[BusyBox] bug in sed for busybox 0.52

Erik Andersen andersen at lineo.com
Tue Jul 31 10:12:34 UTC 2001


On Tue Jul 31, 2001 at 10:05:28AM -0600, Matt Kraai wrote:
> On Tue, Jul 31, 2001 at 10:01:41AM -0600, Erik Andersen wrote:
> > On Fri Jul 13, 2001 at 03:20:43PM -0600, Jim Gleason wrote:
> > > However, my patch does not fix yet another bug in the same code (bigger and
> > > badder in my opinion).  If the substitution commands are to affect the same
> > > line (first one and then another), the code will fail miserably.  First,
> > > you will get the first substitution results output, and then the second
> > > which relied on the second will fail.  Next, if the original line would
> > > have been altered by yet another substitution, we will get another altered
> > > copy of the line output.  The code needs modified to pass around a modified
> > > string and then output it if it has not been totaly removed/deleted.  I
> > > don't have time to make such a modification now, but maybe sometime next
> > > week.
> > 
> > Jim, do you have a testcase yo can provide that demonstrates this 
> > bigger and badder bug, so I can look into it?
> 
> There is such a testcase in the testsuite I posted recently.  It
> contains the following:
> 
> # XFAIL
> test "$(echo foo | busybox sed -e s/foo/bar/ -e s/bar/baz/)" = baz

Ahh, ok.  Thats what I get for still being so far behind...  I'll
grab your testsuite and take a look, thanks,

 -Erik

--
Erik B. Andersen   email:  andersen at lineo.com
--This message was written using 73% post-consumer electrons--





More information about the busybox mailing list