test: wave79 - subtitle_generator纯函数深度测试 + text_splitter边界补充 (+62)

- subtitle_generator: 新增52个测试
  - _hex_to_ass_color 颜色转换（9个）
  - _position_to_ass_alignment 位置映射（6个）
  - _format_ass_time 时间格式化（7个）
  - _escape_ass_text 转义处理（8个）
  - _wrap_text 换行补充（11个）
  - generate_ass_from_timeline 补充（11个）
- text_splitter: 新增10个边界测试
  - max_chars=1、连续换行、纯换行符
  - 首尾空白、超长单句10段
  - 混合标点、尾段合并、短句合并
  - 长纯标点、制表符不是句子结束符

tests/unit/test_subtitle_generator.py

@@ -176,3 +176,582 @@ class TestGenerateAssFromTimeline:
             assert "0:00:01.25" in content
             # 1:01:01.00 格式（3661秒 = 1小时1分1秒）
             assert "1:01:01.00" in content
+
+
+# ── _hex_to_ass_color 深度测试 ──────────────────────────────────────────────
+
+
+class TestHexToAssColor:
+    """_hex_to_ass_color 颜色转换测试"""
+
+    def test_standard_hex_with_hash(self):
+        """带#号的标准6位HEX"""
+        from apps.worker.video_processing.subtitle_generator import _hex_to_ass_color
+
+        # #RRGGBB → &HBBGGRR
+        assert _hex_to_ass_color("#FF0000") == "&H0000FF"  # 红
+        assert _hex_to_ass_color("#00FF00") == "&H00FF00"  # 绿
+        assert _hex_to_ass_color("#0000FF") == "&HFF0000"  # 蓝
+
+    def test_standard_hex_without_hash(self):
+        """不带#号的6位HEX"""
+        from apps.worker.video_processing.subtitle_generator import _hex_to_ass_color
+
+        assert _hex_to_ass_color("FF0000") == "&H0000FF"
+        assert _hex_to_ass_color("00FF00") == "&H00FF00"
+        assert _hex_to_ass_color("0000FF") == "&HFF0000"
+
+    def test_white_and_black(self):
+        """白色和黑色"""
+        from apps.worker.video_processing.subtitle_generator import _hex_to_ass_color
+
+        assert _hex_to_ass_color("#FFFFFF") == "&HFFFFFF"  # 白
+        assert _hex_to_ass_color("#000000") == "&H000000"  # 黑
+
+    def test_lowercase_hex(self):
+        """小写字母HEX"""
+        from apps.worker.video_processing.subtitle_generator import _hex_to_ass_color
+
+        result = _hex_to_ass_color("#aabbcc")
+        # 输出是大写的
+        assert result == "&HCCBBAA"
+
+    def test_mixed_case_hex(self):
+        """大小写混合HEX"""
+        from apps.worker.video_processing.subtitle_generator import _hex_to_ass_color
+
+        result = _hex_to_ass_color("#AaBbCc")
+        assert result == "&HCCBBAA"
+
+    def test_short_hex_returns_default(self):
+        """长度不足6位返回默认白色"""
+        from apps.worker.video_processing.subtitle_generator import _hex_to_ass_color
+
+        assert _hex_to_ass_color("#FFF") == "&H00FFFFFF"  # 3位
+        assert _hex_to_ass_color("FF") == "&H00FFFFFF"  # 2位
+
+    def test_long_hex_returns_default(self):
+        """长度超过6位返回默认白色"""
+        from apps.worker.video_processing.subtitle_generator import _hex_to_ass_color
+
+        assert _hex_to_ass_color("#AABBCCDD") == "&H00FFFFFF"
+
+    def test_empty_string_returns_default(self):
+        """空字符串返回默认白色"""
+        from apps.worker.video_processing.subtitle_generator import _hex_to_ass_color
+
+        assert _hex_to_ass_color("") == "&H00FFFFFF"
+
+    def test_gray_color(self):
+        """灰色调"""
+        from apps.worker.video_processing.subtitle_generator import _hex_to_ass_color
+
+        assert _hex_to_ass_color("#808080") == "&H808080"
+
+
+# ── _position_to_ass_alignment 测试 ────────────────────────────────────────
+
+
+class TestPositionToAssAlignment:
+    """_position_to_ass_alignment 位置映射测试"""
+
+    def test_top_maps_to_8(self):
+        """顶部 → 8"""
+        from apps.worker.video_processing.subtitle_generator import _position_to_ass_alignment
+
+        assert _position_to_ass_alignment("top") == 8
+
+    def test_center_maps_to_5(self):
+        """居中 → 5"""
+        from apps.worker.video_processing.subtitle_generator import _position_to_ass_alignment
+
+        assert _position_to_ass_alignment("center") == 5
+
+    def test_bottom_maps_to_2(self):
+        """底部 → 2"""
+        from apps.worker.video_processing.subtitle_generator import _position_to_ass_alignment
+
+        assert _position_to_ass_alignment("bottom") == 2
+
+    def test_unknown_position_defaults_to_bottom(self):
+        """未知位置默认底部(2)"""
+        from apps.worker.video_processing.subtitle_generator import _position_to_ass_alignment
+
+        assert _position_to_ass_alignment("left") == 2
+        assert _position_to_ass_alignment("right") == 2
+        assert _position_to_ass_alignment("middle") == 2
+
+    def test_empty_string_defaults_to_bottom(self):
+        """空字符串默认底部"""
+        from apps.worker.video_processing.subtitle_generator import _position_to_ass_alignment
+
+        assert _position_to_ass_alignment("") == 2
+
+    def test_uppercase_not_matched(self):
+        """大写不匹配，走默认"""
+        from apps.worker.video_processing.subtitle_generator import _position_to_ass_alignment
+
+        assert _position_to_ass_alignment("TOP") == 2
+        assert _position_to_ass_alignment("CENTER") == 2
+
+
+# ── _format_ass_time 测试 ──────────────────────────────────────────────────
+
+
+class TestFormatAssTime:
+    """_format_ass_time 时间格式化测试"""
+
+    def test_zero_seconds(self):
+        """0秒"""
+        from apps.worker.video_processing.subtitle_generator import _format_ass_time
+
+        assert _format_ass_time(0.0) == "0:00:00.00"
+
+    def test_fractional_seconds(self):
+        """小数秒"""
+        from apps.worker.video_processing.subtitle_generator import _format_ass_time
+
+        assert _format_ass_time(0.5) == "0:00:00.50"
+        assert _format_ass_time(1.25) == "0:00:01.25"
+
+    def test_whole_seconds(self):
+        """整秒"""
+        from apps.worker.video_processing.subtitle_generator import _format_ass_time
+
+        assert _format_ass_time(5.0) == "0:00:05.00"
+        assert _format_ass_time(30.0) == "0:00:30.00"
+
+    def test_minutes_level(self):
+        """分钟级"""
+        from apps.worker.video_processing.subtitle_generator import _format_ass_time
+
+        assert _format_ass_time(60.0) == "0:01:00.00"
+        assert _format_ass_time(90.5) == "0:01:30.50"
+        assert _format_ass_time(599.0) == "0:09:59.00"
+
+    def test_hours_level(self):
+        """小时级"""
+        from apps.worker.video_processing.subtitle_generator import _format_ass_time
+
+        assert _format_ass_time(3600.0) == "1:00:00.00"
+        assert _format_ass_time(3661.5) == "1:01:01.50"
+        assert _format_ass_time(7384.0) == "2:03:04.00"
+
+    def test_centisecond_precision(self):
+        """百分秒精度"""
+        from apps.worker.video_processing.subtitle_generator import _format_ass_time
+
+        assert _format_ass_time(1.01) == "0:00:01.01"
+        assert _format_ass_time(1.99) == "0:00:01.99"
+
+    def test_sub_centisecond_truncated(self):
+        """毫秒级精度会被格式化截断到百分秒"""
+        from apps.worker.video_processing.subtitle_generator import _format_ass_time
+
+        # Python 的 %05.2f 会四舍五入
+        result = _format_ass_time(1.123)
+        assert result.startswith("0:00:01.")
+        assert len(result.split(".")[-1]) == 2
+
+
+# ── _escape_ass_text 测试 ──────────────────────────────────────────────────
+
+
+class TestEscapeAssText:
+    """_escape_ass_text 转义测试"""
+
+    def test_plain_text_unchanged(self):
+        """普通文本不改变"""
+        from apps.worker.video_processing.subtitle_generator import _escape_ass_text
+
+        assert _escape_ass_text("你好世界") == "你好世界"
+        assert _escape_ass_text("Hello World") == "Hello World"
+
+    def test_newline_to_ass_newline(self):
+        """\\n 转 \\N"""
+        from apps.worker.video_processing.subtitle_generator import _escape_ass_text
+
+        assert _escape_ass_text("第一行\n第二行") == "第一行\\N第二行"
+
+    def test_crlf_to_ass_newline(self):
+        """\\r\\n 转 \\N"""
+        from apps.worker.video_processing.subtitle_generator import _escape_ass_text
+
+        assert _escape_ass_text("第一行\r\n第二行") == "第一行\\N第二行"
+
+    def test_cr_to_ass_newline(self):
+        """\\r 转 \\N"""
+        from apps.worker.video_processing.subtitle_generator import _escape_ass_text
+
+        assert _escape_ass_text("第一行\r第二行") == "第一行\\N第二行"
+
+    def test_curly_braces_escaped(self):
+        """花括号转圆括号（防止ASS标签注入）"""
+        from apps.worker.video_processing.subtitle_generator import _escape_ass_text
+
+        assert _escape_ass_text("{text}") == "(text)"
+        assert _escape_ass_text("{{double}}") == "((double))"
+
+    def test_mixed_escapes(self):
+        """混合转义"""
+        from apps.worker.video_processing.subtitle_generator import _escape_ass_text
+
+        text = "你好\n世界{测试}\r\n结束"
+        result = _escape_ass_text(text)
+        assert "\\N" in result
+        assert "(测试)" in result
+        assert "\n" not in result
+        assert "{" not in result
+
+    def test_empty_string(self):
+        """空字符串"""
+        from apps.worker.video_processing.subtitle_generator import _escape_ass_text
+
+        assert _escape_ass_text("") == ""
+
+    def test_only_braces(self):
+        """只有花括号"""
+        from apps.worker.video_processing.subtitle_generator import _escape_ass_text
+
+        assert _escape_ass_text("{}") == "()"
+
+
+# ── _wrap_text 补充测试 ────────────────────────────────────────────────────
+
+
+class TestWrapTextMore:
+    """_wrap_text 补充边界测试"""
+
+    def test_empty_string_returns_empty_list(self):
+        """空字符串返回空列表"""
+        from apps.worker.video_processing.subtitle_generator import _wrap_text
+
+        assert _wrap_text("", 10) == [""]
+
+    def test_single_character(self):
+        """单字符"""
+        from apps.worker.video_processing.subtitle_generator import _wrap_text
+
+        assert _wrap_text("好", 10) == ["好"]
+
+    def test_max_chars_equals_one(self):
+        """max_chars=1 每个字一行"""
+        from apps.worker.video_processing.subtitle_generator import _wrap_text
+
+        result = _wrap_text("一二三四", 1)
+        assert len(result) == 4
+        assert result[0] == "一"
+        assert result[1] == "二"
+
+    def test_punctuation_in_middle(self):
+        """标点在正中间优先从标点断开"""
+        from apps.worker.video_processing.subtitle_generator import _wrap_text
+
+        # 前10字内有句号，优先在句号后断开
+        text = "一二三四五六七八九十。后面的内容继续写下去"
+        result = _wrap_text(text, 15)
+        # 第一行应该包含句号
+        assert "。" in result[0]
+
+    def test_punctuation_at_start_ignored(self):
+        """标点在开头位置（前半部分）不会触发断开"""
+        from apps.worker.video_processing.subtitle_generator import _wrap_text
+
+        text = "。一二三四五六七八九十"
+        result = _wrap_text(text, 10)
+        # 标点在第0位，不会在max_chars//2到max_chars范围内
+        assert len(result) == 2
+        assert len(result[0]) == 10
+
+    def test_no_punctuation_long_text(self):
+        """完全没有标点的长文本硬切"""
+        from apps.worker.video_processing.subtitle_generator import _wrap_text
+
+        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"
+        result = _wrap_text(text, 10)
+        assert len(result) == 3
+        assert len(result[0]) == 10
+        assert len(result[1]) == 10
+        assert len(result[2]) == 5
+
+    def test_exactly_two_lines(self):
+        """恰好两行"""
+        from apps.worker.video_processing.subtitle_generator import _wrap_text
+
+        text = "一" * 20
+        result = _wrap_text(text, 10)
+        assert len(result) == 2
+        assert len(result[0]) == 10
+        assert len(result[1]) == 10
+
+    def test_three_lines(self):
+        """三行"""
+        from apps.worker.video_processing.subtitle_generator import _wrap_text
+
+        text = "一" * 25
+        result = _wrap_text(text, 10)
+        assert len(result) == 3
+        assert len(result[0]) == 10
+        assert len(result[1]) == 10
+        assert len(result[2]) == 5
+
+    def test_multiple_punctuation_points(self):
+        """多个标点，选择最后一个合适的"""
+        from apps.worker.video_processing.subtitle_generator import _wrap_text
+
+        text = "你好。再见。谢谢。抱歉。好的不行了"
+        result = _wrap_text(text, 12)
+        # 应该在最靠后的（在范围内的）标点处断开
+        assert "。" in result[0]
+
+    def test_english_punctuation_wrap(self):
+        """英文标点也会触发换行"""
+        from apps.worker.video_processing.subtitle_generator import _wrap_text
+
+        text = "Hello world. This is a test sentence."
+        result = _wrap_text(text, 20)
+        assert len(result) >= 2
+        assert result[0].endswith(".") or "." in result[0]
+
+    def test_total_length_preserved(self):
+        """换行后总字符数不变"""
+        from apps.worker.video_processing.subtitle_generator import _wrap_text
+
+        text = "这是一段用于测试换行功能的中文文本，包含了各种标点符号。看看效果如何？"
+        result = _wrap_text(text, 10)
+        assert "".join(result) == text
+
+
+# ── generate_ass_from_timeline 补充测试 ────────────────────────────────────
+
+
+class TestGenerateAssFromTimelineMore:
+    """generate_ass_from_timeline 补充测试"""
+
+    def test_single_segment(self):
+        """单段字幕"""
+        from apps.worker.video_processing.subtitle_generator import generate_ass_from_timeline
+        from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline
+        import tempfile
+        from pathlib import Path
+
+        timeline = SubtitleTimeline(
+            segments=[SubtitleSegment(text="你好", start=0.0, end=1.0)],
+            total_duration=1.0,
+        )
+        with tempfile.TemporaryDirectory() as tmpdir:
+            output = Path(tmpdir) / "single.ass"
+            generate_ass_from_timeline(output, timeline, video_width=1920, video_height=1080)
+            content = output.read_text(encoding="utf-8")
+            assert "Dialogue:" in content
+            assert "你好" in content
+            assert content.count("Dialogue:") == 1
+
+    def test_multiple_segments(self):
+        """多段字幕"""
+        from apps.worker.video_processing.subtitle_generator import generate_ass_from_timeline
+        from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline
+        import tempfile
+        from pathlib import Path
+
+        timeline = SubtitleTimeline(
+            segments=[SubtitleSegment(text=f"第{i}段", start=float(i), end=float(i + 1)) for i in range(5)],
+            total_duration=5.0,
+        )
+        with tempfile.TemporaryDirectory() as tmpdir:
+            output = Path(tmpdir) / "multi.ass"
+            generate_ass_from_timeline(output, timeline, video_width=1920, video_height=1080)
+            content = output.read_text(encoding="utf-8")
+            assert content.count("Dialogue:") == 5
+
+    def test_long_text_auto_wrap(self):
+        """长字幕自动换行"""
+        from apps.worker.video_processing.subtitle_generator import generate_ass_from_timeline
+        from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline
+        import tempfile
+        from pathlib import Path
+
+        long_text = "这是一段非常非常长的字幕文本，用来测试自动换行功能是否正常工作。"
+        timeline = SubtitleTimeline(
+            segments=[SubtitleSegment(text=long_text, start=0, end=5)],
+            total_duration=5.0,
+        )
+        with tempfile.TemporaryDirectory() as tmpdir:
+            output = Path(tmpdir) / "wrap.ass"
+            generate_ass_from_timeline(
+                output,
+                timeline,
+                video_width=1920,
+                video_height=1080,
+                subtitle_config={"max_chars_per_line": 10},
+            )
+            content = output.read_text(encoding="utf-8")
+            # 应该包含 \N 换行符
+            assert "\\N" in content
+
+    def test_custom_color_hex(self):
+        """自定义颜色正确转换"""
+        from apps.worker.video_processing.subtitle_generator import generate_ass_from_timeline
+        from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline
+        import tempfile
+        from pathlib import Path
+
+        timeline = SubtitleTimeline(
+            segments=[SubtitleSegment(text="红", start=0, end=1)],
+            total_duration=1.0,
+        )
+        with tempfile.TemporaryDirectory() as tmpdir:
+            output = Path(tmpdir) / "red.ass"
+            generate_ass_from_timeline(
+                output,
+                timeline,
+                video_width=1920,
+                video_height=1080,
+                subtitle_config={"color": "#FF0000"},
+            )
+            content = output.read_text(encoding="utf-8")
+            # 红色 #FF0000 → &H0000FF (BBGGRR)
+            assert "&H0000FF" in content
+
+    def test_position_top(self):
+        """顶部位置"""
+        from apps.worker.video_processing.subtitle_generator import generate_ass_from_timeline
+        from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline
+        import tempfile
+        from pathlib import Path
+
+        timeline = SubtitleTimeline(
+            segments=[SubtitleSegment(text="顶部", start=0, end=1)],
+            total_duration=1.0,
+        )
+        with tempfile.TemporaryDirectory() as tmpdir:
+            output = Path(tmpdir) / "top.ass"
+            generate_ass_from_timeline(
+                output,
+                timeline,
+                video_width=1920,
+                video_height=1080,
+                subtitle_config={"position": "top"},
+            )
+            content = output.read_text(encoding="utf-8")
+            # 顶部对齐是 \\an8 → Style 中 Alignment=8
+            assert "1,1.5,0,8," in content or ",0,8," in content
+
+    def test_position_center(self):
+        """居中位置"""
+        from apps.worker.video_processing.subtitle_generator import generate_ass_from_timeline
+        from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline
+        import tempfile
+        from pathlib import Path
+
+        timeline = SubtitleTimeline(
+            segments=[SubtitleSegment(text="居中", start=0, end=1)],
+            total_duration=1.0,
+        )
+        with tempfile.TemporaryDirectory() as tmpdir:
+            output = Path(tmpdir) / "center.ass"
+            generate_ass_from_timeline(
+                output,
+                timeline,
+                video_width=1920,
+                video_height=1080,
+                subtitle_config={"position": "center"},
+            )
+            content = output.read_text(encoding="utf-8")
+            # 居中对齐是 Alignment=5
+            assert ",0,5," in content
+
+    def test_custom_font_size(self):
+        """自定义字体大小"""
+        from apps.worker.video_processing.subtitle_generator import generate_ass_from_timeline
+        from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline
+        import tempfile
+        from pathlib import Path
+
+        timeline = SubtitleTimeline(
+            segments=[SubtitleSegment(text="大", start=0, end=1)],
+            total_duration=1.0,
+        )
+        with tempfile.TemporaryDirectory() as tmpdir:
+            output = Path(tmpdir) / "big.ass"
+            generate_ass_from_timeline(
+                output,
+                timeline,
+                video_width=1920,
+                video_height=1080,
+                subtitle_config={"size": 48},
+            )
+            content = output.read_text(encoding="utf-8")
+            # Style 行中字号应该是48
+            assert "Default,思源黑体,48," in content
+
+    def test_720p_resolution(self):
+        """720p分辨率"""
+        from apps.worker.video_processing.subtitle_generator import generate_ass_from_timeline
+        from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline
+        import tempfile
+        from pathlib import Path
+
+        timeline = SubtitleTimeline(
+            segments=[SubtitleSegment(text="720p", start=0, end=1)],
+            total_duration=1.0,
+        )
+        with tempfile.TemporaryDirectory() as tmpdir:
+            output = Path(tmpdir) / "720p.ass"
+            generate_ass_from_timeline(output, timeline, video_width=1280, video_height=720)
+            content = output.read_text(encoding="utf-8")
+            assert "PlayResX: 1280" in content
+            assert "PlayResY: 720" in content
+
+    def test_special_characters_in_text(self):
+        """字幕文本含特殊字符（花括号、换行）"""
+        from apps.worker.video_processing.subtitle_generator import generate_ass_from_timeline
+        from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline
+        import tempfile
+        from pathlib import Path
+
+        timeline = SubtitleTimeline(
+            segments=[SubtitleSegment(text="你好{tag}\n世界", start=0, end=1)],
+            total_duration=1.0,
+        )
+        with tempfile.TemporaryDirectory() as tmpdir:
+            output = Path(tmpdir) / "special.ass"
+            generate_ass_from_timeline(output, timeline, video_width=1920, video_height=1080)
+            content = output.read_text(encoding="utf-8")
+            # 花括号被转义
+            assert "(tag)" in content
+            # 换行被转义成 \N
+            assert "\\N" in content
+
+    def test_output_path_creates_parent_dirs(self):
+        """自动创建父目录"""
+        from apps.worker.video_processing.subtitle_generator import generate_ass_from_timeline
+        from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline
+        import tempfile
+        from pathlib import Path
+
+        timeline = SubtitleTimeline(
+            segments=[SubtitleSegment(text="测试", start=0, end=1)],
+            total_duration=1.0,
+        )
+        with tempfile.TemporaryDirectory() as tmpdir:
+            output = Path(tmpdir) / "nested" / "deep" / "out.ass"
+            result = generate_ass_from_timeline(output, timeline, video_width=1920, video_height=1080)
+            assert result.exists()
+            assert result.parent.exists()
+
+    def test_returns_output_path(self):
+        """返回值等于输出路径"""
+        from apps.worker.video_processing.subtitle_generator import generate_ass_from_timeline
+        from packages.domain.subtitle import SubtitleSegment, SubtitleTimeline
+        import tempfile
+        from pathlib import Path
+
+        timeline = SubtitleTimeline(
+            segments=[SubtitleSegment(text="返回", start=0, end=1)],
+            total_duration=1.0,
+        )
+        with tempfile.TemporaryDirectory() as tmpdir:
+            output = Path(tmpdir) / "ret.ass"
+            result = generate_ass_from_timeline(output, timeline, video_width=1920, video_height=1080)
+            assert result == output

tests/unit/test_text_splitter.py

@@ -325,3 +325,88 @@ class TestSplitTextEdgeCases:
         assert len(result) > 1
         for seg in result:
             assert len(seg) <= 20
+
+
+# ── 更多边界场景补充 ─────────────────────────────────────────────────────────
+
+
+class TestSplitTextMoreEdgeCases:
+    """更多边界场景补充"""
+
+    def test_max_chars_one(self):
+        """max_chars=1 每个字符一段"""
+        text = "一二三四五"
+        result = split_text(text, max_chars=1)
+        assert len(result) == 5
+        for seg in result:
+            assert len(seg) == 1
+
+    def test_consecutive_newlines(self):
+        """连续多个换行符"""
+        text = "第一段\n\n\n第二段\n\n第三段"
+        result = split_text(text, max_chars=100)
+        # 合并后应该是一段（内容不长且合并逻辑会被合并）
+        assert len(result) >= 1
+        assert "第一段" in result[0]
+        for seg in result:
+            assert len(seg) <= 100
+
+    def test_only_newlines_only(self):
+        """只有换行符（纯空白被strip掉返回空"""
+        assert split_text("\n\n\n\n") == []
+
+    def test_leading_trailing_whitespace(self):
+        """首尾空白被去除"""
+        text = "   你好世界。   "
+        result = split_text(text, max_chars=100)
+        assert result == ["你好世界。"]
+
+    def test_very_long_single_sentence_many_segments(self):
+        """超长单句被切成很多段"""
+        text = "字" * 1000
+        result = split_text(text, max_chars=100)
+        assert len(result) == 10
+        for seg in result:
+            assert len(seg) == 100
+
+    def test_mixed_punctuation_types(self):
+        """全角半角标点混合"""
+        text = "你好!再见。谢谢?抱歉；好的"
+        result = split_text(text, max_chars=200)
+        assert len(result) == 1
+
+    def test_last_segment_short_merged_to_previous(self):
+        """尾部极短段被合并到前一段"""
+        # 构造第一段接近max_chars，结尾有个短句尾巴
+        long_part = "一二三四五六七八九十" * 9 + "。"  # ~90字
+        tail = "完"  # 1字
+        text = long_part + tail
+        result = split_text(text, max_chars=100)
+        # 尾巴应该被合并
+        combined = "".join(result)
+        assert combined == text.strip()
+        assert len(result) <= 2
+
+    def test_all_short_sentences_merged_into_one(self):
+        """大量短句全部合并成一段"""
+        sentences = ["你好。", "我好。", "他好。", "大家好。", "才是真的好。"]
+        text = "".join(sentences)
+        result = split_text(text, max_chars=200)
+        assert len(result) == 1
+
+    def test_punctuation_only_long(self):
+        """很长的纯标点文本"""
+        text = "。" * 200
+        result = split_text(text, max_chars=50)
+        assert len(result) >= 4
+        for seg in result:
+            assert len(seg) <= 50
+
+    def test_tab_not_sentence_end(self):
+        """制表符不是句子结束符"""
+        text = "这是一段\t包含制表符的文本内容" + "字" * 100
+        result = split_text(text, max_chars=50)
+        # 制表符不在句子结束符集合中，不会触发分段
+        # 制表符会保留在分段内容中
+        has_tab = any("\t" in seg for seg in result)
+        assert has_tab