I was reading the C programming book by Kernighan and and came across putc and fputc. I didn't quite understand the difference between the two or when would one be used over the other. I found some posts on StackOverflow which dealt with this topic but it's still unclear.
As mentioned here (putc needs stdout, vs puts) :
According to Kernighan's book
putcis equivalent tofputcbutputccould be implemented as a macro andputcmay evaluate its stream argument more than once.The difference between
putcandfputcis that by usingputc, you risk running the macro version which is inherently unsafe because it may have to evaluate its stream argument more than once. This causes complications that most people aren't aware of and thus do not watch out for, sofputcis better to use.fputc's macro does not have this problem.
Questions:
putc can be implemented as a macro but what is the problem doing same with fputc?
The second statement mentions of some complications and safety issues. What are those?
putc evaluates its argument more than once. So what are the advantages or disadvantages it poses compared to evaluating the argument.
fputc() writes the character c, cast to an unsigned char, to stream. fputs() writes the string s to stream, without its trailing '\0'.
The putc() function converts c to unsigned char and then writes c to the output stream at the current position. The putchar() is equivalent to putc( c , stdout) . The putc() function can be defined as a macro so the argument can be evaluated multiple times.
The syntax of the fputc() function is as follows: Syntax: int fputc(int ch, FILE *fp); The fputc() function is used to write a single character specified by the first argument to a text file pointed by the fp pointer. After writing a character to the text file, it increments the internal position pointer.
The issue with a macro implementation is that if any of the arguments have side effects, those side effects may be evaluated more than once, possibly leading to undefined behavior. Consider this toy example:
#define SQUARE(x) ((x) * (x))
Most of the time, this will function as expected, but if you pass in an expression such as f(), then the side effect of calling the function f() will happen twice, not once, since the preprocessor is just a text transformer which is ignorant of C:
int f()
{
printf("f() was called\n");
return 42;
}
...
int x = SQUARE(f()); // This calls f() twice! It gets expanded to this:
// int x = (f() * f());
To put this in perspective, the putc function, if implemented as a macro, may evaluate its stream argument more than once. So, if that stream comes from a function:
FILE *get_file()
{
// Potential side effects could happen here
return some_file;
}
...
putc('A', get_file());
Then that could result in the function get_file() getting called more than once, with potentially unwanted side effects.
The solution, of course, is to just call a regular function like fputc() instead of putc(). Since it's not a macro, it doesn't have any potential problems with evaluating its arguments more than once. Macros can be dangerous sometimes, so use them with care.
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With