1.1.0

• codenerix_lib.normalizers no longer downloads the NLTK stopwords corpus by itself. Importing the module never touches the network; the corpus is loaded on the first strong_normalizer() call, which raises MissingStopwordsError (a LookupError) with the install command when it is missing. Install it once per machine with python -m nltk.downloader stopwords. nltk 3.10 refuses downloads through HTTP proxies by default, so the automatic download could fail at import time. - Packaging moved from setup.py + pip-tools to pyproject.toml (hatchling) + uv.lock, aligned with django-codenerix. The distribution metadata is unchanged apart from the license, now declared as the SPDX expression Apache-2.0, and the added Python 3.14 and Typing :: Typed classifiers.