theme_generation/reports.py

"""Markdown reports for generation runs."""

from __future__ import annotations

from pathlib import Path
from typing import Any

from .common import COMMON_DURATION_BEATS, Symbol


def write_generation_report(
    *,
    output_dir: Path,
    title: str,
    description: str,
    settings: dict[str, Any],
    stats: dict[str, Any],
    generated: list[tuple[Symbol, ...]],
    write_abc: bool,
    write_musicxml: bool = False,
) -> None:
    lines = [
        f"# {title}",
        "",
        description,
        "",
        "## Settings",
        "",
    ]
    for key, value in settings.items():
        lines.append(f"- {key}: {value}")

    lines += [
        "",
        "## Common Training Symbols",
        "",
        "| symbol | count |",
        "| --- | ---: |",
    ]
    for symbol, count in stats["top_symbols"]:
        lines.append(f"| `{symbol}` | {count} |")

    lines += ["", "## Common Durations", "", "| duration | count |", "| --- | ---: |"]
    for duration, count in stats["top_durations"]:
        lines.append(f"| {duration} | {count} |")

    lines += ["", "## Generated Samples", ""]
    for index, sequence in enumerate(generated, start=1):
        rpc_text = " ".join(str(symbol.rpc) for symbol in sequence)
        dur_text = " ".join(symbol.duration for symbol in sequence)
        lines += [
            f"### Sample {index:02d}",
            "",
            f"- MIDI: `generated_{index:02d}.mid`",
        ]
        if write_abc:
            lines.append(f"- ABC: `generated_{index:02d}.abc`")
        if write_musicxml:
            lines.append(f"- MusicXML: `generated_{index:02d}.musicxml`")
        lines += [
            f"- relative pcs: `{rpc_text}`",
            f"- durations: `{dur_text}`",
            "",
        ]

    (output_dir / "report.md").write_text("\n".join(lines), encoding="utf-8")


def format_allowed_durations(allowed_durations: set[str]) -> str:
    return ", ".join(sorted(allowed_durations, key=COMMON_DURATION_BEATS.get))