Problem Statement

Determine if a character string is   collapsible. And if so,   collapse the string   (by removing   immediately repeated   characters).

If a character string has   immediately repeated   character(s),   the repeated characters are to be deleted (removed),   but not the primary (1st) character(s).

An   immediately repeated   character is any character that is   immediately   followed by an identical character (or characters).   Another word choice could've been   duplicated character,   but that might have ruled out   (to some readers)   triplicated characters   ···   or more.

{This Rosetta Code task was inspired by a newly introduced   (as of around November 2019)   PL/I   BIF:   collapse.}

In the following character string:

Only the 2nd   t,   e, and   l   are repeated characters,   indicated by underscores (above),   even though they (those characters) appear elsewhere in the character string.

So, after collapsing the string, the result would be:

Another example: In the following character string:

The "collapsed" string would be:

Write a subroutine/function/procedure/routine···   to locate   repeated   characters and   collapse   (delete)   them from the character string.   The character string can be processed from either direction.

Show all output here, on this page:

Use (at least) the following five strings,   all strings are length seventy-two (characters, including blanks),   except the 1st string:

Let's start with the solution:

h
s/.*/<<<&>>>/
x
s/\(.\)\1*/\1/g
s/.*/<<<&>>>/
H
s/.*//
G