From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <db3d1508ae67897d7430cf40918dffef@india.com>
To: 9fans@india.com
Date: Fri, 14 Nov 2014 13:00:33 +0000
From: trebol@india.com
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit
Subject: [9fans] e0|e01 on regexp(3)
Topicbox-Message-UUID: 28109ece-ead9-11e9-9d60-3106f5b1d025

Using a little script I wrote to rename files, I encounter a funny
behavior with '|'.  If the regular expressions share a part, doesn't
matter the order of those,  The first match from the start of the input
is processed.  Now that I'm writing this, it's seems obviously the
right behavior (and I think it is), for example:


	% echo '123' | sed s'/12|12/o/g'
	o3
	% echo '123' | sed s'/23|12/o/g'
	o3

But in large concatenations with strange characters like in:
	$ echo '__-' | sed s'/_-|_-_|-_|__+ ... and a lot of more cra* ... /__/g'

You could expect the output be '__', but it's '__-'

So... if you are a retard like me, don't try to be such smart and use
various sed commands.

trebol.