Skip to content

Commit

Permalink
Remove functional as a runtime module (pytorch#337)
Browse files Browse the repository at this point in the history
Summary:
Pull Request resolved: pytorch#337

The motivation to use runtime module is to override `__get_attr__` (which allows lazily loading operation at runtime, e.g. by dlopen a .so file),
however, this doesn't work well with doc generation (plus it makes code more obscure to read).

Since Python 3.7, `__get_attr__` is natively overridable: https://peps.python.org/pep-0562/
So we can get rid of using runtime module.

Reviewed By: damianr99

Differential Revision: D36530070

fbshipit-source-id: 2318f517daddfc5bf0e215dc02660f36d48eb5de
  • Loading branch information
wenleix authored and facebook-github-bot committed May 20, 2022
1 parent d394630 commit a820577
Show file tree
Hide file tree
Showing 8 changed files with 227 additions and 220 deletions.
1 change: 0 additions & 1 deletion torcharrow/__init__.py
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,6 @@
# LICENSE file in the root directory of this source tree.

from . import pytorch, velox_rt # noqa # noqa
from ._functional import functional
from .icolumn import Column, column, concat, if_else # noqa
from .idataframe import DataFrame, dataframe, me # noqa
from .ilist_column import ListColumn # noqa
Expand Down
214 changes: 0 additions & 214 deletions torcharrow/_functional.py

This file was deleted.

Loading

0 comments on commit a820577

Please sign in to comment.