
    assert tags.album == '专辑'
    assert tags.date == '2024'
    assert tags.track == 3
    assert tags.lyrics == '第一行\n第二行'      # LRC 时间标签已剔除


def test_build_tags_respects_selection():
    tags = build_tags(_info(), ['title'])
    assert tags.title == '歌名'
    assert tags.artist is None
    assert tags.album is None
    assert tags.lyrics is None
    assert tags.track is None


def test_build_tags_accepts_name_alias():
    assert build_tags(_info(), ['name']).title == '歌名'


@pytest.mark.parametrize('kw, field', [
    ({'track_number': 0}, 'track'),
    ({'publish_year': None}, 'date'),
    ({'lyric': ''}, 'lyrics'),
    ({'lyric': '[00:01.00]'}, 'lyrics'),        # 只有时间标签 → 清洗后为空
])
def test_build_tags_omits_empty_values(kw, field):
    assert getattr(build_tags(_info(**kw), None), field) is None


# ---------- write_tags ----------

def test_write_tags_rejects_unsupported_extension(tmp_path):
    p = tmp_path / 'a.wav'
    p.write_bytes(b'RIFF' + b'\x00' * 8)
    assert write_tags(p, TagSet(title='x')) is False


def test_write_tags_returns_false_when_missing(tmp_path):
    assert write_tags(tmp_path / 'nope.mp3', TagSet(title='x')) is False


def test_write_tags_mp3_roundtrip(tmp_path):
    from mutagen.id3 import ID3

    p = tmp_path / 'a.mp3'
    ID3().save(str(p))          # 构造一个只含 ID3 标签区的合法 mp3
    tags = TagSet(title='T', artist='A', album='B', album_artist='A2',
                  date='2020', track=2, lyrics='L')
    assert write_tags(p, tags, cover=_JPEG) is True

    got = ID3(str(p))
    assert got['TIT2'].text[0] == 'T'
    assert got['TPE1'].text[0] == 'A'
    assert got['TPE2'].text[0] == 'A2'
    assert got['TALB'].text[0] == 'B'
    assert got['TRCK'].text[0] == '2'
    # USLT/APIC 的 HashKey 带语言/描述后缀，须用 getall 取
    assert got.getall('USLT')[0].text == 'L'
    assert got.getall('APIC')[0].mime == 'image/jpeg'


def test_write_tags_mp3_keeps_untouched_frames(tmp_path):
    from mutagen.id3 import ID3, TXXX

    p = tmp_path / 'a.mp3'
    audio = ID3()
    audio['TXXX:keep'] = TXXX(encoding=3, desc='keep', text=['v'])
    audio.save(str(p))

    assert write_tags(p, TagSet(title='T')) is True
    assert ID3(str(p))['TXXX:keep'].text[0] == 'v'   # 未触及的帧被保留


def test_write_tags_flac_roundtrip(tmp_path):
    from mutagen.flac import FLAC

    p = tmp_path / 'a.flac'
    p.write_bytes(_MINIMAL_FLAC)
    assert write_tags(p, TagSet(title='T', artist='A', track=4), cover=_JPEG) is True

    got = FLAC(str(p))
    assert got['title'][0] == 'T'
    assert got['artist'][0] == 'A'
    assert got['tracknumber'][0] == '4'
    assert len(got.pictures) == 1
    assert got.pictures[0].mime == 'image/jpeg'


def test_write_tags_failure_is_swallowed(tmp_path):
    """写坏文件不应抛异常（主流程靠返回值判断）。"""
    p = tmp_path / 'broken.flac'
    p.write_bytes(b'fLaC')          # 截断的 FLAC，解析必然失败
    assert write_tags(p, TagSet(title='T')) is False


# ---------- 内容嗅探 ----------

def test_sniff_accepts_id3(tmp_path):
    p = tmp_path / 'x.mp3'
    p.write_bytes(b'ID3\x03\x00\x00\x00' + b'\x00' * 64)
    assert _sniff_audio_header(p) is True


def test_sniff_accepts_flac(tmp_path):
    p = tmp_path / 'x.flac'
    p.write_bytes(_MINIMAL_FLAC)
    assert _sniff_audio_header(p) is True


def test_sniff_accepts_ogg(tmp_path):
    p = tmp_path / 'x.ogg'
    p.write_bytes(b'OggS' + b'\x00' * 64)
    assert _sniff_audio_header(p) is True


def test_sniff_rejects_risk_json(tmp_path):
    """CDN 风控 JSON 不能被当成音频交付。"""
    p = tmp_path / 'x.mp3'
    p.write_bytes(b'{"code": 403, "msg": "risk"}')
    assert _sniff_audio_header(p) is False


def test_sniff_rejects_html_error_page(tmp_path):
    p = tmp_path / 'x.mp3'
    p.write_bytes(b'<!DOCTYPE html><html><body>403</body></html>')
    assert _sniff_audio_header(p) is False


def test_sniff_rejects_empty_file(tmp_path):
    p = tmp_path / 'x.mp3'
    p.write_bytes(b'')
    assert _sniff_audio_header(p) is False


def test_sniff_rejects_missing_file(tmp_path):
    assert _sniff_audio_header(tmp_path / 'nope.mp3') is False