X-Git-Url: https://fleuret.org/cgi-bin/gitweb/gitweb.cgi?p=finddup.git;a=blobdiff_plain;f=finddup.c;h=1ea54cec256ed09f7b8d0d4dd9143a82756f145c;hp=7d5b0580024837894675718e1cb103896f598468;hb=0d200c3b81ec3672daea4cffd15551e4e3788d9d;hpb=5075f86b7871fc3bf9bb8304598d648ae0fe653e diff --git a/finddup.c b/finddup.c index 7d5b058..1ea54ce 100644 --- a/finddup.c +++ b/finddup.c @@ -23,7 +23,7 @@ * */ -#define VERSION_NUMBER "0.6" +#define VERSION_NUMBER "0.9" #define _BSD_SOURCE @@ -40,6 +40,9 @@ #include #include #include +#ifdef WITH_MD5 +#include +#endif /* 1M really helps compared to 64k */ #define READ_BUFFER_SIZE (1024 * 1024) @@ -65,12 +68,17 @@ int show_hits = 1; /* 1 means to show the files from dir2 int show_groups = 1; /* 1 means to show the group IDs when printing file names */ -int ignore_same_inode = 0; /* 1 means that comparison between two file - with same inode will always be false */ +int same_inodes_are_different = 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 */ +#ifdef WITH_MD5 +int use_md5 = 0; /* 1 means we keep an MD5 signature for each file */ +#endif + /********************************************************************/ /* malloc with error checking. */ @@ -90,45 +98,35 @@ int ignore_entry(const char *name) { return strcmp(name, ".") == 0 || strcmp(name, "..") == 0 || - (ignore_dotfiles && name[0] == '.'); -} - -void print_size_sum(size_sum_t s) { - char tmp[100]; - char *a = tmp + sizeof(tmp)/sizeof(char); - *(--a) = '\0'; - if(s) { - while(s) { - *(--a) = s%10 + '0'; - s /= 10; - } - } else { - *(--a) = '0'; - } - printf(a); + (ignore_dotfiles && name[0] == '.' && name[1] != '/'); } /**********************************************************************/ -struct file_with_size { - char *filename; +struct file_node { + struct file_node *next; + char *name; size_t size; ino_t inode; - struct file_with_size *next; - int group_id, dir_id; + 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_with_size *head) { - struct file_with_size *next; +void file_list_delete(struct file_node *head) { + struct file_node *next; while(head) { next = head->next; - free(head->filename); + free(head->name); free(head); head = next; } } -int file_list_length(struct file_with_size *head) { +int file_list_length(struct file_node *head) { int l = 0; while(head) { l++; @@ -139,12 +137,31 @@ int file_list_length(struct file_with_size *head) { /**********************************************************************/ -int same_content(struct file_with_size *f1, struct file_with_size *f2, +int same_content(struct file_node *f1, struct file_node *f2, char *buffer1, char *buffer2) { int fd1, fd2, s1, s2; - fd1 = open(f1->filename, O_RDONLY); - fd2 = open(f2->filename, O_RDONLY); +#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(fd1 >= 0 && fd2 >= 0) { while(1) { @@ -161,6 +178,18 @@ int same_content(struct file_with_size *f1, struct file_with_size *f2, 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)) { @@ -168,6 +197,16 @@ int same_content(struct file_with_size *f1, struct file_with_size *f2, 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, @@ -180,36 +219,37 @@ int same_content(struct file_with_size *f1, struct file_with_size *f2, if(fd1 < 0) { fprintf(stderr, "Can not open \"%s\" error: %s\n", - f1->filename, + f1->name, strerror(errno)); } if(fd2 < 0) { fprintf(stderr, "Can not open \"%s\" error: %s\n", - f2->filename, + f2->name, strerror(errno)); } + exit(EXIT_FAILURE); } } -int same_files(struct file_with_size *f1, struct file_with_size *f2, +int same_files(struct file_node *f1, struct file_node *f2, char *buffer1, char *buffer2) { - if(ignore_same_inode && f1->inode == f2->inode) { + if(same_inodes_are_different && f1->inode == f2->inode) { return 0; } + return f1->size == f2->size && same_content(f1, f2, buffer1, buffer2); } /**********************************************************************/ -struct file_with_size *scan_directory(struct file_with_size *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; - struct file_with_size *tmp; + struct file_node *tmp; char subname[PATH_MAX + 1]; if(lstat(name, &sb) != 0) { @@ -235,13 +275,16 @@ struct file_with_size *scan_directory(struct file_with_size *tail, if(S_ISREG(sb.st_mode)) { if(!ignore_entry(name)) { if(!ignore_empty_files || sb.st_size > 0) { - tmp = safe_malloc(sizeof(struct file_with_size)); + tmp = safe_malloc(sizeof(struct file_node)); tmp->next = tail; - tmp->filename = strdup(name); + tmp->name = strdup(name); tmp->size = sb.st_size; tmp->inode = sb.st_ino; tmp->group_id = -1; tmp->dir_id = -1; +#ifdef WITH_MD5 + tmp->md5_computed = 0; +#endif tail = tmp; } } @@ -251,10 +294,10 @@ struct file_with_size *scan_directory(struct file_with_size *tail, return tail; } -void print_file(struct file_with_size *node) { +void print_file(struct file_node *node) { char tmp[PATH_MAX + 1]; if(show_realpaths) { - if(realpath(node->filename, tmp)) { + if(realpath(node->name, tmp)) { if(show_groups) { printf("%d %s\n", node->group_id, tmp); } else { @@ -262,24 +305,24 @@ void print_file(struct file_with_size *node) { } } else { printf("Can not get the realpath of \"%s\": %s\n", - node->filename, + node->name, strerror(errno)); exit(EXIT_FAILURE); } } else { if(show_groups) { - printf("%d %s\n", node->group_id, node->filename); + printf("%d %s\n", node->group_id, node->name); } else { - printf("%s\n", node->filename); + printf("%s\n", node->name); } } } int compare_nodes(const void *x1, const void *x2) { - const struct file_with_size **f1, **f2; + const struct file_node **f1, **f2; - f1 = (const struct file_with_size **) x1; - f2 = (const struct file_with_size **) x2; + f1 = (const struct file_node **) x1; + f2 = (const struct file_node **) x2; if((*f1)->group_id < (*f2)->group_id) { return -1; @@ -297,9 +340,9 @@ int compare_nodes(const void *x1, const void *x2) { } -void print_result(struct file_with_size *list1, struct file_with_size *list2) { - struct file_with_size *node1, *node2; - struct file_with_size **nodes; +void print_result(struct file_node *list1, struct file_node *list2) { + struct file_node *node1, *node2; + struct file_node **nodes; int nb, n; nb = 0; @@ -313,7 +356,7 @@ void print_result(struct file_with_size *list1, struct file_with_size *list2) { } } - nodes = safe_malloc(nb * sizeof(struct file_with_size *)); + nodes = safe_malloc(nb * sizeof(struct file_node *)); n = 0; for(node1 = list1; node1; node1 = node1->next) { @@ -330,7 +373,7 @@ void print_result(struct file_with_size *list1, struct file_with_size *list2) { } } - qsort(nodes, nb, sizeof(struct file_with_size *), compare_nodes); + qsort(nodes, nb, sizeof(struct file_node *), compare_nodes); for(n = 0; n < nb; n++) { if(!show_groups && n > 0 && nodes[n]->group_id != nodes[n-1]->group_id) { @@ -363,11 +406,12 @@ void print_progress(int max, int n, int *pp) { } void start(const char *dirname1, const char *dirname2) { - struct file_with_size *list1, *list2; - struct file_with_size *node1, *node2; + struct file_node *list1, *list2; + struct file_node *node1, *node2; int not_in, found; int nb_groups, nb_nodes; int list1_length, previous_progress; + struct winsize win; char *buffer1 = safe_malloc(sizeof(char) * READ_BUFFER_SIZE); char *buffer2 = safe_malloc(sizeof(char) * READ_BUFFER_SIZE); @@ -375,6 +419,10 @@ void start(const char *dirname1, const char *dirname2) { not_in = 0; if(show_progress) { + if(isatty(STDOUT_FILENO) && + !ioctl (STDOUT_FILENO, TIOCGWINSZ, (char *) &win)) { + tty_width = win.ws_col; + } fprintf(stderr, "Scanning %s ... ", dirname1); } @@ -421,9 +469,9 @@ void start(const char *dirname1, const char *dirname2) { if(!found) { if(show_realpaths) { - printf("%s\n", realpath(node1->filename, 0)); + printf("%s\n", realpath(node1->name, 0)); } else { - printf("%s\n", node1->filename); + printf("%s\n", node1->name); } } } @@ -477,34 +525,46 @@ void print_help(FILE *out) { 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 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"); - fprintf(out, " -r show the real file paths\n"); - fprintf(out, " -i consider files with same inode as different\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"); +#ifdef WITH_MD5 + fprintf(out, " -m, --md5 use MD5 hashing\n"); +#endif fprintf(out, "\n"); - fprintf(out, "Report bugs and comments to \n"); + fprintf(out, "Report bugs and comments to .\n"); } /**********************************************************************/ +static struct option long_options[] = { + { "help", no_argument, 0, 'h' }, + { "same-inodes-are-different", no_argument, 0, 'i' }, + { "real-paths", no_argument, 0, 'r' }, + { "hide-matchings", no_argument, 0, 'c' }, + { "no-group-ids", no_argument, 0, 'g' }, + { "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 } +}; + 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, "hircgd0p"); - if (c == -1) - break; - + while ((c = getopt_long(argc, argv, "hircgd0pm", + long_options, NULL)) != -1) { switch (c) { case 'h': @@ -526,7 +586,7 @@ int main(int argc, char **argv) { break; case 'i': - ignore_same_inode = 1; + same_inodes_are_different = 1; break; case 'g': @@ -541,21 +601,25 @@ int main(int argc, char **argv) { 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); } } - 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) { - ignore_same_inode = 1; + same_inodes_are_different = 1; start(argv[optind], 0); } else { print_help(stderr);