--- Tim Chase <[EMAIL PROTECTED]> wrote:

> As to why it does correctly handle ">" quoting, par was built to 
> handle such cases, so it gracefully considers "leader" 
> characters.  By default, I believe par considers whitespace, a 
> ">" and the pipe character as its default set of "leader" 
> characters that it will treat in a similar fashion.

Ah.  I might have read something like this in par's somewhat limited
documentation.  Although I freely admit that I find par very confusing
indeed, even though it is very useful for tasks such as this.

> The "par" utility is a rather complex tool with piles of nobs and 
> dials for the twiddling.  In most cases, the default does what is 
> wanted.  However, I don't believe there's any "intelligent" way 
> for it to detect scripts within quoted text.

I thought as much.  What I might do is (as you say) look at trying to see
if there are any markers which define where the shell-code is and change
the regexps.  I know of two cases where this is likely to be the case, so
I'll keep you posted.

> To see what's going on, simply write an excerpt of a quoted 
> script to a file and simply run it through par, as Vim is doing 
> for each block in question.  You'll notice that par is what is 
> doing the mangling.  For just reformatting purpose, it's 
> conceivable to use vim's internal reformatting--however, it too 
> lacks the superpowers to discern code from non-code in a quotation.

I should point out that it's not just shell code that exhibits this
"problem" -- but you're right -- it's most likely par treating various
leaders characters.

> The short answer, sadly, is "I'm afraid there's no nice way to do 
> this".

No worries, I'll just have to convert it back by hand.  :)

Thank you as always, Tim.

Thomas.



        
        
                
___________________________________________________________ 
All new Yahoo! Mail "The new Interface is stunning in its simplicity and ease 
of use." - PC Magazine 
http://uk.docs.yahoo.com/nowyoucan.html

Reply via email to