On Thu, 27 Jun 1996 TBaehr@xxxxxxxx wrote: > OK, I promise to stop this madness. I have a v.7, but it's a trivial change > from v.6. > No - don't stop - I want to see vers 99 - keep on tweaking ... I've been using a III version of your prog for a few days. I'm still tweaking it - I've found that I'm prone to hit F3 to end the search - so now F3 is the same as enter. I've also put the whole search/select into a subroutine that returns an error code. Instead of separate search down/back routines, I pass "b" and "" to the subroutine so it will know which way to search and which way to back up to define the search arg. I've put the search arg on the cmd line so I'm sure of what I've typed. - Also implemented David Auerbach's design such that a failed search starts a new search with the character that caused the error. I need to work on searching for ascii-2, control chars (to search for modes) and functions (to search for XPL function codes) - Last: I've found that in III+ I can just concatenate the backspace char to strip off the last char of the search arg. Still get the proper string for default on next search. But in Vers 4 - doing this causes "BD" to appear in the default search string (odd eh?) In vers III the siz of the string shows that the BD function is in there, but the resulting string is what you'd expect (last char lopped off) - So - (I think this is why Holmgren was pissed that he couldn't find your code within the Enhanced set) - its not just the finished product that draws interest. Multiple versions of the same basic idea are a tradition around here. I'll be interested to see what Robert does to this routine when the leaves begin to fall. Cheers. J SollaAttachment: SRCH6.ZIP
Description: Binary data