Truncate file affect the read process - c

Question

Recently, I am reading the UNIX Systems Programming, and doing some test.
I'm confused that why i truncate the file with truncate command in the terminal, then the read process read the same position return 0 which return the string "a" before. The process open the file, it will cache the file, beacase i change the file content, then read, the result is the same not changed, so why truncate file will affect the process immediately？In the UNIX Systems Programming, it says that v node include the current file size? so the size don't be read from the disk every time, it's cached in the memory.
the read process code source.
#include <unistd.h>
#include <stdio.h>
#include <unistd.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <string.h>
int main(int argc, char *argv[])
{
int fd = open("a.txt", O_RDONLY);
char buf[1024];
int n;
for(int i = 0;i < 10; i++){
if (lseek(fd, 0, 0) == -1){
return 1;
}
n = read(fd, buf, 1);
if (n < 0){
perror("error");
return 1;
}
printf("%d\n", n);
buf[n]=0;
printf("buf %s\n", buf);
sleep(2);
}
return 0;
}
a.txt only have one char:
a

Answer 1

Caching is usually intended to improve performance, without having any other detectable effects.
If the inode was cached by the process, as you describe, then modifications made by another process might not be visible by processes that had cached information. That would be bad.
In reality, disk blocks are cached, inodes are cached, various other structures might be cached, but in each case there is a single cache, in the kernel, which all processes share, so they all have a consistent view.
When one process truncates the file, the in-memory cache is updated, storing the new file size (which will eventually be written to disk, but probably not immediately). When the other process calls read again, the updated file size is read from the in-memory cache, not from disk.