Moved to version 0.7

[finddup.git] / finddup.c

diff --git a/finddup.c b/finddup.c
index ab345d3..90c1351 100644 (file)
--- a/finddup.c
+++ b/finddup.c
@@ -1,8 +1,8 @@
 
 /*
- *  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>
@@ -23,7 +23,7 @@
  *
  */
 
-#define VERSION_NUMBER "0.6"
+#define VERSION_NUMBER "0.7"
 
 #define _BSD_SOURCE
 
@@ -93,21 +93,6 @@ int ignore_entry(const char *name) {
     (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);
-}
-
 /**********************************************************************/
 
 struct file_with_size {
@@ -254,12 +239,17 @@ struct file_with_size *scan_directory(struct file_with_size *tail,
 void print_file(struct file_with_size *node) {
   char tmp[PATH_MAX + 1];
   if(show_realpaths) {
-    if(show_groups) {
-      realpath(node->filename, tmp);
-      printf("%d %s\n", node->group_id, tmp);
+    if(realpath(node->filename, tmp)) {
+      if(show_groups) {
+        printf("%d %s\n", node->group_id, tmp);
+      } else {
+        printf("%s\n", tmp);
+      }
     } else {
-      realpath(node->filename, tmp);
-      printf("%s\n", tmp);
+      printf("Can not get the realpath of \"%s\": %s\n",
+             node->filename,
+             strerror(errno));
+      exit(EXIT_FAILURE);
     }
   } else {
     if(show_groups) {
@@ -342,7 +332,7 @@ void print_progress(int max, int n, int *pp) {
   int width;
   if(show_progress && tty_width > 0) {
     width = tty_width - 7;
-    p = (width * n) / max;
+    p = (width * n) / (max - 1);
     if(p > *pp) {
       for(k = 0; k < p; k++) {
         fprintf(stderr, "+");
@@ -351,8 +341,8 @@ void print_progress(int max, int n, int *pp) {
         fprintf(stderr, "-");
       }
       *pp = p;
-      p = (100 * n) / max;
-      fprintf(stderr, " [% 3d%%]\r", p);
+      p = (100 * n) / (max - 1);
+      fprintf(stderr, " [%3d%%]\r", p);
     }
   }
 }
@@ -361,7 +351,8 @@ 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, pp, l1, n;
+  int nb_groups, nb_nodes;
+  int list1_length, previous_progress;
 
   char *buffer1 = safe_malloc(sizeof(char) * READ_BUFFER_SIZE);
   char *buffer2 = safe_malloc(sizeof(char) * READ_BUFFER_SIZE);
@@ -377,10 +368,11 @@ void start(const char *dirname1, const char *dirname2) {
   if(dirname2) {
     if(strncmp(dirname2, "not:", 4) == 0) {
       not_in = 1;
-      /* We should show groups even in that mode. However they are not
-         properly calculated for now, so we force it off. */
+      /* 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) {
       fprintf(stderr, "%s ... ", dirname2);
@@ -394,15 +386,15 @@ void start(const char *dirname1, const char *dirname2) {
     fprintf(stderr, "done.\n");
   }
 
-  k = 0;
-  pp = -1;
-  n = 0;
-  l1 = file_list_length(list1);
+  nb_groups = 0;
+  previous_progress = -1;
+  nb_nodes = 0;
+  list1_length = file_list_length(list1);
 
   if(not_in) {
     for(node1 = list1; node1; node1 = node1->next) {
-      print_progress(l1, n, &pp);
-      n++;
+      print_progress(list1_length, nb_nodes, &previous_progress);
+      nb_nodes++;
 
       found = 0;
 
@@ -423,8 +415,8 @@ void start(const char *dirname1, const char *dirname2) {
 
   } else {
     for(node1 = list1; node1; node1 = node1->next) {
-      print_progress(l1, n, &pp);
-      n++;
+      print_progress(list1_length, nb_nodes, &previous_progress);
+      nb_nodes++;
 
       for(node2 = list2; node2; node2 = node2->next) {
         if(node1->group_id < 0 || node2->group_id < 0) {
@@ -433,9 +425,9 @@ void start(const char *dirname1, const char *dirname2) {
               if(node2->group_id >= 0) {
                 node1->group_id = node2->group_id;
               } else {
-                node1->group_id = k;
+                node1->group_id = nb_groups;
                 node1->dir_id = 1;
-                k++;
+                nb_groups++;
               }
             }
             if(node2->group_id < 0) {
@@ -466,9 +458,9 @@ void start(const char *dirname1, const char *dirname2) {
 }
 
 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");