... in order to avoid multiple full stops messing up syntax.
We can use `\.` as a literal full stop.
`[^ ]` means anything other than the containing character, so `[^\.]` would mean `anything other than a full stop'.
`*` in sed means this character is repeated any number of times, including 0, so files with no `.` character would still be processed fine.
`$` means `end of the line'.
Apparently sed uses `regex, not globing'.
Regular expressions (``regex'') looks for patterns and is used with find and grep. It interprets `*' as a wildcard, `?' as a single-character wildcard, and [12,1,2] as characters matching one of a set(in this case, `12 or 1 or 2 but not 21'].
If the shell is set to find file ``a*b.txt''then it will pass this first to regex, and hit items like `aab.txt' and `abb.txt'. If it finds nothing, it'll then use globbing, and interpret `a*b.txt' literally.
IFS is the field separator. This is required to denote the different files as marked by a new line, and not the spaces.
# Arguments and Input
The `rm' program takes arguments, but not `stdin' from a keyboard, and therefore programs cannot pipe results into rm.
That said, we can sometimes pipe into rm with `xargs rm' to turn the stdin into an argument. For example, if we have a list of files called `list.txt'then we could use cat as so:
> cat list.txt | xargs rm
... *However*, this wouldn't work if spaces were included, as rm would take everything literally.
