I'm glad you did this, it's interesting. I'm not sure about the binary
results. Doesn't the repeat loop exit early if the search block contains
a few short lines? I think you could get a block of text that might
contain the first instance of a leading digit, but they wouldn't be
checked because their total length would be less than kBrute.
Anyway, for lists with shorter lines the script would need to lower the
kBrute variable to a smaller number, and then the timing might be
different. 10,000 seems like a pretty large exit condition.
I could easily not be understanding what's going on though. These things
warp my brain.
On 2/1/13 6:41 PM, Alex Tweedly wrote:
I just can't resist a good benchmark / coding challenge.
...
function f3 @pD
local r, tLower, tUpper, tMid, c, t, temp
constant kBrute = 10000
put 1 into tLower
put (the number of chars in pD) + 1 into tUpper
repeat 10000 times -- more than enough :-)
if tUpper - tLower < kBrute then exit repeat
put (tUpper + tlower ) / 2 into tMid
put char 1 of line 2 of (char tMid to tUpper of pD) into c
if c is a number then
put tMid into tUpper
else
put tMid into tLower
end if
end repeat
put char tLower to tUpper of pD into temp
filter temp without "[0-9]*"
if temp is empty then
-- no lines start with numeric
return "No lines"
else
put the number of lines in (char 1 to tLower of pD) into t
add the number of lines in temp to t
return t
end if
end f3
--
Jacqueline Landman Gay | jac...@hyperactivesw.com
HyperActive Software | http://www.hyperactivesw.com
_______________________________________________
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode
