NAME¶
MCE::Grep - Parallel grep model similar to the native grep function
VERSION¶
This document describes MCE::Grep version 1.517
SYNOPSIS¶
## Exports mce_grep, mce_grep_f, and mce_grep_s
use MCE::Grep;
## Array or array_ref
my @a = mce_grep { $_ % 5 == 0 } 1..10000;
my @b = mce_grep { $_ % 5 == 0 } [ 1..10000 ];
## File_path, glob_ref, or scalar_ref
my @c = mce_grep_f { /phrase/ } "/path/to/file";
my @d = mce_grep_f { /phrase/ } $file_handle;
my @e = mce_grep_f { /phrase/ } \$scalar;
## Sequence of numbers (begin, end [, step, format])
my @f = mce_grep_s { %_ * 3 == 0 } 1, 10000, 5;
my @g = mce_grep_s { %_ * 3 == 0 } [ 1, 10000, 5 ];
my @h = mce_grep_s { %_ * 3 == 0 } {
begin => 1, end => 10000, step => 5, format => undef
};
DESCRIPTION¶
This module provides a parallel grep implementation via Many-core Engine. MCE
incurs a small overhead due to passing of data. Therefore, a fast code block
will likely run faster using the native grep function in Perl. The overhead
quickly diminishes as the complexity of the code block increases.
my @m1 = grep { $_ % 5 == 0 } 1..1000000; ## 0.137 secs
my @m2 = mce_grep { $_ % 5 == 0 } 1..1000000; ## 0.295 secs
Chunking, enabled by default, greatly reduces the overhead behind the scene. The
time for mce_grep below also includes the time for data exchanges between the
manager and worker processes. More parallelization will be seen when the code
block requires additional CPU time code-wise.
my @m1 = grep { /[2357][1468][9]/ } 1..1000000; ## 0.653 secs
my @m2 = mce_grep { /[2357][1468][9]/ } 1..1000000; ## 0.347 secs
The mce_grep_s function will provide better times, useful when the input data is
simply a range of numbers. Workers generate sequences mathematically among
themselves without any interaction from the manager process. Two arguments are
required for mce_grep_s (begin, end). Step defaults to 1 if begin is smaller
than end, otherwise -1.
my @m3 = mce_grep_s { /[2357][1468][9]/ } 1, 1000000; ## 0.271 secs
Although this document is about MCE::Grep, the MCE::Stream module can write
results immediately without waiting for all chunks to complete. This is made
possible by passing the reference of the array (in this case @m4 and @m5).
use MCE::Stream default_mode => 'grep';
my @m4; mce_stream \@m4, sub { /[2357][1468][9]/ }, 1..1000000;
## Completed in 0.304 secs. That is amazing considering the
## overhead for passing data between the manager and worker.
my @m5; mce_stream_s \@m5, sub { /[2357][1468][9]/ }, 1, 1000000;
## Completed in 0.227 secs. Like with mce_grep_s, specifying a
## sequence specification turns out to be faster due to lesser
## overhead for the manager process.
A good use-case for MCE::Grep is for searching through a large log file much
like one might do using the native grep function. Lets assume the file
contains a hundred thousand records separated by a string "::\n\n"
between each record. The imaginary pattern used also returns less than 50
records.
The native implementation is what one might do actually. What's not clearly
visible here is the initial memory consumption, due to Perl reading the entire
content into memory, prior to grep actually starting. A 300 MB file will
consume roughly 640 MB. The time to run is 1.217 seconds for the file residing
in the OS-level file-system cache.
$/ = "::\n\n";
open my $LOG, "<", "/path/to/log/file";
my @match = grep { $_ =~ /pattern/ } <$LOG>;
close $LOG;
The memory utilization is much better with MCE; 8 workers * 23 MB = 184 MB. MCE
caps at some point, therefore allowing one to process a file much larger than
available memory. The time to run is 0.416 seconds (2.93x faster) which
includes the overhead for chunking and serializing data back to the main
process as if processing serially.
use MCE::Grep;
MCE::Grep::init { RS => "::\n\n" };
my @match = mce_grep_f { $_ =~ /pattern/ } "/path/to/file";
It gets even better for counting though. The native grep function takes 1.136
seconds to run whereas MCE takes just 0.155 seconds (7.33x faster).
## Native Grep
my $count = grep { $_ =~ /pattern/ } <$LOG>;
## MCE Grep
my $count = mce_grep_f { $_ =~ /pattern/ } "/path/to/file";
OVERRIDING DEFAULTS¶
The following list 5 options which may be overridden when loading the module.
use Sereal qw(encode_sereal decode_sereal);
use MCE::Grep
max_workers => 4, ## Default 'auto'
chunk_size => 100, ## Default 'auto'
tmp_dir => "/path/to/app/tmp", ## $MCE::Signal::tmp_dir
freeze => \&encode_sereal, ## \&Storable::freeze
thaw => \&decode_sereal ## \&Storable::thaw
;
There is a simpler way to enable Sereal with MCE 1.5. The following will attempt
to use Sereal if available, otherwise will default back to using Storable for
serialization.
use MCE::Grep Sereal => 1;
## Serialization is through Sereal if available.
my @m2 = mce_grep { $_ % 5 == 0 } 1..10000;
CUSTOMIZING MCE¶
- init
- The init function accepts a hash of MCE options. The gather option, if
specified, will be set to undef due to being used internally by the
module.
use MCE::Grep;
MCE::Grep::init {
chunk_size => 1, max_workers => 4,
user_begin => sub {
print "## ", MCE->wid, " started\n";
},
user_end => sub {
print "## ", MCE->wid, " completed\n";
}
};
my @a = mce_grep { $_ % 5 == 0 } 1..100;
print "\n", "@a", "\n";
-- Output
## 2 started
## 3 started
## 1 started
## 4 started
## 3 completed
## 4 completed
## 1 completed
## 2 completed
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
API DOCUMENTATION¶
- mce_grep { code } list
- Input data can be defined using a list or passing a reference to an array.
my @a = mce_grep { /[2357]/ } 1..1000;
my @b = mce_grep { /[2357]/ } [ 1..1000 ];
- mce_grep_f { code } file
- The fastest of these is the /path/to/file. Workers communicate the next
offset position among themselves without any interaction from the manager
process.
my @c = mce_grep_f { /phrase/ } "/path/to/file";
my @d = mce_grep_f { /phrase/ } $file_handle;
my @e = mce_grep_f { /phrase/ } \$scalar;
- mce_grep_s { code } sequence
- Sequence can be defined as a list, an array reference, or a hash
reference. The functions require both begin and end values to run. Step
and format are optional. The format is passed to sprintf (% may be omitted
below).
my ($beg, $end, $step, $fmt) = (10, 20, 0.1, "%4.1f");
my @f = mce_grep_s { /[1234]\.[5678]/ } $beg, $end, $step, $fmt;
my @g = mce_grep_s { /[1234]\.[5678]/ } [ $beg, $end, $step, $fmt ];
my @h = mce_grep_s { /[1234]\.[5678]/ } {
begin => $beg, end => $end, step => $step, format => $fmt
};
- mce_grep { code } iterator
- An iterator reference can by specified for input data. Iterators are
described under "SYNTAX for INPUT_DATA" at MCE::Core.
my @a = mce_grep { $_ % 3 == 0 } make_iterator(10, 30, 2);
MANUAL SHUTDOWN¶
- finish
- MCE workers remain persistent as much as possible after running. Shutdown
occurs when the script exits. One can manually shutdown MCE by simply
calling finish after running. This resets the MCE instance.
use MCE::Grep;
MCE::Grep::init {
chunk_size => 20, max_workers => 'auto'
};
my @a = mce_grep { ... } 1..100;
MCE::Grep::finish;
INDEX¶
MCE
AUTHOR¶
Mario E. Roy, <marioeroy AT gmail DOT com>
LICENSE¶
This program is free software; you can redistribute it and/or modify it under
the terms of either: the GNU General Public License as published by the Free
Software Foundation; or the Artistic License.
See <
http://dev.perl.org/licenses/> for more information.