X-Git-Url: https://fleuret.org/cgi-bin/gitweb/gitweb.cgi?a=blobdiff_plain;ds=sidebyside;f=finddup.c;h=05c399959c4edb1dba7eb382b649614ebbdeebdb;hb=97491f6c7870239ac72bf775215904a948828f33;hp=9b65d40165c41ccce0f266091ad1be24102fda05;hpb=35064530c38bcca4897f684b07d8de6ca25b0e47;p=finddup.git

diff --git a/finddup.c b/finddup.c
index 9b65d40..05c3999 100644
--- 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>
@@ -139,9 +139,9 @@ 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_with_size *f1, struct file_with_size *f2,
+                 char *buffer1, char *buffer2) {
   int fd1, fd2, s1, s2;
-  char buffer1[READ_BUFFER_SIZE], buffer2[READ_BUFFER_SIZE];
 
   fd1 = open(f1->filename, O_RDONLY);
   fd2 = open(f2->filename, O_RDONLY);
@@ -194,11 +194,12 @@ int same_content(struct file_with_size *f1, struct file_with_size *f2) {
   }
 }
 
-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);
 }
 
 /**********************************************************************/
@@ -341,7 +342,7 @@ 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, "+");
@@ -350,8 +351,8 @@ 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);
     }
   }
 }
@@ -362,6 +363,9 @@ void start(const char *dirname1, const char *dirname2) {
   int not_in, found;
   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(show_progress) {
@@ -373,10 +377,11 @@ 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);
@@ -403,7 +408,7 @@ void start(const char *dirname1, const char *dirname2) {
       found = 0;
 
       for(node2 = list2; !found && node2; node2 = node2->next) {
-        if(same_files(node1, node2)) {
+        if(same_files(node1, node2, buffer1, buffer2)) {
           found = 1;
         }
       }
@@ -424,7 +429,7 @@ void start(const char *dirname1, const char *dirname2) {
 
       for(node2 = list2; node2; node2 = node2->next) {
         if(node1->group_id < 0 || node2->group_id < 0) {
-          if(same_files(node1, node2)) {
+          if(same_files(node1, node2, buffer1, buffer2)) {
             if(node1->group_id < 0) {
               if(node2->group_id >= 0) {
                 node1->group_id = node2->group_id;
@@ -456,12 +461,15 @@ void start(const char *dirname1, const char *dirname2) {
     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");