Negative difference of sizes in archive gives ValueError: math domain error

[jetchirag]

Description

Steps to reproduce:

1. I created a source with a single file and started a backup
2. Added a file and created another archive
3. Remove the previous file and started last backup

When I see the diff of first two backups (in which I added a file), it works fine but when I try to view the diff for last two archives, it gives this error:

Traceback (most recent call last):
  File "/Users/xxx/Projects/vorta/vorta/src/vorta/views/diff_result.py", line 776, in data
    return pretty_bytes(item.data.changed_size)
  File "/Users/xxx/Projects/vorta/vorta/src/vorta/utils.py", line 290, in pretty_bytes
    n = find_best_unit_for_size(size, metric=metric, precision=precision)
  File "/Users/xxx/Projects/vorta/vorta/src/vorta/utils.py", line 269, in find_best_unit_for_size
    n = math.floor(math.log(size * 10**precision, power))
ValueError: math domain error

I traced the error down to

vorta/src/vorta/views/diff_result.py

Lines 775 to 779 in 6bc5321

	elif column == 2:
	return pretty_bytes(item.data.changed_size)
	else:
	# size
	return pretty_bytes(item.data.size)

value of item.data.changed_size and item.data.size is in negative because log can't take negative value

n = math.floor(math.log(size * 10**precision, power))

Suggested fix:

n = math.floor(math.log(abs(size) * 10**precision, power))

It would still calculate the best unit and display +/- as per size. I'm not sure if I'm missing something. If this is a bug, please assign this to me so I can open a PR.

Reproduction

• I tried to reproduce the issue.
• I was able to reproduce the issue.

OS

macOS 12.2.1

Version of Vorta

0.8.10 / master branch

What did you install Vorta with?

Pip

Version of Borg

No response

Logs

Traceback (most recent call last):
  File "/Users/xxx/Projects/vorta/vorta/src/vorta/views/diff_result.py", line 776, in data
    return pretty_bytes(item.data.changed_size)
  File "/Users/xxx/Projects/vorta/vorta/src/vorta/utils.py", line 290, in pretty_bytes
    n = find_best_unit_for_size(size, metric=metric, precision=precision)
  File "/Users/xxx/Projects/vorta/vorta/src/vorta/utils.py", line 269, in find_best_unit_for_size
    n = math.floor(math.log(size * 10**precision, power))
ValueError: math domain error

[jetchirag]

@m3nu @real-yfprojects Can you assign it to me if you have verified the bug?

[real-yfprojects]

I noticed that one too but somehow forgot to fix it. It was introduced with #1598.

[jetchirag]

I've opened a PR with my preferred fix.

[real-yfprojects]

Thanks. Btw we get a notification for the PR too.