Several meg files is nothing nowdays to read into memory at once
Ok so maybe you can, however, if you could do the same with just reading one line into the buffer surely this would be better?
Recall the OP said:-
Are you saying he should read all those lines into memory?There is 939197 lines in this file.
I'm just saying if you could make the code more efficient then surely that would be the optimal choice? Right?
Yes -- that's only about 70 meg (or less, assuming each line is 80 or fewer characters) file which can easily fit in most computers today.Originally Posted by treenef
The only reason I mentioned reading the whole file at once was because of the requirement to determine the offset from the beginning of the file to the start of the search string found. Reading one line at a time in text mode on MS-Windows operating system can lead to erroneous file offsets. It would be easier, and probably faster too, to do that in memory.
Yes the OP was a bit vague with his description of this. Personally, if the program said this word was found in line whatever, that would be satisfactory for most people.The only reason I mentioned reading the whole file at once was because of the requirement to determine the offset from the beginning of the file to the start of the search string found.
Why he would want to determine the offset from the beginning of the file is a little confusing. Moreover, by this does he mean how many words into the file it is, or how many chars - or what?
Anyway, I guess that's for the OP to decide.
Good Lord, I'm soooo happy that I may program in C++ like a real man, where boolean checks are performed on void pointers. Not like those Java pussies who'd probably call operator bool() to find out if something is true or falseTo convert an istream into a boolean, the compiler calls a member function called istream:: operator void*().
hth
-nv
She was so Blonde, she spent 20 minutes looking at the orange juice can because it said "Concentrate."
When in doubt, read the FAQ.
Then ask a smart question.
