windows command line giving error on calculation - variables

Question

I found this nice little tidbit of code here: https://stackoverflow.com/a/5262637/2128987
#echo off
set starttime=%TIME%
set startcsec=%STARTTIME:~9,2%
set startsecs=%STARTTIME:~6,2%
set startmins=%STARTTIME:~3,2%
set starthour=%STARTTIME:~0,2%
set /a starttime=(%starthour%*60*60*100)+(%startmins%*60*100)+(%startsecs%*100)+(%startcsec%)
:TimeThis
robocopy /e /NFL /NDL /NJH /NJS /nc /ns /np folder%rndfolder% %drvltr%:\f%dirnew%\
set endtime=%time%
set endcsec=%endTIME:~9,2%
set endsecs=%endTIME:~6,2%
set endmins=%endTIME:~3,2%
set endhour=%endTIME:~0,2%
if %endhour% LSS %starthour% set /a endhour+=24
set /a endtime=(%endhour%*60*60*100)+(%endmins%*60*100)+(%endsecs%*100)+(%endcsec%)
set /a timetaken= ( %endtime% - %starttime% )
set /a timetakens= %timetaken% / 100
set timetaken=%timetakens%.%timetaken:~-2%
echo.
echo Took: %timetaken% sec.
As a standalone program it works great. I am using it with a robocopy command basically to determine how long it takes to write a file.
I add one extra variable in it because I want to keep the raw seconds for calculation purposes. So I add the extra line set timeraw=%timetaken%:
set /a timetaken= ( %endtime% - %starttime% )
***set timeraw=%timetaken%***
set /a timetakens= %timetaken% / 100
set timetaken=%timetakens%.%timetaken:~-2%
My batch file also uses setlocal enabledelayedexpansion
Well sometimes it does not properly calculate the "starttime" or "endtime". It's keeps it as the raw time in 08:30:22.35 type format and results in the error:
Invalid number. Numeric constants are either decimal (17),hexadecima (0x11), or octal (021)
Well obviously because it contains non-numeric characters like the : character.
My batch file goes in a continuous loop forever as I am using it to read, write, delete files and folders for a specific torture test condition.
Any idea why it would intermittently not calculate the starttime or endtime variables?
edit:
I made some changes to my overall script. I no longer need enabledelayedexpansion, cleaned up some if then statements, and simplified code a little. But I still occasionally get it where the starttime or endtime variables remain as the raw time format of HH:MM:SS.CS and causes error in calculation.

Answer 1

Old question, but there are probably blocks of parentheses and when you change a variable within parentheses then you need to use delayed expansion.
Run this and examine the differences.
#echo off
set a=nothing
if z==z (
set a=b
echo %a%
)
pause
setlocal enabledelayedexpansion
set a=nothing
if z==z (
set a=b
echo !a!
)
pause

Answer 2

Gee - a question nearly a year old, with no answer.
I'll assume that the problem has now been solved, so as a matter of record, I'd conclude that the
"sometimes it does not properly calculate" is because the hour/minute/second/hundredths will contain "08" or "09" which are not octal numbers.
The solution is
set /a startcsec=1%STARTTIME:~9,2% - 100
and repeat with each of the other 3 start time-segments; then repeat again with the end parts.
In addition, it could be that the hour is being presented with 0s suppressed. In this case, I'd suggest
set starttime=0%TIME: =%
set starttime=%startTIME:~-11%
set /a startcsec=1%STARTTIME:~9,2% - 100
where the first line prefixes the time with a '0', and replaces Space with [nothing]
the second selects just the last 11 characters of the result
and the last is the familiar form, using the resultant hh:mm:ss.cc format.
(obviously, the remainder of the substring-and-calculate method needs also to be implemented)