GREP remove white space at start and end of paragraph

Question

Hi,
Question for the GREP experts: I'd like to construct a GREP expression
that will remove all white space at the start AND end of a paragraph in
one go.
What I've got so far works, but you have to run it twice to work:
I'll try and post it at the bottom of this message, but since I'm
posting via email it probably will come out scrambled, so I'll describe
what I've got so far:
Two expressions in parenthesis separated by the OR character (pipe). The
first expression searches for one or more space, tab or white-space at
the beginning of a paragraph.
The second expression searches for the same at the end of the paragraph.
Each expression is in parenthesis, and the two groups are separated, as
I say, by an OR sign.
However, this has to be run twice for it to work. For it removes all the
spaces at the start of paragraphs, then when run again all the spaces at
the end.
Any ideas how to get it to work in one go?
Thanks,
Ariel
The expression is:
(^[ \ts]+)|([ \ts)+$)

Answer 1

Arïel, your GREP appears a bit mashed up in the forum -- was it something like this?
(^[\t\s]+)|([\t\s]+$)
If so:
1. "\s" is "any whitespace", so you would not need to include "\t" as well. But I would not recommend this, as a HUGE disadvantage of "\s" is it also includes hard and soft returns: double returns are seen as "any whitespace" (the first return) and then "end of line" (the second return), and the first is removed. ("Disadvantage" is subject to what you want to accomplish, by the way. I usually don't want this so I list each possible whitespace character that may be removed.)
2. The parentheses around each 'section' of this GREP are not necessary. The OR operator is 'highest level' -- in this case, since you have an entire left part OR an entire right part, you can omit all parentheses.
3. This works for me ...
^[\t ]+|[\t ]+$
... and despite Peter's not-too-shabby suggestion, I would still recommend using mine instead. It's easier on ID (since it only has to remove text, not "remove-and-then-insert-again" (which is exactly what using the $.. operators do), and, more importantly, there are limits to what $1 can 'pick up'. I can't recall the special circumstances, but I have had text attributes shift, and footnotes and anchored objects disappear/being mucked up/otherwise mangled.