core.scripts.download_large_files#

Attributes#

Functions#

parse_args()

change_path_for_pypi(→ list[pathlib.Path])

Modify or exclude files from download if running in a PyPi-installed

download_file_group(→ None)

Download the given file group.

Module Contents#

core.scripts.download_large_files.S3_ROOT = 'https://dl.fbaipublicfiles.com/opencatalystproject/data/large_files/'#
core.scripts.download_large_files.FILE_GROUPS#
core.scripts.download_large_files.parse_args()#
core.scripts.download_large_files.change_path_for_pypi(files_to_download: list[pathlib.Path], par_dir: str, install_dir: pathlib.Path, test_par_dir: pathlib.Path | None) list[pathlib.Path]#

Modify or exclude files from download if running in a PyPi-installed build.

Installation of FAIR-Chem with PyPi does not include the entire directory structure of the fairchem repo. As such, files outside of src can’t be downloaded and those in src should actually be in the site-packages directory. If a user wants these files, they must build from the git repo.

If the tests have been separately downloaded (e.g. from the git repo), then we can download if we’ve been told where those tests have been downloaded to. Note that we can’t divine that location from anything in fairchem.core because they would have to be somewhere “unexpected” since we’ve built with PyPi which shouldn’t have tests at all.

Parameters:
  • files_to_download – List of files to be downloaded

  • par_dir – the parent directory of the PyPi build, probably “site-packages”

  • install_dir – path to where fairchem.core was installed

  • test_par_dir – path to where tests have been downloaded (not necessarily the same as install_dir)

Returns:

modified list of files to be downloaded

core.scripts.download_large_files.download_file_group(file_group: str, test_par_dir: pathlib.Path | None = None) None#

Download the given file group.

Parameters:
  • file_group – Name of group of files to download

  • test_par_dir – Parent directory where fairchem tests have been downloaded

core.scripts.download_large_files.args#