Remove trailing whitespace or remove all whitespaces on both sides?

[baldassarreFe]

This line doesn't do what the comment says:

natsort/natsort/__main__.py

Lines 180 to 181 in dc1edbb

	# Remove trailing whitespace from all the entries
	entries = [e.strip() for e in args.entries]

As it is now, it will remove all whitespaces on both sides of the string. Should it only remove the trailing one (str.rstrip())? Should it remove only one character or all whitespaces?

Example:

python -m natsort 'tmp/a57/path2 ' '  tmp/a23/path1' ' tmp/a1/path1'
tmp/a1/path1
tmp/a23/path1
tmp/a57/path2

[SethMMorton]

Agreed. I wrote that over a decade ago when I did not have any idea what I was doing.

[SethMMorton]

Would you also be interested in contributing to this fix?

[baldassarreFe]

Sure, why not! This one is actually pretty easy.

But let's decide on the desired behavior.

• Strip left+right or only right?
• Strip one whitespace/null character or all?

[SethMMorton]

So, I gave this some thought. Trailing whitespace should not actually affect the ordering, so stripping it has no real effect.

Leading whitespace certainly has an effect and I think stripping it is incorrect.

At this point, I am leaning towards not doing any stripping at all.

[baldassarreFe]

Well said, not stripping at all is probably a good choice.

The only use case for stripping leading blanks would be sorting the output of another program that prints a formatted table like this:

      JobID     JobName              Submit    Elapsed    State
----------- ----------- ------------------- ---------- --------
  3082235_2   job_array 2025-08-04T07:17:51   00:25:02  RUNNING
  3090244_2   job_array 2025-08-04T07:18:04   00:25:02  RUNNING
    3093865  single_job 2025-08-04T07:22:17   00:30:21  RUNNING

But in this case users can just filter the output through tr to remove the leading spaces.

---

Another thing, if we completely remove the stripping, we need to be midful of the trailing \n that is left by readlines(). It will not affect sorting, but it will affect printing the result.

For comparison:

❯ cat tmp.txt
foo
bar baz
biz buz

In [1]: with open("tmp.txt") as f:
   ...:     x=f.read().splitlines()
   ...: repr(x)
Out[1]: "['foo', 'bar baz', 'biz buz']"

In [2]: with open("tmp.txt") as f:
   ...:     x=f.readlines()
   ...: repr(x)
Out[2]: "['foo\\n', 'bar baz\\n', 'biz buz\\n']"

[SethMMorton]

Were you planning on making a PR for this?

[baldassarreFe]

Tbh, I forgot 😅 . Let me have a look now