dupe: ignores "-" ?

[NotNull]

Test setup:

Code: Select all

T:\TEST>dir /b
#abc.test
$abc.test
%abc.test
'abc.test
(abc.test
)abc.test
+abc.test
-abc-.test
-abc.test
@abc.test
abc-.test
[abc.test
]abc.test
_abc.test
`abc.test
{abc.test
}abc.test
~abc.test



T:\TEST>es.exe  t:\test\ dupe:name
T:\TEST\'abc.test
T:\TEST\-abc-.test
T:\TEST\-abc.test
T:\TEST\abc-.test

Same goes for dupe:path (using a different setup)
Same goes for unique:

[void]

Everything finds duplicates using its built-in natural sort.

The Everything natural sort:

• ignores case
• ignores
  -
  and
  '
• normalizes text
• removes diacritics
• treats digit sequences as numbers

If you want to find duplicates using Unicode code point values instead of the natural sort, prefix your dupe: search with the case: modifier.

For example:

case:dupe:

[therube]

See also, Understanding Sort Sequence 2 - Code Point.

[NotNull]

@void, therube: Thanks for the explanations. Will keep that in mind.

One question remains:

The Everything natural sort:
[...]
treats digit sequences as numbers

What is the underlying mechanism when numbers are involved?
For example for 00001.test and 01.test.

[void]

Basically, it's regex:\d+

00001 is the same as 01
Both are treated as the number 1
However, if the numbers are the same, Everything will then check the number of leading zeroes.
More leading zeroes are shown first.

000 < 00
00001 < 01

So while 00001 and 01 are the same number, they still compare differently because of the extra leading zeroes.

[NotNull]

All clear. No more questions, your honor