Printable View
While dealing with Unix text files, I found that a lot of people prefer file manipulation writing shell scripts rather than processing using C.
My question is that, what is more efficient, shell programming or C programming?
A C program, when run compiled, works as single process, but every shell command is a process, doesn't it mean that C code is more efficient?
Thanks for any help :-)
C programming, definitely.
Although it depends what you define as "efficient". If you want to do an operation on ten million rows of a text file, C will be much faster. However, it'll take you longer to write, so in total, a shell script might be faster.
A lot of shell commands are implemented as shell builtins, so they won't spawn another process.
Also depends on the shell what is available or builtin.
Yesterday, someone needed 380,000 addresses, RIGHT NOW.
I wrote a perl script. Consider learning perl.
Efficiency is in the eye of the beholder. Compiled C is faster, but for a one-up of 40 pages of text, who cares?. If it is run 1000 times a day, use perl, C, FORTRAN, whatever. Not script.
Actually, real efficiency is cost-based. It it more expensive to have a user wait 5 seconds vs ten seconds or more efficeint to pay a programmer to work an extra hour? Depends on your shop.
Are you running the smallest unix box avaialble with 500 users? Then pay the programmer.
Yes, I agree that efficiency depends on the particular problem.
However, another advantage of using C is that, if the applications ever need to be ported to some different platform say, Windows form Unix, then changing C source code will be much easier than re-writing shell scripts.
:rolleyes:
Not necessarily. A lot of things rely on POSIX features which Windows has a very poor implementation of.
Plus, you can just run the shell program on Windows (Cygwin can do both of these, though).
Not everyone wants to install cygwin on their windows computer.
Give the programmer a little more time and he should be able to write the app in a completly portable way.