From: Francois Fleuret <francois@fleuret.org>
Date: Wed, 15 Dec 2010 19:03:54 +0000 (+0100)
Subject: Added the -e option to execute a command for each group of identical files.
X-Git-Url: https://www.fleuret.org/cgi-bin/gitweb/gitweb.cgi?p=finddup.git;a=commitdiff_plain;h=069e440749c66d6fd7a9aa89ef7c31df1b3f34ea

Added the -e option to execute a command for each group of identical files.
---

diff --git a/finddup.1 b/finddup.1
index abfc209..8a284ec 100644
--- a/finddup.1
+++ b/finddup.1
@@ -68,6 +68,10 @@ show the real path of the files
 .TP
 \fB-i\fR, \fB--same-inodes-are-different\fR
 files with same inode are considered as different
+.TP
+\fB-e \fI<command>\fR, \fB--exec \fI<command>\fR
+execute the provided command for each group of identical files, with
+their names as arguments
 
 .SH "BUGS"
 
diff --git a/finddup.c b/finddup.c
index b1a8474..c0a51a1 100644
--- a/finddup.c
+++ b/finddup.c
@@ -30,6 +30,7 @@
 #include <sys/types.h>
 #include <sys/stat.h>
 #include <sys/param.h>
+#include <sys/wait.h>
 #include <dirent.h>
 #include <stdlib.h>
 #include <stdio.h>
@@ -74,6 +75,9 @@ int same_inodes_are_different = 0; /* 1 means that comparison between
 int sort_by_time = 0; /* 1 means to sort files in each group according
                          to the modification time */
 
+char *command_to_exec = 0; /* the name of the command to exec for each
+                              group of identical files */
+
 /********************************************************************/
 
 /* malloc with error checking.  */
@@ -323,6 +327,49 @@ int compare_nodes(const void *x1, const void *x2) {
   }
 }
 
+void exec_command(int nb, struct file_node **nodes) {
+  char **args;
+  int max_group_size = 0, group_size, m, n, status;
+  pid_t pid;
+
+  for(n = 0; n < nb; n++) {
+    if(n == 0 || nodes[n]->group_id != nodes[n-1]->group_id) {
+      group_size = 0;
+    }
+    group_size++;
+    if(group_size > max_group_size) {
+      max_group_size = group_size;
+    }
+  }
+
+  args = safe_malloc((max_group_size + 2) * sizeof(char *));
+  args[0] = command_to_exec;
+
+  n = 0;
+  while(n < nb) {
+    m = n;
+    while(n < nb && nodes[n]->group_id == nodes[m]->group_id) {
+      args[n - m + 1] = nodes[n]->name;
+      n++;
+    }
+    args[n - m + 1] = 0;
+    pid = fork();
+    if(pid < 0) {
+    } else if(pid == 0) {
+      if(execvp(command_to_exec, args) < 0) {
+        perror("execvp");
+        exit(EXIT_FAILURE);
+      }
+    } else {
+      while(wait(&status) != pid);
+      if(status > 0) {
+        exit(EXIT_FAILURE);
+      }
+    }
+  }
+
+  free(args);
+}
 
 void print_result(struct file_node *list1, struct file_node *list2) {
   struct file_node *node1, *node2;
@@ -359,11 +406,15 @@ void print_result(struct file_node *list1, struct file_node *list2) {
 
   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) {
-      printf("\n");
+  if(command_to_exec) {
+    exec_command(nb, nodes);
+  } else {
+    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]);
     }
-    print_file(nodes[n]);
   }
 
   free(nodes);
@@ -544,6 +595,9 @@ void usage(FILE *out) {
   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");
+  fprintf(out, "   -e <command>, --exec <command>\n");
+  fprintf(out, "                              execute the provided command for each group of\n");
+  fprintf(out, "                              identical files, with their names as arguments\n");
   fprintf(out, "\n");
   fprintf(out, "Report bugs and comments to <francois@fleuret.org>.\n");
 }
@@ -561,6 +615,7 @@ static struct option long_options[] = {
   { "ignore-dots", no_argument, 0, 'd' },
   { "ignore-empty", no_argument, 0, '0' },
   { "show-progress", no_argument, 0, 'p' },
+  { "exec", 1, 0, 'e' },
   { 0, 0, 0, 0 }
 };
 
@@ -569,7 +624,7 @@ int main(int argc, char **argv) {
 
   setlocale (LC_ALL, "");
 
-  while ((c = getopt_long(argc, argv, "vhircgtd0pm",
+  while ((c = getopt_long(argc, argv, "vhircgtd0pme:",
                           long_options, NULL)) != -1) {
     switch (c) {
 
@@ -616,6 +671,13 @@ int main(int argc, char **argv) {
       show_hits = 0;
       break;
 
+    case 'e':
+      if(command_to_exec != 0) {
+        free(command_to_exec);
+      }
+      command_to_exec = strdup(optarg);
+      break;
+
     default:
       usage(stderr);
       exit(EXIT_FAILURE);