📘 What’s New in v1.4.0
- Introduced a
--version flag in the CLI to display the installed package version. (source)
- Added a
__version__ attribute in __init__.py. (source)
- Improved batch CLI functionality: you can now use
--input-dir to read entire directories (e.g., .txt and .md files), not just single files. (source)
- Added flexible CLI output: you can specify an
--output-dir and have each chunk saved as a separate file (e.g., filename_chunk_1.txt). (source)
- Alias
--input-file for --file was added, for consistency when specifying inputs. (source)
- Improved CLI output formatting: added newlines between chunks for readability. (source)
- Better CLI error messages: clearer guidance when no input arguments are provided. (source)
- Introduced a deprecation warning: using
--batch together with --file (or --input-file) is discouraged, encouraging --input-dir usage. (source)
- Optimized startup time by lazily importing
mpire in core.py only when batch processing is used. (source)
- Branding change: project renamed from
chunklet to chunklet-py to improve visibility and avoid confusion. (source)
📈 Why This Release Matters
- The rename to
chunklet-py distinguishes the library in the Python ecosystem, especially on PyPI.
- CLI improvements (
--input-dir / --output-dir) make large-scale chunking practical.
- Version reporting (
--version + __version__) helps with debugging and automation.
- Usability and performance have been enhanced without breaking major functionality.
🔗 Additional Information
⚠️ Users should note the rebranding from chunklet → chunklet-py to avoid import or installation issues.
