From 630c461cbdb6610420863dcfe7af247d62bf004e Mon Sep 17 00:00:00 2001
From: Francois Fleuret <francois@fleuret.org>
Date: Wed, 10 Mar 2010 19:15:54 +0100
Subject: [PATCH] Added the -0 option to ignore empty files.

---
 finddup.1 |  3 +++
 finddup.c | 57 +++++++++++++++++++++++++++++++++++--------------------
 2 files changed, 39 insertions(+), 21 deletions(-)

diff --git a/finddup.1 b/finddup.1
index 71e0e57..e021a81 100644
--- a/finddup.1
+++ b/finddup.1
@@ -38,6 +38,9 @@ display help and exit
 \fB-d\fR
 ignore files and directories starting with a dot
 .TP
+\fB-0\fR
+ignore empty files
+.TP
 \fB-c\fR
 do not show which files from DIR2 corresponds to files from DIR1
 (hence, show only the files from DIR1 which have an identical twin in
diff --git a/finddup.c b/finddup.c
index adf138e..bca9c5c 100644
--- a/finddup.c
+++ b/finddup.c
@@ -42,6 +42,7 @@
 #include <fcntl.h>
 
 #define BUFFER_SIZE 4096
+#define LARGE_BUFFER_SIZE 65536
 
 typedef int64_t size_sum_t;
 
@@ -50,6 +51,8 @@ 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 */
 
@@ -135,15 +138,15 @@ int file_list_length(struct file_with_size *head) {
 
 int same_content(struct file_with_size *f1, struct file_with_size *f2) {
   int fd1, fd2, s1, s2;
-  char buffer1[BUFFER_SIZE], buffer2[BUFFER_SIZE];
+  char buffer1[LARGE_BUFFER_SIZE], buffer2[LARGE_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, LARGE_BUFFER_SIZE);
+      s2 = read(fd2, buffer2, LARGE_BUFFER_SIZE);
 
       if(s1 < 0 || s2 < 0) {
         close(fd1);
@@ -226,13 +229,17 @@ struct file_with_size *scan_directory(struct file_with_size *tail,
     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;
+          tail = tmp;
+        }
+      }
     }
   }
 
@@ -392,17 +399,19 @@ void start(const char *dirname1, const char *dirname2) {
       }
 
       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)) {
+            if(node1->group_id < 0) {
+              if(node2->group_id >= 0) {
+                node1->group_id = node2->group_id;
+              } else {
+                node1->group_id = k;
+                k++;
+              }
+            }
+            if(node2->group_id < 0) {
+              node2->group_id = node1->group_id;
             }
-          }
-          if(node2->group_id < 0) {
-            node2->group_id = node1->group_id;
           }
         }
       }
@@ -412,6 +421,7 @@ void start(const char *dirname1, const char *dirname2) {
   print_result(list1, list2);
 
   file_list_delete(list1);
+
   if(dirname2) {
     file_list_delete(list2);
   }
@@ -424,6 +434,7 @@ void print_help(FILE *out) {
   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");
@@ -444,7 +455,7 @@ int main(int argc, char **argv) {
   setlocale (LC_ALL, "");
 
   while (1) {
-    c = getopt(argc, argv, "hrcgdp");
+    c = getopt(argc, argv, "hrcgd0p");
     if (c == -1)
       break;
 
@@ -460,6 +471,10 @@ int main(int argc, char **argv) {
       ignore_dotfiles = 1;
       break;
 
+    case '0':
+      ignore_empty_files = 1;
+      break;
+
     case 'r':
       show_realpaths = 1;
       break;
-- 
2.20.1