Question

Question:

Many files on our computers, such as executables and many music and video files, are binary files (in contrast to text files). The bytes in these files must be interpreted in ways that depend on the file format. In this exercise, we write a program data-extract to extract integers from a file and save them to an output file. The format of the binary files in this exercise is very simple. The file stores n integers (of type int). Each integer consists of 4 bytes (in the native format on the Mimir platform). We number the integers in a file from 0 to n-1. So the first 4 bytes (bytes 0 to 3) are integer 0, the next 4 bytes (bytes 4 to 7) are integer 1, and so on. The size of the file is always a multiple of 4. A program data-gen is provided to generate binary files we can work on. It takes three arguments from the command line:seed, n, and filename. The program writes n pseudo-random numbers determined by seed to file filename. For example, the following command writes 100 integers to file a.dat. The 100 integers are random numbers determined by seed 3100. Please study the source code data-gen.c to learn how integers are written to the output file. $ ./data-gen 3100 100 a.dat The task of this exercise is to complete the code in data-extract.c. The program data-extract takes at least three command line arguments (as shown below), and extract integers from file input-file and saves the extracted integers to file output-file. $./data-extract input-file output-file range [range ..] The integers to be extracted from input-file are specified by ranges. A range can be two numbers separated by '-', like start-end, or a single number. start-end specifies integers starting from integer start to integer end. If a range is a single number pos, it specifies integer pos, which is the same as pos-pos. For example, the following command extract 10 integers from a.dat to t1.dat. The integers written to t1.dat are integers at indices 0, 1, 15, 16, 17, 18, 19, 1, 2, and 3 from a.dat. Note that an integer in a.dat can be extracted multiple times. $ ./data-extract a.dat t1.dat 0 1 15-19 1 2 3 If there are 100 integers in a.dat, the following command moves the integer at the beginning to the end. $ ./data-extract a.dat t2.dat 1-99 0 In the starter code, main() parses the command line arguments. The remaining tasks are as follows. 1. Add necessary statements in main() to open and close files properly (before and after the for loop). If any file operation fails, call my_error() to report error and exit. See comments in the starter code. 2. Complete the copy_integers() function, which has the following prototype. int copy_integers(FILE *outfp, FILE *infp, int start, int end); outfp is the outpot stream and infp is the input stream. The function copies integers start to end from infp to outfp. The function returns -1 on errors and 0 on success. Do not call my_error() in this function. Read the comments in the starter code for more implementation details. Some output files (*.dat) are provided. They are generated with the following commands. $ ./data-gen 3100 100 a.dat $ ./data-extract a.dat t1.dat 0 1 15-19 1 2 3 $ ./data-extract a.dat t2.dat 1-99 0 $ ./data-extract a.dat t3.dat 10 20 30 40 50 5 15 25 35 45 $ ./data-extract a.dat t4.dat 30-49 50-59 80-99 10-40 Some tools can help us to compare and examine binary files. Below are some examples. Lines that starts with '#' are comments explaining the following commands. See the manual pages for more options of cmp, xxd, and diff commands. # compare a.dat and b.dat $ cmp -b a.dat b.dat # compare a.dat and b.dat, and show detailed information $ cmp -bl a.dat b.dat # list groups of 4 bytes file a.dat in hexadecimal $ xxd -g4 a.dat # compare the hex dump of a.dat and b.dat $ diff <(xxd a.dat) <(xxd b.dat)

Datagen.c code:

#include <stdlib.h>

#include <stdio.h>

void my_error(char *s)

{

// print our own error message to standard error output

fprintf(stderr, "Error:%s\n", s);

// Use perror() to print the descriptive message about the error

perror("errno:");

exit(EXIT_FAILURE);

}

int main(int argc, char *argv[])

{

int n;

unsigned int seed;

if (argc != 4) {

fprintf(stderr, "Usage: %s <seed> <n> <filename>\n", argv[0]);

exit(EXIT_FAILURE);

}

// not checking the value of seed and n

seed = atoi(argv[1]);

n = atoi(argv[2]);

srand(seed);

// try to open the file specified on the command line

FILE *fp = fopen (argv[3], "w");

// check the return values!!

if (fp == NULL)

my_error("fopen() returned NULL.");

for (int i = 0; i < n; i ++) {

int r = rand(); // get a pseudo-random value

// write the integer to the file and check the return value!!

if (fwrite(&r, sizeof(r), 1, fp) != 1)

my_error("fwrite() failed");

}

fclose(fp);

return 0;

}

Data-extract skeleton code (fill in parts with TODO):

#include <stdlib.h>

#include <stdio.h>

void my_error(char *s)

{

// print our own error message to standard error output

fprintf(stderr, "Error:%s\n", s);

// Use perror() to print the descriptive message about the error

perror("errno");

exit(EXIT_FAILURE);

}

/* This function copy integers from file infp to file outfp.

* The integers copied are specified by start and end.  

*

* For example,

* start is 0, end is 0.

*      The first int in infp is copied to outfp.

* start is 10, end is 20.

*      11 integers are copied from infp to outfp, starting from integer 10.

*

* Functions that may be used in this function include

* fseek(), fread(), fwrite().

*

* Check the return values of function calls.

*

* Return values:

* 0:   success

* -1:  function calls failed

* */

int copy_integers(FILE *outfp, FILE *infp, int start, int end)

{

// TODO

return 0;

}

int main(int argc, char *argv[])

{

if (argc < 4) {

fprintf(stderr, "Usage: %s <input-filename> <output-filename> range [range ..]\n", argv[0]);

exit(EXIT_FAILURE);

}

FILE    *infp, *outfp;

// TODO

// open inputfile for read

// If it fails, call my_error("Cannot open input file.");

// open output for write

// If it fails, call my_error("Cannot open output file.");

for (int i = 3; i < argc; i ++) {

int start, end;

// example of using sscanf()

if (sscanf(argv[i], "%d-%d", &start, &end) != 2) {

if (sscanf(argv[i], "%d", &start) != 1)

my_error("Invalid range");

end = start;

}

if (start < 0 || end < 0 || end < start)

my_error("start and end must be >= 0 and start must be <= end");

if (copy_integers(outfp, infp, start, end))

my_error("copy_integers() reteurned a non-zero value.");

}

// TODO

// close files

// On error,

// call my_error("Cannot close input file.") or

//      my_error("Cannot close output file.")

return 0;

}