1 .TH "FINDDUP" "1.1" "Apr 2010" "Francois Fleuret" "User Commands"
3 \" This man page was written by Francois Fleuret <francois@fleuret.org>
4 \" and is distributed under a Creative Commons Attribution-Share Alike
9 finddup \- Find files common to two directories (or not)
13 \fBfinddup\fP [OPTION]... [DIR1 [[and:|not:]DIR2]]
17 With one directory as argument, \fBfinddup\fP prints the duplicated
18 files found in it. If no directory is provided, it uses the current
21 With two directories, it prints either the files common to both DIR1
22 and DIR2 or, with the `not:' prefix, the ones present in DIR1 and not
23 in DIR2. The `and:' prefix is assumed by default and necessary only if
24 you have a directory name starting with `not:'.
26 This command compares files by first comparing their sizes, hence goes
29 When looking for identical files, \fBfinddup\fP associates a group ID
30 to every content, and prints it along the file names. Use the \fB-g\fP
35 is virtually the same as
40 \fB-h\fR, \fB--help\fR
43 \fB-d\fR, \fB--ignore-dots\fR
44 ignore files and directories starting with a dot
46 \fB-0\fR, \fB--ignore-empty\fR
49 \fB-c\fR, \fB--hide-matchings\fR
50 do not show which files from DIR2 correspond to files from DIR1
51 (hence, show only the files from DIR1 which have an identical twin in
54 \fB-g\fR, \fB--no-group-ids\fR
55 do not show the file group IDs
57 \fB-t\fR, \fB--time-sort\fR
58 sort files in each group according to the modification times
60 \fB-p\fR, \fB--show-progress\fR
61 show progress information in stderr
63 \fB-r\fR, \fB--real-paths\fR
64 show the real path of the files
66 \fB-i\fR, \fB--same-inodes-are-different\fR
67 files with same inode are considered as different
71 None known, probably many. Valgrind does not complain though.
73 Since files with same inodes are considered as different when looking
74 for duplicates in a single directory, there are weird behaviors -- not
75 bugs -- with hard links.
77 The current algorithm is dumb, as it does not use any hashing of the
80 Here are the things I tried, which did not help at all: (1) Computing
81 md5s on the whole files, which is not satisfactory because files are
82 often not read entirely, hence the md5s can not be properly computed,
83 (2) computing XORs of the first 4, 16 and 256 bytes with rejection as
84 soon as one does not match, (3) reading files in parts of increasing
85 sizes so that rejection could be done with only a small fraction read
86 when possible, (4) using mmap instead of open/read.
90 The format of the output should definitely be improved. Not clear how.
92 Their could be some fancy option to link two instances of the command
93 running on different machines to reduce network disk accesses. This
94 may not help much though.
101 List duplicated files in directory ./blah/, show a progress bar,
102 ignore empty files, and ignore files and directories starting with a
106 .B finddup sources not:/mnt/backup
109 List all files found in \fB./sources/\fR which do not have
110 content-matching equivalent in \fB/mnt/backup/\fR.
113 .B finddup -g tralala cuicui
116 List groups of files with same content which exist both in
117 \fB./tralala/\fR and \fB./cuicui/\fR. Do not show group IDs, instead
118 write empty lines between groups of files of same content.
122 Written by Francois Fleuret <francois@fleuret.org> and distributed
123 under the terms of the GNU General Public License version 3 as
124 published by the Free Software Foundation. This is free software: you
125 are free to change and redistribute it. There is NO WARRANTY, to the
126 extent permitted by law.