X-Git-Url: http://git.hungrycats.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=faster-dupemerge;h=7e899ddc8eb8ae4f87770d42a33efe4acca43b95;hb=6910de85c5daa8b99a1ea7d6e4614584fbfdfec5;hp=f1413b47e10a7d60364c14d532efdcdb049a76f0;hpb=1c211f4055c67cfbf832d057d5b8bb3c7bf6d21c;p=dupemerge

diff --git a/faster-dupemerge b/faster-dupemerge
index f1413b4..7e899dd 100755
--- a/faster-dupemerge
+++ b/faster-dupemerge
@@ -4,7 +4,7 @@ use Fcntl qw(:DEFAULT :flock);
 use File::Compare;
 use File::Temp;
 
-# Copyright (C) 2003-2010 Zygo Blaxell <faster-dupemerge@mailtoo.hungrycats.org>
+# Copyright (C) 2002-2010 Zygo Blaxell <faster-dupemerge@mailtoo.hungrycats.org>
 
 # This program is free software; you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
@@ -134,24 +134,33 @@ hard links).
         --debug         show all steps in duplication discovery process
                         (implies --verbose)
 
-	--dry-run	do not lock files or make changes to filesystem
+        --dry-run       do not lock files or make changes to filesystem
 
         --find          pass next options (up to --) to find command
 
-	--humane	human-readable statistics (e.g. 1 048 576)
+        --humane        human-readable statistics (e.g. 1 048 576)
 
         --lock FILE     exit immediately (status 10) if unable to obtain a 
                         flock(LOCK_EX|LOCK_NB) on FILE
 
-	--lock-rm	remove lock file at exit
+        --lock-rm       remove lock file at exit
+
+        --progress      output single-character progress indicators:
+                        C - compare
+                        H - hash
+                        L - link(2)
+                        R - rename(2)
+                        S, s - lstat(2) (see source for details)
+                        U - unlink(2)
+                        . - all inodes with similar attributes done
 
         --sort          pass next options (up to --) to sort command
 
         --timestamps    mtime may be different for identical files
 
-	--skip-compare	skip byte-by-byte file comparisons
+        --skip-compare  skip byte-by-byte file comparisons
 
-	--skip-hash	skip calculation of hash function on files
+        --skip-hash     skip calculation of hash function on files
 
         --trust         old name for --skip-compare
                         (trust the hash function)
@@ -331,10 +340,11 @@ hash_file:
 			print STDERR "\t\tDigesting file $filename\n" if $debug;
 			if ((-l $filename) || ! -f _) {
 				warn "Bogon file " . tick_quote($filename);
-				$surprises++;
+				$input_bogons++;
+				delete $inode_to_file_name{$candidate}->{$filename};
 				next;
 			}
-			eval { 
+			eval {
 				$digest = digest($filename); 
 			};
 			if ($@) {
@@ -370,7 +380,7 @@ incumbent_file:
 						my ($incumbent_dev,$incumbent_ino,$incumbent_mode,$incumbent_nlink,$incumbent_uid,$incumbent_gid,$incumbent_rdev,$incumbent_size,$incumbent_atime,$incumbent_mtime,$incumbent_ctime,$incumbent_blksize,$incumbent_blocks) = lstat($incumbent_file);
 						print STDERR "\t\tINCUMBENT dev=$incumbent_dev ino=$incumbent_ino mode=$incumbent_mode nlink=$incumbent_nlink uid=$incumbent_uid gid=$incumbent_gid rdev=$incumbent_rdev size=$incumbent_size atime=$incumbent_atime mtime=$incumbent_mtime ctime=$incumbent_ctime blksize=$incumbent_blksize blocks=$incumbent_blocks _=$incumbent_file\n" if $debug;
 
-						if (!defined($incumbent_blocks)) {
+						if (!defined($incumbent_blocks) || ! (-f _)) {
 							warn "lstat: $incumbent_file: $!";
 							$surprises++;
 							next incumbent_file;
@@ -391,7 +401,7 @@ candidate_file:
 							my ($candidate_dev,$candidate_ino,$candidate_mode,$candidate_nlink,$candidate_uid,$candidate_gid,$candidate_rdev,$candidate_size,$candidate_atime,$candidate_mtime,$candidate_ctime,$candidate_blksize,$candidate_blocks) = lstat($candidate_file);
 							print STDERR "\t\t\tCANDIDATE dev=$candidate_dev ino=$candidate_ino mode=$candidate_mode nlink=$candidate_nlink uid=$candidate_uid gid=$candidate_gid rdev=$candidate_rdev size=$candidate_size atime=$candidate_atime mtime=$candidate_mtime ctime=$candidate_ctime blksize=$candidate_blksize blocks=$candidate_blocks _=$candidate_file\n" if $debug;
 
-							if (!defined($candidate_blocks)) {
+							if (!defined($candidate_blocks) || ! (-f _)) {
 								warn "lstat: $candidate_file: $!";
 								$surprises++;
 								next candidate_file;
@@ -553,10 +563,10 @@ candidate_file:
 			delete $inode_to_file_name{$candidate};
 		}
 	}
+	print STDERR '.' if $progress;
 
 end_merge:
 
-	print STDERR '.' if $progress;
 	print STDERR "Merge done.\n" if $debug;
 	undef %inode_to_file_name;
 }
@@ -569,12 +579,6 @@ while (<FIND>) {
 
 	print STDERR "weak_key=$weak_key inode=$inode name=$name\n" if $debug;
 
-	unless (! (-l $name) && (-f _)) {
-		warn "Bogon file " . tick_quote($name);
-		$input_bogons++;
-		next;
-	}
-
 	$input_links++;
 	merge_files if $weak_key ne $current_key;
 	$current_key = $weak_key;
@@ -633,6 +637,7 @@ if ($humane) {
 
 $stats_blob =~ s/([^\n]*\n[^\n]*? )(\s+)( [^\n]*\n)/$1 . ('.' x length($2)) . $3/oemg;
 
+print STDERR "\n" if $progress;
 print STDERR $stats_blob;
 
 exit(0);