/*
- * finddup is a simple utility find duplicated files, files common to
- * several directories, or files present in one directory and not in
- * another one.
+ * finddup is a simple utility to find duplicated files, files common
+ * to several directories, or files present in one directory and not
+ * in another one.
*
* Copyright (c) 2010 Francois Fleuret
* Written by Francois Fleuret <francois@fleuret.org>
#include <getopt.h>
#include <fcntl.h>
-#define BUFFER_SIZE 4096
+/* 1M really helps compared to 64k */
+#define READ_BUFFER_SIZE (1024 * 1024)
typedef int64_t size_sum_t;
int ignore_dotfiles = 0; /* 1 means ignore files and directories
starting with a dot */
+int ignore_empty_files = 0; /* 1 means ignore empty files */
+
int show_realpaths = 0; /* 1 means ignore files and directories
starting with a dot */
int ignore_same_inode = 0; /* 1 means that comparison between two file
with same inode will always be false */
+int tty_width = -1; /* Positive value means what width to use to show
+ the progress bar */
+
/********************************************************************/
/* malloc with error checking. */
size_t size;
ino_t inode;
struct file_with_size *next;
- int group_id;
+ int group_id, dir_id;
};
void file_list_delete(struct file_with_size *head) {
/**********************************************************************/
-int same_content(struct file_with_size *f1, struct file_with_size *f2) {
+int same_content(struct file_with_size *f1, struct file_with_size *f2,
+ char *buffer1, char *buffer2) {
int fd1, fd2, s1, s2;
- char buffer1[BUFFER_SIZE], buffer2[BUFFER_SIZE];
fd1 = open(f1->filename, O_RDONLY);
fd2 = open(f2->filename, O_RDONLY);
if(fd1 >= 0 && fd2 >= 0) {
while(1) {
- s1 = read(fd1, buffer1, BUFFER_SIZE);
- s2 = read(fd2, buffer2, BUFFER_SIZE);
+ s1 = read(fd1, buffer1, READ_BUFFER_SIZE);
+ s2 = read(fd2, buffer2, READ_BUFFER_SIZE);
if(s1 < 0 || s2 < 0) {
close(fd1);
close(fd2);
return 1;
} else {
- if(strncmp(buffer1, buffer2, s1)) {
+ if(memcmp(buffer1, buffer2, s1)) {
close(fd1);
close(fd2);
return 0;
}
}
-int same_files(struct file_with_size *f1, struct file_with_size *f2) {
+int same_files(struct file_with_size *f1, struct file_with_size *f2,
+ char *buffer1, char *buffer2) {
if(ignore_same_inode && f1->inode == f2->inode) {
return 0;
}
- return f1->size == f2->size && same_content(f1, f2);
+ return f1->size == f2->size && same_content(f1, f2, buffer1, buffer2);
}
/**********************************************************************/
closedir(dir);
} else {
if(S_ISREG(sb.st_mode)) {
- tmp = safe_malloc(sizeof(struct file_with_size));
- tmp->next = tail;
- tmp->filename = strdup(name);
- tmp->size = sb.st_size;
- tmp->inode = sb.st_ino;
- tmp->group_id = -1;
- tail = tmp;
+ if(!ignore_entry(name)) {
+ if(!ignore_empty_files || sb.st_size > 0) {
+ tmp = safe_malloc(sizeof(struct file_with_size));
+ tmp->next = tail;
+ tmp->filename = strdup(name);
+ tmp->size = sb.st_size;
+ tmp->inode = sb.st_ino;
+ tmp->group_id = -1;
+ tmp->dir_id = -1;
+ tail = tmp;
+ }
+ }
}
}
} else if((*f1)->group_id > (*f2)->group_id) {
return 1;
} else {
- return 0;
+ if((*f1)->dir_id < (*f2)->dir_id) {
+ return -1;
+ } else if((*f1)->dir_id > (*f2)->dir_id) {
+ return 1;
+ } else {
+ return 0;
+ }
}
}
if(node1->group_id >= 0) { nb++; }
}
- if(show_hits) {
+ if(list2) {
for(node2 = list2; node2; node2 = node2->next) {
if(node2->group_id >= 0) { nb++; }
}
}
}
- if(show_hits) {
+ if(list2) {
for(node2 = list2; node2; node2 = node2->next) {
if(node2->group_id >= 0) {
nodes[n++] = node2;
qsort(nodes, nb, sizeof(struct file_with_size *), compare_nodes);
for(n = 0; n < nb; n++) {
+ if(!show_groups && n > 0 && nodes[n]->group_id != nodes[n-1]->group_id) {
+ printf("\n");
+ }
print_file(nodes[n]);
}
free(nodes);
}
+void print_progress(int max, int n, int *pp) {
+ int p, k;
+ int width;
+ if(show_progress && tty_width > 0) {
+ width = tty_width - 7;
+ p = (width * n) / (max - 1);
+ if(p > *pp) {
+ for(k = 0; k < p; k++) {
+ fprintf(stderr, "+");
+ }
+ for(; k < width; k++) {
+ fprintf(stderr, "-");
+ }
+ *pp = p;
+ p = (100 * n) / (max - 1);
+ fprintf(stderr, " [%3d%%]\r", p);
+ }
+ }
+}
+
void start(const char *dirname1, const char *dirname2) {
struct file_with_size *list1, *list2;
struct file_with_size *node1, *node2;
int not_in, found;
- int k, p, pp, l1, n;
+ int k, pp, l1, n;
+
+ char *buffer1 = safe_malloc(sizeof(char) * READ_BUFFER_SIZE);
+ char *buffer2 = safe_malloc(sizeof(char) * READ_BUFFER_SIZE);
not_in = 0;
if(dirname2) {
if(strncmp(dirname2, "not:", 4) == 0) {
not_in = 1;
+ /* groups are not computed in the not: mode */
+ show_groups = 0;
+ dirname2 += 4;
+ } else if(strncmp(dirname2, "and:", 4) == 0) {
dirname2 += 4;
}
if(show_progress) {
if(not_in) {
for(node1 = list1; node1; node1 = node1->next) {
- if(show_progress) {
- p = (100 * n)/l1;
- if(p > pp) {
- fprintf(stderr, "%d%%\n", p);
- pp = p;
- }
- n++;
- }
+ print_progress(l1, n, &pp);
+ n++;
found = 0;
for(node2 = list2; !found && node2; node2 = node2->next) {
- if(same_files(node1, node2)) {
+ if(same_files(node1, node2, buffer1, buffer2)) {
found = 1;
}
}
} else {
for(node1 = list1; node1; node1 = node1->next) {
- if(show_progress) {
- p = (100 * n)/l1;
- if(p > pp) {
- fprintf(stderr, "%d%%\n", p);
- pp = p;
- }
- n++;
- }
+ print_progress(l1, n, &pp);
+ n++;
for(node2 = list2; node2; node2 = node2->next) {
- if(same_files(node1, node2)) {
- if(node1->group_id < 0) {
- if(node2->group_id >= 0) {
- node1->group_id = node2->group_id;
- } else {
- node1->group_id = k;
- k++;
+ if(node1->group_id < 0 || node2->group_id < 0) {
+ if(same_files(node1, node2, buffer1, buffer2)) {
+ if(node1->group_id < 0) {
+ if(node2->group_id >= 0) {
+ node1->group_id = node2->group_id;
+ } else {
+ node1->group_id = k;
+ node1->dir_id = 1;
+ k++;
+ }
+ }
+ if(node2->group_id < 0) {
+ node2->group_id = node1->group_id;
+ node2->dir_id = 2;
}
- }
- if(node2->group_id < 0) {
- node2->group_id = node1->group_id;
}
}
}
}
}
- print_result(list1, list2);
+ if(show_progress) {
+ fprintf(stderr, "\n");
+ }
- file_list_delete(list1);
if(dirname2) {
+ print_result(list1, list2);
+ file_list_delete(list1);
file_list_delete(list2);
+ } else {
+ print_result(list1, 0);
+ file_list_delete(list1);
}
+
+ free(buffer1);
+ free(buffer2);
}
void print_help(FILE *out) {
- fprintf(out, "Usage: finddup [OPTION]... DIR1 [[not:]DIR2]\n");
+ fprintf(out, "Usage: finddup [OPTION]... DIR1 [[and:|not:]DIR2]\n");
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.\n");
+ 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 show this help\n");
fprintf(out, " -d ignore dot files and directories\n");
+ fprintf(out, " -0 ignore empty files\n");
fprintf(out, " -c do not show which files in DIR2 corresponds to those in DIR1\n");
fprintf(out, " -g do not show the file groups\n");
fprintf(out, " -p show progress\n");
int main(int argc, char **argv) {
int c;
struct file_with_size *root;
+ struct winsize win;
root = 0;
setlocale (LC_ALL, "");
while (1) {
- c = getopt(argc, argv, "hrcgdp");
+ c = getopt(argc, argv, "hircgd0p");
if (c == -1)
break;
ignore_dotfiles = 1;
break;
+ case '0':
+ ignore_empty_files = 1;
+ break;
+
case 'r':
show_realpaths = 1;
break;
}
}
+ if(show_progress &&
+ isatty(STDOUT_FILENO) &&
+ !ioctl (STDOUT_FILENO, TIOCGWINSZ, (char *) &win)) {
+ tty_width = win.ws_col;
+ }
+
if(optind + 2 == argc) {
start(argv[optind], argv[optind + 1]);
} else if(optind + 1 == argc) {