Re: scanning user space

["Richard B Baird" <rbaird@xxxxxxxxxxxxxxxxxxxxx>]

Pete,

Yes, my parsing could probably use some work, but it's reletively clean
now.

the last csv I posted was 77mb, and it took a couple hours to do the whole
thing - reading, data parsing, date conversion, data validation and update
another PF.

on the other hand, I was using fgets to read one record at a time (average
size 30-40 bytes), so based on your benches, i'd say i could save a bunch
of time by at least increasing my read buffer size.

and I'll probably end up writing the parsing routine to process this
particular file and screw flexibility.

as a learning experience, this all has been invaluable.

thanks again for everyone's help,

rick

-------original message---------
I didn't do anything except read the data - no parsing. I was interested in
the different timings of the various methods of reading the contents of the
file. I wasn't proposing to do Rick's job for him. You're right, of course,
that the crux of the matter is the parsing. I suspect that is the area
where
the real time savings are to be made.

Pete