*
*/
-#define VERSION_NUMBER "0.8"
+#define VERSION_NUMBER "0.9"
#define _BSD_SOURCE
#include <locale.h>
#include <getopt.h>
#include <fcntl.h>
+#ifdef WITH_MD5
+#include <openssl/md5.h>
+#endif
/* 1M really helps compared to 64k */
#define READ_BUFFER_SIZE (1024 * 1024)
int tty_width = -1; /* Positive value means what width to use to show
the progress bar */
+#ifdef WITH_MD5
+int use_md5 = 0; /* 1 means we keep an MD5 signature for each file */
+#endif
+
/********************************************************************/
/* malloc with error checking. */
return
strcmp(name, ".") == 0 ||
strcmp(name, "..") == 0 ||
- (ignore_dotfiles && name[0] == '.');
+ (ignore_dotfiles && name[0] == '.' && name[1] != '/');
}
/**********************************************************************/
ino_t inode;
int group_id; /* one per identical file content */
int dir_id; /* 1 for DIR1, and 2 for DIR2 */
+#ifdef WITH_MD5
+ int md5_computed;
+ unsigned char md5[MD5_DIGEST_LENGTH];
+#endif
};
void file_list_delete(struct file_node *head) {
char *buffer1, char *buffer2) {
int fd1, fd2, s1, s2;
+#ifdef WITH_MD5
+ MD5_CTX c1, c2;
+
+ if(use_md5) {
+ if(f1->md5_computed && f2->md5_computed) {
+ if(!memcmp(f1->md5, f2->md5, MD5_DIGEST_LENGTH)) {
+ return 0;
+ }
+ } else {
+ if(!f1->md5_computed) {
+ MD5_Init(&c1);
+ }
+ if(!f2->md5_computed) {
+ MD5_Init(&c2);
+ }
+ }
+ }
+#endif
+
fd1 = open(f1->name, O_RDONLY);
fd2 = open(f2->name, O_RDONLY);
if(s1 == 0) {
close(fd1);
close(fd2);
+#ifdef WITH_MD5
+ if(use_md5) {
+ if(!f1->md5_computed) {
+ MD5_Final(f1->md5, &c1);
+ f1->md5_computed = 1;
+ }
+ if(!f2->md5_computed) {
+ MD5_Final(f2->md5, &c2);
+ f2->md5_computed = 1;
+ }
+ }
+#endif
return 1;
} else {
if(memcmp(buffer1, buffer2, s1)) {
close(fd2);
return 0;
}
+#ifdef WITH_MD5
+ if(use_md5) {
+ if(!f1->md5_computed) {
+ MD5_Update(&c1, buffer1, s1);
+ }
+ if(!f2->md5_computed) {
+ MD5_Update(&c2, buffer2, s2);
+ }
+ }
+#endif
}
} else {
fprintf(stderr,
f2->name,
strerror(errno));
}
+
exit(EXIT_FAILURE);
}
}
/**********************************************************************/
-struct file_node *scan_directory(struct file_node *tail,
- const char *name) {
+struct file_node *scan_directory(struct file_node *tail, const char *name) {
DIR *dir;
struct dirent *dir_e;
struct stat sb;
tmp->inode = sb.st_ino;
tmp->group_id = -1;
tmp->dir_id = -1;
+#ifdef WITH_MD5
+ tmp->md5_computed = 0;
+#endif
tail = tmp;
}
}
fprintf(out, "Version %s (%s)\n", VERSION_NUMBER, UNAME);
fprintf(out, "Without DIR2, lists duplicated files found in DIR1. With DIR2, lists files common to both directories. With the not: prefix, lists files found in DIR1 which do not exist in DIR2. The and: prefix is the default and should be used only if you have a directory starting with 'not:'\n");
fprintf(out, "\n");
- fprintf(out, " -h, --help\n");
- fprintf(out, " show this help\n");
- fprintf(out, " -d, --ignore-dots\n");
- fprintf(out, " ignore dot files and directories\n");
- fprintf(out, " -0, --ignore-empty\n");
- fprintf(out, " ignore empty files\n");
- fprintf(out, " -c, --hide-matchings\n");
- fprintf(out, " do not show which files in DIR2 corresponds to those in DIR1\n");
- fprintf(out, " -g, --no-group-ids\n");
- fprintf(out, " do not show the file groups\n");
- fprintf(out, " -p, --show-progress\n");
- fprintf(out, " show progress\n");
- fprintf(out, " -r, --real-paths\n");
- fprintf(out, " show the real file paths\n");
+ /* 01234567890123456789012345678901234567890123456789012345678901234567890123456789*/
+ fprintf(out, " -h, --help show this help\n");
+ fprintf(out, " -d, --ignore-dots ignore dot files and directories\n");
+ fprintf(out, " -0, --ignore-empty ignore empty files\n");
+ fprintf(out, " -c, --hide-matchings do not show which files in DIR2 corresponds to\n");
+ fprintf(out, " those in DIR1\n");
+ fprintf(out, " -g, --no-group-ids do not show the file groups\n");
+ fprintf(out, " -p, --show-progress show progress\n");
+ fprintf(out, " -r, --real-paths show the real file paths\n");
fprintf(out, " -i, --same-inodes-are-different\n");
- fprintf(out, " consider files with same inode as different\n");
+ fprintf(out, " consider files with same inode as different\n");
+#ifdef WITH_MD5
+ fprintf(out, " -m, --md5 use MD5 hashing\n");
+#endif
fprintf(out, "\n");
- fprintf(out, "Report bugs and comments to <francois@fleuret.org>\n");
+ fprintf(out, "Report bugs and comments to <francois@fleuret.org>.\n");
}
/**********************************************************************/
{ "ignore-dots", no_argument, 0, 'd' },
{ "ignore-empty", no_argument, 0, '0' },
{ "show-progress", no_argument, 0, 'p' },
+ { "md5", no_argument, 0, 'm' },
{ 0, 0, 0, 0 }
};
setlocale (LC_ALL, "");
- while ((c = getopt_long(argc, argv, "hircgd0p",
+ while ((c = getopt_long(argc, argv, "hircgd0pm",
long_options, NULL)) != -1) {
switch (c) {
show_hits = 0;
break;
+ case 'm':
+#ifdef WITH_MD5
+ use_md5 = 1;
+#else
+ fprintf(stderr,
+ "finddup has not be compiled with MD5 hashing.\n");
+ exit(EXIT_FAILURE);
+#endif
+ break;
+
default:
exit(EXIT_FAILURE);
}