On Fri, Jan 02, 2004 at 11:31:46AM +0100, LAUN Wolfgang wrote: > Festina lente ;-) Nil illegitimo carborundum? > Yes, it fixes the assertion failure and all that may come from an > insufficient allocation. But note that the eval does not return an > error, so the format can be run. Observe: [snip] > A patch against toke.c (5.8.2) appears to fix this: Thanks, applied as 22055. > Finally I'd like to draw your attention to the fact that some NULL > character *embedded* in the eval'ed string terminates format parsing > prematurely (possibly because strchr is used in toke.c in scan_formline), > resulting in an error ("Missing right curly or square bracket"). Any > need to fix this, accepting embedded NULLs? My personal take on this (not that I ever use formats!) is that since <<EOF preserves \0's, I don't see why formats should discard them. Would it be possible to extend your 'pass through control chars' patch to handle nulls too? Dave. -- SCO - a train crash in slow motion