dclean/dclean.c

#include "dclean.h"
#include "recordio.h"

char buffer[512]; // character buffer for sprintf

int main(int argc, char *argv[])
{
	if (argc != 3) {
		sprintf(buffer, "usage: dclean directory_master directory_copy\n");
		write(1, buffer, strlen(buffer));
		return EXIT_FAILURE;
	}
	char *master = argv[1]; // read in the arguments
	char *copy = argv[2];

	clean_txt();										// clean the directory
	find_directories(master, "dclean-master.txt");		// find all files in the master directory
	sortfile("dclean-master.txt", "dclean-master.txt"); // sort the file by filesize
	find_directories(copy, "dclean-copy.txt");			// find all of the files in the copy directory
	sortfile("dclean-copy.txt", "dclean-copy.txt");		// sort the file by filesize
	find_same();										// find which files are the same
}

//-----------------------------------------------------------
// FUNCTION clean_txt:
//      this is mostly here to make sure that all of the .txt files are cleared before the program runs
// PARAMETER USAGE :
//      none
// FUNCTION CALLED :
//      void run_external_process(int fds[2], const char** args);
//-----------------------------------------------------------
void clean_txt()
{
	const char *a[3] = {"rm", "dclean-master.txt", NULL};
	const char *b[3] = {"rm", "dclean-copy.txt", NULL};
	const char *c[3] = {"rm", ".rinx.dclean-master.txt", NULL};
	const char *d[3] = {"rm", ".rinx.dclean-copy.txt", NULL};
	int s[2]; // pointer for the pipe; (we dont actually use the pipe)
	run_external_process(s, a);
	run_external_process(s, b);
	run_external_process(s, c);
	run_external_process(s, d);
}

//-----------------------------------------------------------
// FUNCTION find_same:
//      Seaches through the Master and Copy files and checks to see if the files listed are the same using 'diff'
// PARAMETER USAGE :
//      none
// FUNCTION CALLED :
//      int rio_open(const char *pathname, int flags, mode_t mode);
//			int rio_lseek(int fd, int offset, int whence);
//			void * rio_read(int fd, int * return_value);
//			int identical(char * file1, char * file2);
//-----------------------------------------------------------
void find_same()
{
	const char s[2] = {'*', '\n'}; //* and \n character for use in tokenizer
	char *token;
	char *tokencp;
	char *rc;
	char *rccp;
	int r, rcp, i, j, ir, jr, idn, n;

	int fd = rio_open("dclean-master.txt", O_RDONLY, 0644); // open the list of master files
	if (fd < 0)
	{
		fprintf(stderr, "%s\n", strerror(EIO));
		exit(EXIT_FAILURE);
	}; // error checking

	int fdcp = rio_open("dclean-copy.txt", O_RDONLY, 0644); // open the list of copy files
	if (fdcp < 0)
	{
		fprintf(stderr, "%s\n", strerror(EIO));
		exit(EXIT_FAILURE);
	}; // error checking

	int fdsh = open("clean-duplicates.sh", O_RDWR | O_CREAT | O_TRUNC, 0644); // open the shell script
	if (fdsh < 0)
	{
		fprintf(stderr, "%s\n", strerror(EIO));
		exit(EXIT_FAILURE);
	}; // error checking

	n = write(fdsh, "#!/bin/sh\n", strlen("#!/bin/sh\n")); // write this to the top so the .sh works
	if (n < 0)
	{
		fprintf(stderr, "%s\n", strerror(EIO));
		exit(EXIT_FAILURE);
	}; // error checking

	int eol = rio_lseek(fd, 0, SEEK_END);	  // we now know the length of the master file
	int eolcp = rio_lseek(fdcp, 0, SEEK_END); // we now know the length of the copy file
	r = rio_lseek(fd, 0, SEEK_SET);			  // seek back to the beginning
	rcp = rio_lseek(fdcp, 0, SEEK_SET);

	for (i = 0; i < eol; i++)
	{							 // loop through all files in the master file
		ir = i;					 // we have to use this since rio_read returns a value back through &ir
		rc = rio_read(fd, &ir);	 // read the ith line of the file
		token = strtok(rc, s);	 // the first token will always be the filesize
		token = strtok(NULL, s); // this is the filepath

		for (j = 0; j < eolcp; j++)
		{								// loop through all files in the copy file
			jr = j;						// we have to use this since rio_read returns a value back through &jr
			rccp = rio_read(fdcp, &jr); // read the jth line of the file

			tokencp = strtok(rccp, s); // the first token will always be the filesize
			tokencp = strtok(NULL, s); // this is the filepath

			idn = identical(token, tokencp); // see if the two files are identical

			if (idn == 1)
			{											   // if they are identical then add them to the shell script
				n = write(fdsh, "rm ", strlen("rm "));	   // rm (remove command)
				n = write(fdsh, tokencp, strlen(tokencp)); // the file in question
				n = write(fdsh, "\n", strlen("\n"));
				if (n < 0)
				{
					fprintf(stderr, "%s\n", strerror(EIO));
					exit(EXIT_FAILURE);
				}; // error checking
			}

			free(rccp); // we free this as to reset the buffer
		}
		rio_lseek(fdcp, 0, SEEK_SET); // reset pointer because otherwise we would only search the master file 1 time
		free(rc);					  // free the buffer for the master file
	}
}

//-----------------------------------------------------------
// FUNCTION sortfile:
//			Sorts the inputfile using /usr/bin/sort
// PARAMETER USAGE :
//      char * inputfile : the file for input
//			char * outputfile : the file for output
// FUNCTION CALLED :
//      none
//-----------------------------------------------------------
int sortfile(char *inputfile, char *outputfile)
{
	// equivilent to: sort -n dclean-master.txt > sorted.txt
	unsigned char byte;
	const char *a[4] = {"sort", "-n", inputfile, NULL}; // construct arguments

#ifdef DEBUG
	sprintf(buffer, "Sorting File [%s]\n", inputfile);
	write(1, buffer, strlen(buffer));
#endif

	pid_t pid;	// the pid of the child/parent
	int status; // the waiting integer

	if ((pid = fork()) == 0)
	{																	// forking diff (child process)
		int fd = open(outputfile, O_RDWR | O_CREAT, S_IRUSR | S_IWUSR); // open file (create it if it doesnt exist)
		if (fd < 0)
		{
			fprintf(stderr, "Open (outputfile) Failed because of: %s\n", strerror(EIO));
			return -1;
		};
		dup2(fd, 1);	 // make stdout go to file
		execvp(a[0], a); // create the sort process
		exit(0);		 // exit the child
	}
	else if (pid < 0)
	{ // error checking
		sprintf(buffer, "Fork Failed!\n");
		write(1, buffer, strlen(buffer));
		exit(-1);
	}
	else
	{ // parent process
		while (wait(&status) != pid)
			; // wait for completion;
	}
	return 0;
}

//-----------------------------------------------------------
// FUNCTION find_directories:
//			Runs "iodir" to generate the .txt files
// PARAMETER USAGE :
//      char * file_path : the file for input
//			char * out_name : the file for output
// FUNCTION CALLED :
//      void run_external_process(int fds[2], const char** args);
//-----------------------------------------------------------
int find_directories(char *file_path, char *out_name)
{
	const char *a[4] = {"iodir", file_path, out_name, NULL}; // construct arguments
	int fds[2];												 // pointer for the pipe we will create in a line or so
#ifdef DEBUG
	sprintf(buffer, "Generating Directory File [%s] For path: [%s]\n", out_name, file_path);
	write(1, buffer, strlen(buffer));
#endif
	run_external_process(fds, a); // generate the files using iodir [I edited it slightly]
	return 0;
}

//-----------------------------------------------------------
// FUNCTION identical:
//			Runs "diff file1 file2" to generate see if they are the same
// PARAMETER USAGE :
//      char * file1 : "master" file
//			char * file2 : "copy" file
// FUNCTION CALLED :
//      void run_external_process(int fds[2], const char** args);
//-----------------------------------------------------------
int identical(char *file1, char *file2)
{
	const char *a[4] = {"diff", file1, file2, NULL}; // construct arguments
	int fds[2];										 // pointer for the pipe we will create in a line or so
	run_external_process(fds, a);					 // run diff
	// continue with parent process
	ssize_t size = read(fds[0], buffer, sizeof(buffer)); // read from the pipe
	if (size == 0)
	{ // diff returns zero when the files are the same
#ifdef DEBUG
		sprintf(buffer, "\"%s\" is a copy of \"%s\"\n", file2, file1);
		write(1, buffer, strlen(buffer));
#endif
		return 1; // they are identical
	}
	else
	{
#ifdef DEBUG
		sprintf(buffer, "\"%s\" and \"%s\" differ\n", file1, file2);
		write(1, buffer, strlen(buffer));
#endif
		return 0; // they are different
	}
	return -1; // something weird went on; throw an error
}

//-----------------------------------------------------------
// FUNCTION run_external_process:
//			Runs child process and returns the childs output to the parent
// PARAMETER USAGE :
//      int fds[2] : the pointers to the pipe
//			const char** args : the arguments for the execvp
// FUNCTION CALLED :
//      none
//-----------------------------------------------------------
void run_external_process(int fds[2], const char **args)
{
	pid_t pid;	// the pid of the child/parent
	int status; // the waiting integer
	pipe(fds);	// create a new pipe pipe

	if ((pid = fork()) == 0)
	{						   // forking the child process
		close(fds[0]);		   // close the input of the pipe
		close(1);			   // close stdout
		dup(fds[1]);		   // dup the pipes stdout (pipe now sends child's stdout)
		execvp(args[0], args); // create the diff process
		exit(0);			   // exit the child
	}
	else if (pid < 0)
	{ // error checking
		sprintf(buffer, "Fork Failed!\n");
		write(1, buffer, strlen(buffer));
		exit(-1);
	}
	else
	{				   // parent process
		close(fds[1]); // close output of pipe which redirects stdout of the child
		while (wait(&status) != pid)
			; // wait for completion;
	}
}