Lazy assignment of python variables

Question

I have the following if statement and I feel it could be made simpler to read by breaking the terms up into their own named variables.

if os.path.isfile(os.path.join(dirname, 'file.json')) \
or (os.path.isdir(os.path.join(dirname, 'folder'))
and os.listdir(os.path.join(dirname, 'folder')) != []):

So it would look something like this:

file_exists = os.path.isfile(os.path.join(dirname, 'file.json'))
directory_exists = os.path.isdir(os.path.join(dirname, 'folder'))
directory_not_empty = os.listdir(os.path.join(dirname, 'folder')) != [])
if file_exists or (directory_exists and directory_not_empty):

The problem with this is that in the first implementation the and can short circuit and if it does the directory_not_empty term is not evaluated, which could cause an error. When they are broken into variables they are eagerly executed at assignment. My solution to this was to use a lambda to cause the directory_not_empty term to simulate lazy assignment.

file_exists = os.path.isfile(os.path.join(dirname, 'file.json'))
directory_exists = os.path.isdir(os.path.join(dirname, 'folder'))
directory_not_empty = lambda:os.listdir(os.path.join(dirname, 'folder')) != [])
if file_exists or (directory_exists and directory_not_empty()):

I think this looks quite elegant however the pep8 standard doesn't like assignment of lambdas and using a one line function definition feels very clunky here. The call on directory_not_empty and not the others also feels clunky. Is there an elegant way of doing this?

Is comparing a list to an integer what looks clunky to me.

Stop harming Monica
– Stop harming Monica

2019-04-17 08:43:19 +00:00
Commented Apr 17, 2019 at 8:43 — Stop harming Monica
– Stop harming Monica, Commented Apr 17, 2019 at 8:43
Ow! My eyes! Use import os.path already.

candied_orange
– candied_orange

2019-04-17 14:52:11 +00:00
Commented Apr 17, 2019 at 14:52 — candied_orange
– candied_orange, Commented Apr 17, 2019 at 14:52

Robert Harvey · Accepted Answer · 2019-04-17 14:39:18Z

1

If you want to do it this way, I would make the directory_not_empty variable handle both the check for the presence of the directory and check if its non-empty. Also, if the long lengths are annoying then I would put the directory in its own variable.

my_dir = os.path.join(dirname, 'folder')
directory_not_empty = os.path.isdir(my_dir) and (os.listdir(my_dir) != [])

edited Apr 17, 2019 at 14:39

Robert Harvey

201k55 gold badges470 silver badges683 bronze badges

answered Apr 17, 2019 at 11:39

SpaceMonkey55

1262 bronze badges

Add a comment |

dericke · Accepted Answer · 2021-01-26 17:41:44Z

1

Using pathlib, if feasible, may also help readability, by itself or in combination with the accepted answer:

# setup
from pathlib import Path
dirname = Path(dirname)

# Dropped into original code
if (dirname / "file.json").is_file() \
or ((dirname / "folder").is_dir()
and (dirname / "folder").iterdir() != []):

# Dropped into the accepted answer
my_dir = dirname / "folder"
directory_not_empty = my_dir.is_dir() and (my_dir.iterdir() != [])

answered Jan 26, 2021 at 17:41

dericke

1112 bronze badges

Add a comment |

Stack Exchange Network

Lazy assignment of python variables

2 Answers 2

Hot Network Questions

Lazy assignment of python variables

2 Answers 2

Related

Hot Network Questions