xiaoxia-saas/tests/unit/test_subtitle_domain.py

"""字幕领域模型单元测试."""

from __future__ import annotations

from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline, SubtitleWord


class TestSubtitleWord:
    """SubtitleWord 测试."""

    def test_basic_properties(self):
        word = SubtitleWord(text="你好", start=1.0, end=1.5)
        assert word.text == "你好"
        assert word.start == 1.0
        assert word.end == 1.5
        assert word.duration == 0.5

    def test_duration_zero_when_end_before_start(self):
        word = SubtitleWord(text="test", start=2.0, end=1.0)
        assert word.duration == 0.0

    def test_duration_zero_when_same_time(self):
        word = SubtitleWord(text="test", start=1.0, end=1.0)
        assert word.duration == 0.0


class TestSubtitleSegment:
    """SubtitleSegment 测试."""

    def test_basic_properties(self):
        seg = SubtitleSegment(text="大家好", start=0.0, end=2.0)
        assert seg.text == "大家好"
        assert seg.start == 0.0
        assert seg.end == 2.0
        assert seg.duration == 2.0
        assert seg.char_count == 3
        assert seg.words == []

    def test_duration_with_words(self):
        words = [
            SubtitleWord(text="大", start=0.0, end=0.5),
            SubtitleWord(text="家", start=0.5, end=1.0),
            SubtitleWord(text="好", start=1.0, end=1.5),
        ]
        seg = SubtitleSegment(text="大家好", start=0.0, end=1.5, words=words)
        assert seg.duration == 1.5
        assert seg.char_count == 3
        assert len(seg.words) == 3

    def test_duration_zero_when_end_before_start(self):
        seg = SubtitleSegment(text="test", start=3.0, end=1.0)
        assert seg.duration == 0.0


class TestSubtitleTimelineBasics:
    """SubtitleTimeline 基础属性测试."""

    def test_empty_timeline(self):
        tl = SubtitleTimeline()
        assert tl.segment_count == 0
        assert tl.total_chars == 0
        assert tl.language == "zh"
        assert tl.total_duration == 0.0

    def test_single_segment(self):
        seg = SubtitleSegment(text="测试", start=0.0, end=1.0)
        tl = SubtitleTimeline(segments=[seg])
        assert tl.segment_count == 1
        assert tl.total_chars == 2

    def test_multiple_segments(self):
        segs = [
            SubtitleSegment(text="第一句", start=0.0, end=1.0),
            SubtitleSegment(text="第二句", start=1.0, end=2.0),
            SubtitleSegment(text="第三句", start=2.0, end=3.0),
        ]
        tl = SubtitleTimeline(segments=segs, total_duration=3.0)
        assert tl.segment_count == 3
        assert tl.total_chars == 9
        assert tl.total_duration == 3.0

    def test_custom_language(self):
        tl = SubtitleTimeline(language="en")
        assert tl.language == "en"


class TestSubtitleTimelineMergeShort:
    """合并短字幕片段测试."""

    def test_empty_or_single_no_change(self):
        tl = SubtitleTimeline()
        result = tl.merge_short_segments()
        assert result.segment_count == 0

        seg = SubtitleSegment(text="短", start=0.0, end=0.5)
        tl2 = SubtitleTimeline(segments=[seg])
        result2 = tl2.merge_short_segments()
        assert result2.segment_count == 1

    def test_merge_short_segments(self):
        segs = [
            SubtitleSegment(text="你好", start=0.0, end=0.5),
            SubtitleSegment(text="世界", start=0.5, end=1.0),
            SubtitleSegment(text="今天天气很好", start=1.0, end=2.0),
        ]
        tl = SubtitleTimeline(segments=segs)
        result = tl.merge_short_segments(min_chars=4)
        # "你好"+"世界"=4字，合并；"今天天气很好"=6字，保留
        assert result.segment_count == 2
        assert result.segments[0].text == "你好世界"
        assert result.segments[0].start == 0.0
        assert result.segments[0].end == 1.0
        assert result.segments[1].text == "今天天气很好"

    def test_merge_trailing_short_to_last(self):
        segs = [
            SubtitleSegment(text="一二三四五六七八", start=0.0, end=1.0),
            SubtitleSegment(text="短", start=1.0, end=1.2),
            SubtitleSegment(text="尾", start=1.2, end=1.4),
        ]
        tl = SubtitleTimeline(segments=segs)
        result = tl.merge_short_segments(min_chars=4)
        # "一二三四五六七八"=8字 → 保留
        # "短"+"尾"=2字 < 4 → 合并到上一段
        assert result.segment_count == 1
        assert result.segments[0].text == "一二三四五六七八短尾"

    def test_merge_with_words(self):
        words1 = [SubtitleWord(text="你", start=0.0, end=0.25), SubtitleWord(text="好", start=0.25, end=0.5)]
        words2 = [SubtitleWord(text="世", start=0.5, end=0.75), SubtitleWord(text="界", start=0.75, end=1.0)]
        segs = [
            SubtitleSegment(text="你好", start=0.0, end=0.5, words=words1),
            SubtitleSegment(text="世界", start=0.5, end=1.0, words=words2),
        ]
        tl = SubtitleTimeline(segments=segs)
        result = tl.merge_short_segments(min_chars=4)
        assert result.segment_count == 1
        assert len(result.segments[0].words) == 4

    def test_preserves_language_and_duration(self):
        segs = [SubtitleSegment(text="短", start=0.0, end=0.5)]
        tl = SubtitleTimeline(segments=segs, language="ja", total_duration=0.5)
        result = tl.merge_short_segments(min_chars=4)
        assert result.language == "ja"
        assert result.total_duration == 0.5


class TestSubtitleTimelineSplitLong:
    """拆分长字幕片段测试."""

    def test_short_segments_no_change(self):
        segs = [SubtitleSegment(text="短句", start=0.0, end=1.0)]
        tl = SubtitleTimeline(segments=segs)
        result = tl.split_long_segments(max_chars=10)
        assert result.segment_count == 1

    def test_split_by_sentence_punctuation(self):
        text = "今天天气很好。我们出去散步吧！"
        seg = SubtitleSegment(text=text, start=0.0, end=10.0)
        tl = SubtitleTimeline(segments=[seg])
        result = tl.split_long_segments(max_chars=8)
        assert result.segment_count >= 2
        assert result.segments[0].text.endswith("。")
        assert result.total_chars == len(text)

    def test_split_long_text_no_punctuation_hard_cut(self):
        text = "一二三四五六七八九十十一十二十三十四十五十六十七十八"
        seg = SubtitleSegment(text=text, start=0.0, end=10.0)
        tl = SubtitleTimeline(segments=[seg])
        result = tl.split_long_segments(max_chars=8)
        assert result.segment_count > 1
        # 所有片段都不超过 max_chars
        for s in result.segments:
            assert s.char_count <= 8

    def test_split_time_proportional(self):
        text = "一二三四。五六七八。"
        seg = SubtitleSegment(text=text, start=0.0, end=10.0)
        tl = SubtitleTimeline(segments=[seg])
        result = tl.split_long_segments(max_chars=4)
        assert result.segment_count >= 2
        # 总时长保持一致
        assert abs(result.segments[-1].end - 10.0) < 0.01

    def test_split_with_words(self):
        words = [
            SubtitleWord(text="一", start=0.0, end=0.5),
            SubtitleWord(text="二", start=0.5, end=1.0),
            SubtitleWord(text="三", start=1.0, end=1.5),
            SubtitleWord(text="四", start=1.5, end=2.0),
        ]
        text = "一二三四五六七八"
        seg = SubtitleSegment(text=text, start=0.0, end=4.0, words=words)
        tl = SubtitleTimeline(segments=[seg])
        result = tl.split_long_segments(max_chars=4)
        assert result.segment_count >= 2
        # 词的总数应该不变
        total_words = sum(len(s.words) for s in result.segments)
        assert total_words == 4

    def test_split_preserves_language(self):
        seg = SubtitleSegment(text="test", start=0.0, end=1.0)
        tl = SubtitleTimeline(segments=[seg], language="en")
        result = tl.split_long_segments(max_chars=2)
        assert result.language == "en"


class TestSplitTextByPunctuation:
    """标点拆分静态方法测试."""

    def test_empty_text(self):
        result = SubtitleTimeline._split_text_by_punctuation("", 10)
        assert result == []

    def test_short_text_no_split(self):
        result = SubtitleTimeline._split_text_by_punctuation("短文本", 10)
        assert len(result) == 1

    def test_split_by_period(self):
        result = SubtitleTimeline._split_text_by_punctuation("第一句。第二句。", 4)
        assert len(result) >= 2
        assert "。" in result[0]

    def test_split_by_exclamation(self):
        result = SubtitleTimeline._split_text_by_punctuation("你好！世界！", 3)
        assert len(result) >= 2

    def test_split_by_comma_when_long(self):
        text = "这是一个很长的句子，中间有逗号分隔，后面还有内容"
        result = SubtitleTimeline._split_text_by_punctuation(text, 8)
        assert len(result) >= 2

    def test_no_punctuation_hard_cut(self):
        text = "一二三四五六七八九十十一十二十三十四十五"
        result = SubtitleTimeline._split_text_by_punctuation(text, 8)
        assert len(result) > 1
        for part in result:
            assert len(part) <= 8

    def test_sentence_end_triggers_split_when_half_max(self):
        # 句末标点在 max_chars//2 以上就拆分
        text = "你好世界。abcdefghij"
        result = SubtitleTimeline._split_text_by_punctuation(text, 10)
        # "你好世界。"=5字 < 10但>=5(half)，应该拆分
        assert len(result) >= 2


class TestMergeSegments:
    """_merge_segments 静态方法测试."""

    def test_merge_empty(self):
        result = SubtitleTimeline._merge_segments([])
        assert result.text == ""
        assert result.start == 0
        assert result.end == 0

    def test_merge_single(self):
        seg = SubtitleSegment(text="test", start=1.0, end=2.0)
        result = SubtitleTimeline._merge_segments([seg])
        assert result.text == "test"
        assert result.start == 1.0
        assert result.end == 2.0

    def test_merge_multiple(self):
        segs = [
            SubtitleSegment(text="第一", start=0.0, end=1.0),
            SubtitleSegment(text="第二", start=1.0, end=2.0),
        ]
        result = SubtitleTimeline._merge_segments(segs)
        assert result.text == "第一第二"
        assert result.start == 0.0
        assert result.end == 2.0


# ── merge_short_segments 更多边界 ──────────────────────────────────────


class TestMergeShortSegmentsMore:
    """merge_short_segments 补充边界测试."""

    def test_all_short_segments_merged_into_one(self):
        """所有片段都很短，合并成一段."""
        segs = [
            SubtitleSegment(text="你", start=0.0, end=0.1),
            SubtitleSegment(text="好", start=0.1, end=0.2),
            SubtitleSegment(text="世", start=0.2, end=0.3),
            SubtitleSegment(text="界", start=0.3, end=0.4),
        ]
        tl = SubtitleTimeline(segments=segs)
        result = tl.merge_short_segments(min_chars=8)
        assert result.segment_count == 1
        assert result.segments[0].text == "你好世界"
        assert result.segments[0].start == 0.0
        assert result.segments[0].end == 0.4

    def test_min_chars_very_small(self):
        """min_chars 很小（1），几乎不合并."""
        segs = [
            SubtitleSegment(text="一", start=0.0, end=0.5),
            SubtitleSegment(text="二", start=0.5, end=1.0),
            SubtitleSegment(text="三", start=1.0, end=1.5),
        ]
        tl = SubtitleTimeline(segments=segs)
        result = tl.merge_short_segments(min_chars=1)
        # 每个1字 >= min_chars(1)，所以都独立成段
        assert result.segment_count == 3

    def test_min_chars_very_large(self):
        """min_chars 很大，全部合并."""
        segs = [
            SubtitleSegment(text="第一句很长的内容", start=0.0, end=1.0),
            SubtitleSegment(text="第二句也不短", start=1.0, end=2.0),
            SubtitleSegment(text="第三句还行", start=2.0, end=3.0),
        ]
        tl = SubtitleTimeline(segments=segs)
        result = tl.merge_short_segments(min_chars=100)
        assert result.segment_count == 1
        assert result.total_chars == tl.total_chars

    def test_merge_preserves_total_duration(self):
        """合并后总时长不变."""
        segs = [
            SubtitleSegment(text="短1", start=0.0, end=0.5),
            SubtitleSegment(text="短2", start=0.5, end=1.0),
            SubtitleSegment(text="很长的一段内容", start=1.0, end=3.0),
        ]
        tl = SubtitleTimeline(segments=segs, total_duration=3.0)
        result = tl.merge_short_segments(min_chars=4)
        assert result.total_duration == 3.0

    def test_merge_preserves_word_level_info(self):
        """合并后词级信息完整保留."""
        w1 = [SubtitleWord(text="你", start=0.0, end=0.3)]
        w2 = [SubtitleWord(text="好", start=0.3, end=0.6)]
        segs = [
            SubtitleSegment(text="你", start=0.0, end=0.3, words=w1),
            SubtitleSegment(text="好", start=0.3, end=0.6, words=w2),
        ]
        tl = SubtitleTimeline(segments=segs)
        result = tl.merge_short_segments(min_chars=4)
        assert result.segment_count == 1
        assert len(result.segments[0].words) == 2
        assert result.segments[0].words[0].text == "你"
        assert result.segments[0].words[1].text == "好"

    def test_trailing_short_merged_to_last(self):
        """尾部极短段合并到最后一段."""
        segs = [
            SubtitleSegment(text="一二三四五六七八九十", start=0.0, end=1.0),
            SubtitleSegment(text="尾", start=1.0, end=1.1),
        ]
        tl = SubtitleTimeline(segments=segs)
        result = tl.merge_short_segments(min_chars=5)
        # 第一段10字 >= 5 → 先入merged
        # "尾" = 1字 < 5 → 进入buffer
        # 循环结束，buffer非空，检查是否合并到最后一段
        # merged[-1] = 10字 + 1字 = 11字 <= max_chars(这里没限制)
        assert result.segment_count == 1
        assert result.segments[0].text == "一二三四五六七八九十尾"


# ── split_long_segments 更多边界 ──────────────────────────────────────


class TestSplitLongSegmentsMore:
    """split_long_segments 补充边界测试."""

    def test_exactly_max_chars_no_split(self):
        """恰好等于 max_chars 不拆分."""
        text = "一二三四五六七八九十"  # 10字
        seg = SubtitleSegment(text=text, start=0.0, end=5.0)
        tl = SubtitleTimeline(segments=[seg])
        result = tl.split_long_segments(max_chars=10)
        assert result.segment_count == 1

    def test_one_over_max_chars_splits(self):
        """超过1个字符就拆分."""
        text = "一二三四五六七八九十一"  # 11字
        seg = SubtitleSegment(text=text, start=0.0, end=5.0)
        tl = SubtitleTimeline(segments=[seg])
        result = tl.split_long_segments(max_chars=10)
        assert result.segment_count >= 2
        assert result.total_chars == 11

    def test_mixed_short_and_long_segments(self):
        """长短片段混合，只拆分长的."""
        segs = [
            SubtitleSegment(text="短", start=0.0, end=0.5),
            SubtitleSegment(text="很长很长很长的一段文字内容超过限制", start=0.5, end=3.0),
            SubtitleSegment(text="也短", start=3.0, end=3.5),
        ]
        tl = SubtitleTimeline(segments=segs)
        result = tl.split_long_segments(max_chars=10)
        # 第1段和第3段保持，第2段被拆分
        assert result.segment_count > 3
        assert result.segments[0].text == "短"
        assert result.segments[-1].text == "也短"

    def test_total_chars_preserved_after_split(self):
        """拆分后总字符数不变."""
        text = "这是第一段。这是第二段很长的内容。还有第三段。"
        seg = SubtitleSegment(text=text, start=0.0, end=10.0)
        tl = SubtitleTimeline(segments=[seg])
        result = tl.split_long_segments(max_chars=8)
        assert result.total_chars == len(text)

    def test_total_duration_preserved(self):
        """拆分后最后一段的结束时间等于原始结束时间."""
        text = "一二三四。五六七八。九十一二。"
        seg = SubtitleSegment(text=text, start=1.0, end=5.0)
        tl = SubtitleTimeline(segments=[seg])
        result = tl.split_long_segments(max_chars=5)
        assert abs(result.segments[-1].end - 5.0) < 0.01
        assert abs(result.segments[0].start - 1.0) < 0.01


# ── _split_text_by_punctuation 更多边界 ───────────────────────────────


class TestSplitTextByPunctuationMore:
    """_split_text_by_punctuation 补充边界测试."""

    def test_comma_splits_when_over_max(self):
        """逗号在超长时触发拆分."""
        text = "一二三四五六七八，八七六五四三二一"
        result = SubtitleTimeline._split_text_by_punctuation(text, 10)
        assert len(result) >= 2
        assert "，" in result[0]

    def test_colon_semicolon_splits(self):
        """冒号分号也能触发拆分."""
        text = "首先：第一点内容；第二点内容更多一些。"
        result = SubtitleTimeline._split_text_by_punctuation(text, 8)
        assert len(result) >= 2

    def test_punctuation_at_start(self):
        """标点在开头位置."""
        text = "。这是正文内容开始了"
        result = SubtitleTimeline._split_text_by_punctuation(text, 10)
        # 开头的句号在 max_chars//2 之前，不会触发句末标点拆分
        # 但整个文本不长，可能不拆
        assert len(result) >= 1

    def test_consecutive_punctuation(self):
        """连续多个标点符号."""
        text = "你好！！！世界？？？"
        result = SubtitleTimeline._split_text_by_punctuation(text, 5)
        assert len(result) >= 1

    def test_english_sentence_end(self):
        """英文句末标点也触发拆分."""
        text = "Hello world. This is a test sentence with enough words."
        result = SubtitleTimeline._split_text_by_punctuation(text, 15)
        assert len(result) >= 2

    def test_no_punctuation_hard_cut_evenly(self):
        """没有标点时均匀硬切."""
        text = "一二三四五六七八九十一二三四五六七八九十一二三四五"  # 25字
        result = SubtitleTimeline._split_text_by_punctuation(text, 10)
        assert len(result) == 3
        assert len(result[0]) == 10
        assert len(result[1]) == 10
        assert len(result[2]) == 5

    def test_single_character_text(self):
        """单字符文本."""
        result = SubtitleTimeline._split_text_by_punctuation("好", 10)
        assert result == ["好"]

    def test_only_punctuation(self):
        """只有标点符号."""
        result = SubtitleTimeline._split_text_by_punctuation("。！？", 10)
        assert len(result) == 1

    def test_mixed_fullwidth_halfwidth_punctuation(self):
        """全角半角标点混合."""
        text = "你好。再见!谢谢？不用谢。"
        result = SubtitleTimeline._split_text_by_punctuation(text, 5)
        assert len(result) >= 2