On Thu, 22 Jan 2015 15:06:49 +0100 Truls Becken <[email protected]> wrote:
> Did you consider using the entire string from the -d argument instead > of being UTF-8 aware? That would fulfill the requirement of allowing > any unicode character, and up it by allowing e.g. ", " or " :: ". This is a very nice idea! I was actually thinking about the issue with "\t". I already wrote an escaping function in tr.c, but the priority here was for me to build the basis for a more flexible implementation. I'll see if I can work on it this evening. The way cut(1) fortunately was designed allows me to do that trivially. And while at it, I'll implement the escaping-function to allow escape characters. Cheers FRIGN -- FRIGN <[email protected]>
