frankb@ruuinf.cs.ruu.nl (Frank Breure) (03/19/90)
Hello,
while I was trying to upgrade to 1.5.5 I encountered a problem with
"unshar".
The 1.5.5 postings had a shell commands like:
sed 's/^X//' > 'IBM_00.uue' << '+ END-OF-FILE ''IBM_00.uue'
The old unshar was not able to recognize the end marker
'+ END-OF-FILE IBM_00.uue', because this end marker is specified as a
concatenation of two strings.
The following cdiff will make unshar recongizing this kind of shar-files:
---------------------- cut here ------------------------
*** unshar.c~ Fri Mar 16 21:30:29 1990
--- unshar.c Fri Mar 16 21:52:31 1990
***************
*** 97,108 ****
switch (*buf) { /* Now check first char */
case '\'':
buf++;
! while (*buf != '\'') *temp++ = *buf++;
*temp = 0;
return(out);
case '\"':
buf++;
! while (*buf != '\"') *temp++ = *buf++;
*temp = 0;
return(out);
case 0:
--- 97,112 ----
switch (*buf) { /* Now check first char */
case '\'':
buf++;
! /* this assumes 'aa''bb' being the same as 'aabb' */
! while (*buf != '\'' || (*buf ++ == '\'' && *buf ++ == '\''))
! *temp++ = *buf++;
*temp = 0;
return(out);
case '\"':
buf++;
! /* this assumes "aa""bb" being the same as "aabb" */
! while (*buf != '\"' || (*buf ++ == '\"' && *buf ++ == '\"'))
! *temp++ = *buf++;
*temp = 0;
return(out);
case 0:
------------------- cut here ---------------------
--
"My opinions may be subject of change without notice."
Frank Breure (Student at Haagse Hogeschool, The Hague)
workplace at: Dept. of Computer Science, Utrecht University
UUCP: ...!uunet!mcsun!hp4nl!ruuinf!frankb Email: frankb@cs.ruu.nl