This question already addresses how to remove duplicate lines, but enforces that the list is sorted first.
I would like to perform the remove contiguous duplicate lines step (i.e. uniq) without first sorting them.
Example before:
Foo
Foo
Bar
Bar
Example after:
Foo
Bar
Just found the solution here. The following regex works correctly:
g/^\(.*\)$\n\1$/d
:%!uniq
if you're on a unix system, or a system that has the uniq program
If you want to remove non-contiguous duplicates you could use
:g/^\(.*\)\ze\n\%(.*\n\)*\1$/d
(which will remove all but the last copy of a line)
which would change
Foo
Bar
Foo
Bar
Foo
Baz
Foo
Quux
to
Bar
Baz
Foo
Quux
If you want to remove all but the first copy, try
:g/^/m0
:g/^\(.*\)\ze\n\%(.*\n\)*\1$/d
:g/^/m0
which would change
Foo
Bar
Foo
Bar
Foo
Baz
Foo
Quux
to
Foo
Bar
Baz
Quux
If you just want to remove contiguous duplicate lines, just use uniq without sorting anything.
:%!uniq
:%s/^\(.*\)\(\n\1\)\+$/\1/ge
this is my answer for you
I know this is old, but it's worth mentioning the following also works if you don't mind sorting as well (I know the OP wanted to avoid it):
:sort u
来源：https://stackoverflow.com/questions/2738826/removing-contiguous-duplicate-lines-in-vi-without-sorting