changelog_from_update_log 実行例

help出力 changelog_from_update_log.py --help

usage: changelog_from_update_log.py [-h] [-o OUTPUT] [--root-dir1 ROOT_DIR1]
                                    [--root-dir2 ROOT_DIR2] [--ai-lib AI_LIB]
                                    [--config CONFIG] [--provider PROVIDER]
                                    [--model MODEL] [--language {ja,en}]
                                    [--max-direct-chars MAX_DIRECT_CHARS]
                                    [--source-chunk-chars SOURCE_CHUNK_CHARS]
                                    [--temperature TEMPERATURE]
                                    [--reasoning-effort {minimal,low,medium,high}]
                                    [--dry-run]
                                    inputs [inputs ...]

Create a Markdown ChangeLog either from one compare-update log or by directly
comparing OLD_FILE and NEW_FILE.

positional arguments:
  inputs                One argument: compare_update_diff.py log. Two
                        arguments: OLD_FILE NEW_FILE.

options:
  -h, --help            show this help message and exit
  -o OUTPUT, --output OUTPUT
                        Output Markdown file (default: "CHANGELOG.md")
  --root-dir1 ROOT_DIR1
                        Override root_dir1 written in the log (log mode only)
  --root-dir2 ROOT_DIR2
                        Override root_dir2 written in the log (log mode only)
  --ai-lib AI_LIB       Path to tkai_lib_litellm.py
  --config CONFIG       AI config env file (default: "translate.env")
  --provider PROVIDER   AI provider, e.g. "openai" or "gemini"
  --model MODEL         Model name. If omitted, AI_MODEL env or gpt-4o-mini is
                        used
  --language {ja,en}    ChangeLog language (default: "ja")
  --max-direct-chars MAX_DIRECT_CHARS
                        Maximum combined old+new source characters for one
                        direct AI comparison (default: 200000)
  --source-chunk-chars SOURCE_CHUNK_CHARS
                        Approximate source characters per chunk for very large
                        files (default: 40000)
  --temperature TEMPERATURE
                        Optional AI temperature. Default: do not send
  --reasoning-effort {minimal,low,medium,high}
                        Optional reasoning_effort for compatible models
  --dry-run             Show source paths without calling AI

生成されたデータファイル

（データファイルが見つかりませんでした）

生成された画像一覧

（画像ファイルが見つかりませんでした）