I wanted to know how to deal with a situation when I set the size of buffer in fgets to n bytes and the user enters 2n bytes? fgets will reaf the first n bytes, and the other n bytes will be left in stdin, right? Is it a good idea to flush stdin after each fgets, is there any other optiob to deal with that situation?
You know when the input was truncated by the absence of a '\n' newline at the end of the input string.
Except possibly for the last line in a text file, which might not contain any newline.
If there is more input, you can realloc the buffer (originally acquired with malloc) and repeat with the appropriate buffer pointer (and size) to append to what you already have.
No need to flush anything, just pass the right pointer and additional buffer size to fgets.
In any case, flushing input with fflush() is not well defined.
Related
this is probably quite easy to figure out, maybe i'm just looking in the wrong places, but how does one test if fgets has read an oversized input? In the code below, i'm trying to skip further processing for empty lines and oversized ones and go straight to the next line, for empty lines it works just fine.
Printing the strlen(buffer) when using line lengths < maxsize and it gives me expected values.
However when i enter lines that exceed the maxsize, it prints a value over 9000, which should still exceed the maxsize, and therefore enter the if-clause, but this doesn't happen. I've tried casting the return value of strlen into an int, didn't work.
What am i missing here? Thanks for any replies :)
char buffer[102];
while (fgets(buffer,100,stdin)!=NULL){
size_t maxsize = 102;
printf("%ld",strlen(buffer));
if(strcmp(buffer,"\n")==0||strlen(buffer)>maxsize){
continue;
}
//further processing
}
I
in the code:
char buffer[102];
while (fgets(buffer,100,stdin)!=NULL){
You don't need to give two more characters to buffer. The parameter size of fgets just can be the application of the sizeof operator, as in:
char buffer[102];
while (fgets(buffer, sizeof buffer, stdin) != NULL) {
That will give you space for lines of up to 101 characters (to leave space to the string terminator) including (or not, see below) the new line character.
But, answering your question, I understand that you want to know what happens if your input in one line is indeed bigger that the buffer size you provided, what happens then to the input, and how fgets deal with this:
Fgets() reads as many characters as it finds a \n in the input, or the buffer fills completely (this is, after including the \0 character that it must append to the string to terminate it) So, fgets() will fill as many characters in the buffer as the buffer has, minus one, reserved for the null string terminator, and the rest of the line will be read in the next fgets() (or another call to any of the functions of the stdio package).
So, basically, lines longer than one less than the buffer size are split in pieces, in which all except the last don't actually end in a new line, and the last will have the new line included, and will be shorter, all with a length of the length you specified minus one, but the last piece, in which the length is what it requires (again, always less than or equal than the length specified minus one)
I really confused with how exactly a buffer work. So I write a little snippet to verify:
#include<stdio.h>
#define BUF_SIZE 1024
char buf[BUF_SIZE];
char arr[20];
int main()
{
FILE* fs=fopen("test.txt","r");
setvbuf(fs,buf,_IOFBF,1024);
fread(arr,1,1,fs);
printf("%s",arr);
getchar();
return 0;
}
As you see, I set the file stream fs to fully buffered stream(I know most of the time it would default to fully-buffered. just making sure). And I also set its related buffer to be size 1024, which mean that the stream would not be flushed until it contain 1024 bytes of stuff(right?).
In my opinion, the routine of fread() is that, it read data from the file stream, store it at its buffer buf,and then the data in the buf would be send to the arr as soon as it is full of 1024 bytes of data(right?).
But now, I read only one character from the stream!!And also, there is are only four characters in the file test.txt. why can I find something in the arr in case that there is only one char(I can print that one character out)
The distinctions between fully-buffered, line-buffered, and unbuffered really only matter for output streams. I'm pretty sure that input streams are pretty much always act like they're fully buffered.
But even for fully-buffered input streams, there's at least one case where the buffer won't be fully full, and as you've discovered, that's where there aren't enough characters left in the input to fill the buffer. If there are only 4 characters in the file, then when the system goes to fill the buffer, it gets those 4 characters and puts them in the buffer, and then you can start taking them out, as usual.
(The same situation would arise any time the file contains a number of characters that's not an exact multiple of the buffer size. For example, if the input file contained 1028 characters, then after filling the buffer with the first 1024 characters and letting you read them, the next time it filled the buffer, it'd end up with 4 again.)
What were you expecting it to do in this case? Block waiting to read 1,020 more characters from the file (that were never going to come)?
P.S. You said "the stream would not be flushed until it contained 1024 bytes of stuff, right?" But flushing is only defined for output streams; it doesn't mean anything for input streams.
From what I understand, an input buffer works different to what you suggested: if you request one Byte to be read, the system reads 1023 more Bytes into the buffer, so on the next 1023 subsequent read calls it can return data directly from the buffer instead of having to read from the file.
I am writing a C program, which has a 5-element array to store a string. And I am using gets() to get input. When I typed in more than 5 characters and then output the string, it just gave me all the characters I typed in. I know the string is terminated by a \0 so even I exceeded my array, it will still output the whole thing.
But what I am curious is where exactly gets() stores input, either buffer or just directly goes to my array?
What if I type in a long long string, will gets() try to store characters in the memories that should not be touched? Would it gives me a segment fault?
That's why gets is an evil. It does not check array bound and often invokes undefined behavior. Never use gets, instead you can use fgets.
By the way, now gets is no longer be a part of C. It has been removed in C11 standard in favor of a new safe alternative, gets_s1 (see the wiki). So, better to forget about gets.
1. C11: K.3.5.4.1 The gets_s function
Synopsis
#define _ _STDC_WANT_LIB_EXT1_ _ 1
#include <stdio.h>
char *gets_s(char *s, rsize_t n);
gets() will store the characters in the 5-element buffer. If you type in more than 4 characters, the end of string character will be missed and the result may not work well in any string operations in your program.
excerpt from man page on Ubuntu Linux
gets() reads a line from stdin into the buffer pointed to by s until
either a terminating newline or EOF, which it replaces with a null byte
('\0'). No check for buffer overrun is performed
The string is stored in the buffer and if it is too long it is stored in contiguous memory after the buffer. This can lead to unintended writing over of data or a SEGV fault or other problems. It is a security issue as it can be used to inject code into programs.
gets() stores the characters you type directly into your array and you can safely use/modify them. But indeed, as haccks and unxnut correctly state, gets doesn't care about the size of the array you give it to store its chars in, and when you type more characters than the array has space for you might eventually get a segmentation fault or some other weird results.
Just for the sake of completeness, gets() reads from a buffered file called stdin which contains the chars you typed. More specifically, it takes the chars until it reaches a newline. That newline too is put into your array and next the '\0' terminator. You should, as haccks says, use fgets which is very much alike:
char buf[100]; // the input buffer
fgets(buf, 100, stdin); // reads until it finds a newline (your enter) but never
// more than 99 chars, using the last char for the '\0'
// you can now use and modify buf
I am having difficulty with a feature of a segment of code that is designed to illustrate the fgets() function for input. Before I proceed, I would like to make sure that my understanding of I/O and streams is correct and that I'm not completely off base:
Input and Output in C has no specific viable function for working with strings. The one function specific for working with strings is the 'gets()' function, which will accept input beyond the limits of the char array to store the input (thus making it effectively illegal for all but backward compatibility), and create buffer overflows.
This brings up the topic of streams, which to the best of my understanding is a model to explain I/O in a program. A stream is considered 'flowing water' on which the data utilized by programs is conveyed. See links: (also as a conveyor belt)
Can you explain the concept of streams?
What is a stream?
In the C language, there are 3 predefined ANSII streams for standard input and output, and 2 additional streams if using windows or DOS which are as follows:
stdin (keyboard)
stdout (screen)
stderr (screen)
stdprn (printer)
stdaux (serial port)
As I understand, to make things manageable it is okay to think of these as rivers that exist in your operating system, and a program uses I/O functions to put data in them, take data out of them, or change the direction of where the streams are flowing (such as reading or writing a file would require). Never think of the 'beginning' or 'end' of the streams: this is handled by the operating system. What you need to be concerned with is where the water takes your data, and that is mediated by use of specific functions (such as printf(), puts(), gets(), fgets(), etc.).
This is where my questions start to take form. Now I am interested in getting a grasp on the fgets() function and how it ties into streams. fgets() uses the 'stdin' stream (naturally) and has the built in fail safe (see below) that will not allow user input to exceed the array used to store the input. Here is the outline of the fgets() function, rather its prototype (which I don't see why one would ever need to declare it?):
char *fgets(char *str , int n , FILE *fp);
Note the three parameters that the fgets function takes:
p1 is the address of where the input is stored (a pointer, which will likely just be the name of the array you use, e.g., 'buffer')
p2 is the maximum length of characters to be input (I think this is where my question is!)
p3 specifies the input stream, which in this code is 'stdin' (when would it ever be different?)
Now, the code I have below will allow you to type characters until your heart is content. When you hit return, the input is printed on the screen in rows of the length of the second parameter minus 1 (MAXLEN -1). When you enter a return with no other text, the program terminates.
#include <stdio.h>
#define MAXLEN 10
int main(void)
{
char buffer[MAXLEN];
puts("Enter text a line at a time: enter a blank line to exit");
while(1)
{
fgets(buffer, MAXLEN, stdin); //Read comments below. Note 'buffer' is indeed a pointer: just to array's first element.
if(buffer[0] == '\n')
{
break;
}
puts(buffer);
}
return 0;
}
Now, here are my questions:
1) Does this program allow me to input UNLIMITED characters? I fail to see the mechanism that makes fgets() safer than gets(), because my array that I am storing input in is of a limited size (256 in this case). The only thing that I see happening is my long strings of input being parsed into MAXLEN - 1 slices? What am I not seeing with fgets() that stops buffer overflow that gets() does not? I do not see in the parameters of fgets() where that fail-safe exists.
2) Why does the program print out input in rows of MAXLEN-1 instead of MAXLEN?
3) What is the significance of the second parameter of the fgets() function? When I run the program, I am able to type as many characters as I want. What is MAXLEN doing to guard against buffer overflow? From what I can guess, when the user inputs a big long string, once the user hits return, the MAXLEN chops up the string in to MAXLEN sized bites/bytes (both actually work here lol) and sends them to the array. I'm sure I'm missing something important here.
That was a mouthful, but my lack of grasp on this very important subject is making my code weak.
Question 1
You can actually type as much character as your command line tool will allow you per input. However, you call to fgets() will handle only MAXLEN in your example because you tell him to do so.
Moreover, there is no safe check inside fgets(). The second parameter you gave to fgets is the "safety" argument. Try to give to change your call to fgets to fgets(buffer, MAXLEN + 10, stdin); and then type more than MAXLEN characters. Your program will crash because you are accessing unallocated memory.
Question 2
When you make a call to fgets(), it will read MAXLEN - 1 characters because the last one is reserved to the character code \0 which usually means end of string
The second parameter of fgets() is not the number of character you want to store but the maximum capacity of your buffer. And you always have to think about string termination character \0
Question 3
If you undestood the 2 answer before, you will be able to answer to this one by yourself. Try to play with this value. And use a different value than the one used for you buffer size.
Also, you said
p3 specifies the input stream, which in this code is 'stdin' (when would it ever be different?)
You can use fgets to read files stored on your computer. Here is an example :
char buffer[20];
FILE *stream = fopen("myfile.txt", "r"); //Open the file "myfile.txt" in readonly mode
fgets(buffer, 20, stream); //Read the 19 first characters of the file "myfile.txt"
puts(buffer);
When you call fgets(), it lets you type in as much as you want into stdin, so everything stays in stdin. It seems fgets() takes the first 9 characters, attaches a null character, and assigns it to buffer. Then puts() displays buffer then creates a newline.
The key is it's in a while loop -- the code loops again then takes what was remaining in stdin and feeds it into fgets(), which takes the next 9 characters and repeats. Stdin just still had stuff "in queue".
Input and Output in C has no specific viable function for working with strings.
There are several functions for outputting strings, such as printf and puts.
Strings can be input with fgets or scanf; however there is no standard function that both inputs and allocates memory. You need to pre-allocate some memory, and then read some characters into that memory.
Your analogy of a stream as a river is not great. Rivers flow whether or not you are taking items out of them, but streams don't. A better analogy might be a line of people at the gates to a stadium.
C also has the concept of a "line", lines are marked by having a '\n' character at the end. In my analogy let's say the newline character is represented by a short person.
When you do fgets(buf, 20, stdin) it is like "Let the next 19 people in, but if you encounter a short person during this, let him through but not anybody else". Then the fgets function creates a string out of these 0 to 19 characters, by putting the end-of-string marker on the end; and that string is placed in buf.
Note that the second argument to fgets is the buffer size , not the number of characters to read.
When you type in characters, that is like more people joining the queue.
If there were fewer than 19 people and no short people, then fgets waits for more people to arrive. In standard C there's no way to check if people are waiting without blocking to wait for them if they aren't.
By default, C streams are line buffered. In my analogy, this is like there is a "pre-checking" gate earlier on than the main gate, where all people that arrive go into a holding pen until a short person arrives; and then everyone from the holding pen plus that short person get sent onto the main gate. This can be turned off using setvbuf.
Never think of the 'beginning' or 'end' of the streams: this is handled by the operating system.
This is something you do have to worry about. stdin etc. are already begun before you enter main(), but other streams (e.g. if you want to read from a file on your hard drive), you have to begin them.
Streams may end. When a stream is ended, fgets will return NULL. Your program must handle this. In my analogy, the gate is closed.
We know that stdin is, by default, a buffered input; the proof of that is in usage of any of the mechanisms that "leave data" on stdin, such as scanf():
int main()
{
char c[10] = {'\0'};
scanf("%9s", c);
printf("%s, and left is: %d\n", c, getchar());
return 0;
}
./a.out
hello
hello, and left is 10
10 being newline of course...
I've always been curious, is there any way to "peek" at the stdin buffer without removing whatever may reside there?
EDIT
A better example might be:
scanf("%9[^.]", c);
With an input of "at.ct", now I have "data" (ct\n) left on stdin, not just a newline.
Portably, you can get the next character in the input stream with getchar() and then push it back with ungetc(), which results in a state as if the character wasn't removed from the stream.
The ungetc function pushes the character specified by c (converted to an unsigned char) back onto the input stream pointed to by stream. Pushed-back characters will be returned by subsequent reads on that stream in the reverse order of their pushing.
Only one character of pushback is guaranteed by the standard, but usually, you can push back more.
As mentioned in the other answers resp. the comments there, in practice, you can almost certainly peek at the buffer if you provide your own buffer with setvbuf, although that is not without problems:
If buf is not a null pointer, the array it points to may be used instead of a buffer allocated by the setvbuf function
that leaves the possibility that the provided buffer may not be used at all.
The contents of the array at any time are indeterminate.
that means you have no guarantee that the contents of the buffer reflects the actual input (and it makes using the buffer undefined behaviour if it has automatic storage duration, if we're picky).
However, in practice the principal problem would be finding out where in the buffer the not-yet-consumed part of the buffered input begins and where it ends.
If you want to look at the stdin buffer without changing it, you could tell it to use a another buffer with setbuf, using an array you can access:
char buffer[BUFSIZ];
if (setbuf(stdin, buffer) != 0)
// error
getchar();
printf("%15s\n", buffer);
This let you see something more than ungetc, but I don't think you can go further in a portable way.
Actually this is legal but is not correct for the standard, quoting from it about the setvbuf (setbuf has the same behavior):
The contents of the array at any time are indeterminate.
So this is not what you need if you're looking for complete portability and standard-compliance, but I can't imagine why the buffer should not contain what is expected. However, it seems to work on my computer.
Beware that you have to provide an array of at least BUFSIZ characters to setbuf, and you must not do any I/O operation on the stream before it. If you need more flexibility, take a look at setvbuf.
You could set your own buffer with setvbuf on stdin, and peek there whenever you want.