line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
package Linux::NFS::BigDir; |
2
|
1
|
|
|
1
|
|
530
|
use strict; |
|
1
|
|
|
|
|
1
|
|
|
1
|
|
|
|
|
24
|
|
3
|
1
|
|
|
1
|
|
3
|
use warnings; |
|
1
|
|
|
|
|
1
|
|
|
1
|
|
|
|
|
27
|
|
4
|
1
|
|
|
1
|
|
8
|
use Exporter 'import'; |
|
1
|
|
|
|
|
1
|
|
|
1
|
|
|
|
|
25
|
|
5
|
1
|
|
|
1
|
|
3
|
use Carp; |
|
1
|
|
|
|
|
1
|
|
|
1
|
|
|
|
|
69
|
|
6
|
1
|
|
|
1
|
|
7
|
use constant BUF_SIZE => 4096; |
|
1
|
|
|
|
|
1
|
|
|
1
|
|
|
|
|
53
|
|
7
|
1
|
|
|
1
|
|
651
|
use File::Temp 'tempfile'; |
|
1
|
|
|
|
|
12418
|
|
|
1
|
|
|
|
|
49
|
|
8
|
1
|
|
|
1
|
|
5
|
use Fcntl; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
368
|
|
9
|
|
|
|
|
|
|
require 'syscall.ph'; |
10
|
|
|
|
|
|
|
|
11
|
|
|
|
|
|
|
our $VERSION = '0.001'; # VERSION |
12
|
|
|
|
|
|
|
|
13
|
|
|
|
|
|
|
=pod |
14
|
|
|
|
|
|
|
|
15
|
|
|
|
|
|
|
=head1 NAME |
16
|
|
|
|
|
|
|
|
17
|
|
|
|
|
|
|
Linux::NFS::BigDir - use Linux getdents syscall to read large directories over NFS |
18
|
|
|
|
|
|
|
|
19
|
|
|
|
|
|
|
=head1 SYNOPSIS |
20
|
|
|
|
|
|
|
|
21
|
|
|
|
|
|
|
use Linux::NFS::BigDir qw(getdents); |
22
|
|
|
|
|
|
|
# entries_ref is an array reference |
23
|
|
|
|
|
|
|
my $entries_ref = getdents($very_large_dir); |
24
|
|
|
|
|
|
|
|
25
|
|
|
|
|
|
|
=head1 DESCRIPTION |
26
|
|
|
|
|
|
|
|
27
|
|
|
|
|
|
|
This module was created to solve a very specific problem: you have a directory over NFS, mounted by |
28
|
|
|
|
|
|
|
a Linux OS, and that directory has a very large number of items (files, directories, etc). The number of entries |
29
|
|
|
|
|
|
|
is so large that you have trouble to list the contents with C or even C from the shell. In extreme |
30
|
|
|
|
|
|
|
cases, the operation just "hangs" and will provide a feedback hours later. |
31
|
|
|
|
|
|
|
|
32
|
|
|
|
|
|
|
I observed this behavior only with NFS version 3 (and wasn't able to simulate it with local EXT3/EXT4): you might find in different situations, |
33
|
|
|
|
|
|
|
but in that case it migh be a wrong configuration regarding the filesystem. Ask your administrator first. |
34
|
|
|
|
|
|
|
|
35
|
|
|
|
|
|
|
If you can't fix (or get fixed) the problem, then you might want to try to use this module. It will use the C |
36
|
|
|
|
|
|
|
syscall from Linux. You can check the documentation about this syscall with C in a shell. |
37
|
|
|
|
|
|
|
|
38
|
|
|
|
|
|
|
In short, this syscall will return a data structure, but you probably will want to use only the name of each entry in the directory. |
39
|
|
|
|
|
|
|
|
40
|
|
|
|
|
|
|
How can this be useful? Here are some directions: |
41
|
|
|
|
|
|
|
|
42
|
|
|
|
|
|
|
=over |
43
|
|
|
|
|
|
|
|
44
|
|
|
|
|
|
|
=item 1. |
45
|
|
|
|
|
|
|
|
46
|
|
|
|
|
|
|
You want to remove all directory content. |
47
|
|
|
|
|
|
|
|
48
|
|
|
|
|
|
|
=item 2. |
49
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
|
You want to remove files from the directory with a pattern in their filename (using regular expressions, for example). |
51
|
|
|
|
|
|
|
|
52
|
|
|
|
|
|
|
=item 3. |
53
|
|
|
|
|
|
|
|
54
|
|
|
|
|
|
|
You want to select specific files by their filenames and then test something else (like atime). |
55
|
|
|
|
|
|
|
|
56
|
|
|
|
|
|
|
=back |
57
|
|
|
|
|
|
|
|
58
|
|
|
|
|
|
|
These are examples, but it should cover the vast majority of what you want to do. C syscall will be more effective because |
59
|
|
|
|
|
|
|
it will not call C of each of those files before returning the information to you. That means, you will have the opportunity to filter |
60
|
|
|
|
|
|
|
whatever you need and then call C if you really need. |
61
|
|
|
|
|
|
|
|
62
|
|
|
|
|
|
|
I came up at C after researching about "how to remove million of files". After a while I reached an C program example that uses C |
63
|
|
|
|
|
|
|
to print the filenames under the directory. By using it, I was able to cleanup directories with thousands (or even millions) of files in a couple of minutes, |
64
|
|
|
|
|
|
|
instead of many hours. |
65
|
|
|
|
|
|
|
|
66
|
|
|
|
|
|
|
This module is a Perl implementation of that. |
67
|
|
|
|
|
|
|
|
68
|
|
|
|
|
|
|
=head1 FUNCTIONS |
69
|
|
|
|
|
|
|
|
70
|
|
|
|
|
|
|
The sub C is exported by demand. |
71
|
|
|
|
|
|
|
|
72
|
|
|
|
|
|
|
=cut |
73
|
|
|
|
|
|
|
|
74
|
|
|
|
|
|
|
our @EXPORT_OK = qw(getdents); |
75
|
|
|
|
|
|
|
|
76
|
|
|
|
|
|
|
=head2 getdents |
77
|
|
|
|
|
|
|
|
78
|
|
|
|
|
|
|
Expects the full path to the directory as a parameter. |
79
|
|
|
|
|
|
|
|
80
|
|
|
|
|
|
|
Returns an array reference with all the fullpath to each of the file inside that directory. |
81
|
|
|
|
|
|
|
|
82
|
|
|
|
|
|
|
Meanwhile simple, you should be careful regarding memory restrictions. If you have too files, you program may try to allocate too much memory, with all the |
83
|
|
|
|
|
|
|
undesired effects. |
84
|
|
|
|
|
|
|
|
85
|
|
|
|
|
|
|
=cut |
86
|
|
|
|
|
|
|
|
87
|
|
|
|
|
|
|
sub getdents { |
88
|
0
|
|
|
0
|
|
|
my ( $dir, $output ) = @_; |
89
|
0
|
0
|
|
|
|
|
confess "directory $dir is not available" unless ( -d $dir ); |
90
|
0
|
|
|
|
|
|
sysopen( my $fd, $dir, O_RDONLY | O_DIRECTORY ); |
91
|
0
|
|
|
|
|
|
my @items; |
92
|
|
|
|
|
|
|
|
93
|
0
|
|
|
|
|
|
while (1) { |
94
|
0
|
|
|
|
|
|
my $buf = "\0" x BUF_SIZE; |
95
|
0
|
|
|
|
|
|
my $read = syscall( &SYS_getdents, fileno($fd), $buf, BUF_SIZE ); |
96
|
|
|
|
|
|
|
|
97
|
0
|
0
|
0
|
|
|
|
if ( ( $read == -1 ) and ( $! != 0 ) ) { |
98
|
0
|
|
|
|
|
|
confess "failed to syscall getdents: $!"; |
99
|
|
|
|
|
|
|
} |
100
|
|
|
|
|
|
|
|
101
|
0
|
0
|
|
|
|
|
last if ( $read == 0 ); |
102
|
|
|
|
|
|
|
|
103
|
0
|
|
|
|
|
|
while ( $read != 0 ) { |
104
|
0
|
|
|
|
|
|
my ( $ino, $off, $len, $name ) = unpack( "L!L!SZ*", $buf ); |
105
|
0
|
|
|
|
|
|
push( @items, ( $dir . '/' . $name ) ); |
106
|
0
|
|
|
|
|
|
substr( $buf, 0, $len ) = ''; |
107
|
0
|
|
|
|
|
|
$read -= $len; |
108
|
|
|
|
|
|
|
} |
109
|
|
|
|
|
|
|
|
110
|
|
|
|
|
|
|
} |
111
|
|
|
|
|
|
|
|
112
|
0
|
|
|
|
|
|
return @items; |
113
|
|
|
|
|
|
|
} |
114
|
|
|
|
|
|
|
|
115
|
|
|
|
|
|
|
=head1 INSTALL |
116
|
|
|
|
|
|
|
|
117
|
|
|
|
|
|
|
You should install this module as any Perl module, but before that be sure to execute L before trying to run any function from this module! |
118
|
|
|
|
|
|
|
|
119
|
|
|
|
|
|
|
In some system, you might need to use the system administrator account to run L or even run some manual steps to fix files locations. |
120
|
|
|
|
|
|
|
|
121
|
|
|
|
|
|
|
If you got errors like: |
122
|
|
|
|
|
|
|
|
123
|
|
|
|
|
|
|
Error: Can't locate bits/syscall.ph in @INC (did you run h2ph?) (@INC contains: /home/me/Projetos/Linux-NFS-BigDir/.build/MHr69O96uB/blib/lib /home/me/Projetos/Linux-NFS-BigDir/.build/MHr69O96uB/blib/arch /home/me/perl5/perlbrew/perls/perl-5.24.0/lib/site_perl/5.24.0/x86_64-linux /home/me/perl5/perlbrew/perls/perl-5.24.0/lib/site_perl/5.24.0 /home/me/perl5/perlbrew/perls/perl-5.24.0/lib/5.24.0/x86_64-linux /home/me/perl5/perlbrew/perls/perl-5.24.0/lib/5.24.0 .) at /home/me/perl5/perlbrew/perls/perl-5.24.0/lib/site_perl/5.24.0/sys/syscall.ph line 9. |
124
|
|
|
|
|
|
|
|
125
|
|
|
|
|
|
|
It might means that the expected header files are not in the expected standard location. For instance, on a Ubuntu system you might need to create additional links: |
126
|
|
|
|
|
|
|
|
127
|
|
|
|
|
|
|
ln -s /home/me/perl5/perlbrew/perls/perl-5.24.0/lib/site_perl/5.24.0/x86_64-linux/x86_64-linux-gnu/bits /home/me/perl5/perlbrew/perls/perl-5.24.0/lib/site_perl/5.24.0/bits |
128
|
|
|
|
|
|
|
|
129
|
|
|
|
|
|
|
You will have to troubleshoot this by looking at the C<$Config{'installsitearch'}> to see where are located your .ph files, then check the content of each .ph and compare with the real location of the C header files. |
130
|
|
|
|
|
|
|
|
131
|
|
|
|
|
|
|
Even though you might be using something like perlbrew (or compiling perl yourself), you will need to use the root account to fix this. |
132
|
|
|
|
|
|
|
|
133
|
|
|
|
|
|
|
=head1 SEE ALSO |
134
|
|
|
|
|
|
|
|
135
|
|
|
|
|
|
|
=over |
136
|
|
|
|
|
|
|
|
137
|
|
|
|
|
|
|
=item * |
138
|
|
|
|
|
|
|
|
139
|
|
|
|
|
|
|
L |
140
|
|
|
|
|
|
|
|
141
|
|
|
|
|
|
|
=item * |
142
|
|
|
|
|
|
|
|
143
|
|
|
|
|
|
|
L. |
144
|
|
|
|
|
|
|
|
145
|
|
|
|
|
|
|
=item * |
146
|
|
|
|
|
|
|
|
147
|
|
|
|
|
|
|
L |
148
|
|
|
|
|
|
|
|
149
|
|
|
|
|
|
|
=item * |
150
|
|
|
|
|
|
|
|
151
|
|
|
|
|
|
|
The manual page of C. |
152
|
|
|
|
|
|
|
|
153
|
|
|
|
|
|
|
=item * |
154
|
|
|
|
|
|
|
|
155
|
|
|
|
|
|
|
L. |
156
|
|
|
|
|
|
|
|
157
|
|
|
|
|
|
|
=back |
158
|
|
|
|
|
|
|
|
159
|
|
|
|
|
|
|
=head1 AUTHOR |
160
|
|
|
|
|
|
|
|
161
|
|
|
|
|
|
|
Alceu Rodrigues de Freitas Junior, Earfreitas@cpan.orgE |
162
|
|
|
|
|
|
|
|
163
|
|
|
|
|
|
|
=head1 COPYRIGHT AND LICENSE |
164
|
|
|
|
|
|
|
|
165
|
|
|
|
|
|
|
This software is copyright (c) 2016 of Alceu Rodrigues de Freitas Junior, Earfreitas@cpan.orgE |
166
|
|
|
|
|
|
|
|
167
|
|
|
|
|
|
|
This file is part of Linux-NFS-BigDir distribution. |
168
|
|
|
|
|
|
|
|
169
|
|
|
|
|
|
|
Linux-NFS-BigDir is free software: you can redistribute it and/or modify |
170
|
|
|
|
|
|
|
it under the terms of the GNU General Public License as published by |
171
|
|
|
|
|
|
|
the Free Software Foundation, either version 3 of the License, or |
172
|
|
|
|
|
|
|
(at your option) any later version. |
173
|
|
|
|
|
|
|
|
174
|
|
|
|
|
|
|
Linux-NFS-BigDir is distributed in the hope that it will be useful, |
175
|
|
|
|
|
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of |
176
|
|
|
|
|
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
177
|
|
|
|
|
|
|
GNU General Public License for more details. |
178
|
|
|
|
|
|
|
|
179
|
|
|
|
|
|
|
You should have received a copy of the GNU General Public License |
180
|
|
|
|
|
|
|
along with Linux-NFS-BigDir. If not, see . |
181
|
|
|
|
|
|
|
|
182
|
|
|
|
|
|
|
=cut |
183
|
|
|
|
|
|
|
|
184
|
|
|
|
|
|
|
1; |