Pylint’s sole argument seems to be performance, which to me seems a clear case of premature optimization. The issue that introduced the checks, pylint-dev/pylint#4365, contains no reasoning at all, just the assertion that the literals would be “would be clearer”.
I like list() and dict() and set() and tuple() better myself for empty collections. Let’s enforce that instead.
Split from #1860. Pylint:
https://github.com/hpc/charliecloud/blob/d67937332094a7d8de8e9bd2e112d30d117c50c0/lib/build.py#L334
Pylint’s sole argument seems to be performance, which to me seems a clear case of premature optimization. The issue that introduced the checks, pylint-dev/pylint#4365, contains no reasoning at all, just the assertion that the literals would be “would be clearer”.
I like
list()anddict()andset()andtuple()better myself for empty collections. Let’s enforce that instead.