Auto update Conda Environment

[JohT]

🚀 Feature

• Update conda environment if its outdated compared to the environment.yml: Previously, Jupyter notebooks failed to import libraries that had been added lately. An already existing Conda environment "codegraph" was sufficient, even it was outdated. Now, it will automatically be updated if necessary so that there are no more import errors.

• Add PREPARE_CONDA_ENVIRONMENT to skip Conda environment setup: Previously, Conda environment activation was skipped when the codegraph environment was already active. Now, PREPARE_CONDA_ENVIRONMENT="false needs to be set additionally to explicitly skip that part. This is needed in GitHub Action pipelines because conda init doesn't work as expected but is taken care of by setup-miniconda.

• Introduce script testing: The first (for now framework-free) script test is implemented in the pipeline 🎉.

⚙️ Optimization

• Improve change file detection):
  • Log output is now colored (red = error, dark grey = info)
  • Given --paths are now validated
  • File statistics are now correctly extracted for MacOS and Linux

🛠 Fix

• Defer download URL check for offline mode: Previously, it was not possible to get an artifact from the download script in offline mode, even if it had already been downloaded and ready to use in the cache. This is now resolved by deferring the check of the URL until right before the actual download, since it needs an internet connection.
• Fix wrong variable for Jupyter notebook directory: Conda environment creation still used an old variable from another file that kept working since these files are called consecutively. However, can break easily and is now resolved.