def _srt_time(seconds: float) -> str:
    total = int(seconds)
    ms = int((seconds - total) * 1000)
    h, m, s = total // 3600, (total % 3600) // 60, total % 60
    return f"{h:02}:{m:02}:{s:02},{ms:03d}"


def _segment_fields(segment) -> tuple:
    if isinstance(segment, dict):
        return segment["start"], segment["end"], segment["text"], segment.get("speaker")
    return segment.start, segment.end, segment.text, None


def to_srt(segments: list, with_speaker: bool = True) -> str:
    lines = []
    for i, seg in enumerate(segments, start=1):
        start, end, text, speaker = _segment_fields(seg)
        lines.append(str(i))
        lines.append(f"{_srt_time(start)} --> {_srt_time(end)}")
        if with_speaker and speaker:
            lines.append(f"({speaker}): {text.strip()}")
        else:
            lines.append(text.strip())
        lines.append("")
    return "\n".join(lines)


def to_txt(segments: list, with_speaker: bool = True) -> str:
    lines = []
    for seg in segments:
        _, _, text, speaker = _segment_fields(seg)
        if with_speaker and speaker:
            lines.append(f"({speaker}): {text.strip()}")
        else:
            lines.append(text.strip())
    return "\n".join(lines)