| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # SimpleVideoPlugin.pm -- Plugin for multimedia with some simple video
|
|---|
| 4 | # processing
|
|---|
| 5 | #
|
|---|
| 6 | # A component of the Greenstone digital library software from the New
|
|---|
| 7 | # Zealand Digital Library Project at the University of Waikato, New
|
|---|
| 8 | # Zealand.
|
|---|
| 9 | #
|
|---|
| 10 | # Copyright (C) 2012 New Zealand Digital Library Project
|
|---|
| 11 | #
|
|---|
| 12 | # This program is free software; you can redistribute it and/or modify
|
|---|
| 13 | # it under the terms of the GNU General Public License as published by
|
|---|
| 14 | # the Free Software Foundation; either version 2 of the License, or
|
|---|
| 15 | # (at your option) any later version.
|
|---|
| 16 | #
|
|---|
| 17 | # This program is distributed in the hope that it will be useful, but
|
|---|
| 18 | # WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 19 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|---|
| 20 | # General Public License for more details.
|
|---|
| 21 | #
|
|---|
| 22 | # You should have received a copy of the GNU General Public License
|
|---|
| 23 | # along with this program; if not, write to the Free Software
|
|---|
| 24 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
|---|
| 25 | #
|
|---|
| 26 | ###########################################################################
|
|---|
| 27 |
|
|---|
| 28 | package SimpleVideoPlugin;
|
|---|
| 29 |
|
|---|
| 30 | use File::Temp qw/ tempdir /;
|
|---|
| 31 |
|
|---|
| 32 | use BasePlugin;
|
|---|
| 33 | use MetadataRead;
|
|---|
| 34 | use util;
|
|---|
| 35 |
|
|---|
| 36 | use strict;
|
|---|
| 37 | use warnings;
|
|---|
| 38 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
|---|
| 39 |
|
|---|
| 40 | sub BEGIN
|
|---|
| 41 | {
|
|---|
| 42 | @SimpleVideoPlugin::ISA = ('MetadataRead', 'BasePlugin');
|
|---|
| 43 | }
|
|---|
| 44 |
|
|---|
| 45 | my $arguments = [ { 'name' => "process_exp",
|
|---|
| 46 | 'desc' => "{BasePlugin.process_exp}",
|
|---|
| 47 | 'type' => "regexp",
|
|---|
| 48 | 'reqd' => "no",
|
|---|
| 49 | 'deft' => &get_default_process_exp() },
|
|---|
| 50 | { 'name' => "streamingHQsize",
|
|---|
| 51 | 'desc' => "{VideoPlugin.streamingsize}",
|
|---|
| 52 | 'type' => "int",
|
|---|
| 53 | 'deft' => "720",
|
|---|
| 54 | 'reqd' => "no" },
|
|---|
| 55 | { 'name' => "streamingHQVideoBitrate",
|
|---|
| 56 | 'desc' => "{VideoPlugin.streamingbitrate}",
|
|---|
| 57 | 'type' => "int",
|
|---|
| 58 | 'deft' => "496",
|
|---|
| 59 | 'reqd' => "no" },
|
|---|
| 60 | { 'name' => "streamingHQAudioBitrate",
|
|---|
| 61 | 'desc' => "{VideoPlugin.streamingbitrate}",
|
|---|
| 62 | 'type' => "int",
|
|---|
| 63 | 'deft' => "80",
|
|---|
| 64 | 'reqd' => "no" },
|
|---|
| 65 | { 'name' => "videoDeinterlacingFilter",
|
|---|
| 66 | 'desc' => "Activate a deinterlacing filter to increase the quality of TV footage",
|
|---|
| 67 | 'type' => "enum",
|
|---|
| 68 | 'list' => [{'name' => "true", 'desc' => "{common.true}"},
|
|---|
| 69 | {'name' => "false", 'desc' => "{common.false}"}],
|
|---|
| 70 | 'deft' => "false",
|
|---|
| 71 | 'reqd' => "no" },
|
|---|
| 72 | { 'name' => "isParallel",
|
|---|
| 73 | 'desc' => "Will the import use parallel processing? (maybe this should be set by parallel-import.pl somehow)",
|
|---|
| 74 | 'type' => "enum",
|
|---|
| 75 | 'list' => [{'name' => "true", 'desc' => "{common.true}"},
|
|---|
| 76 | {'name' => "false", 'desc' => "{common.false}"}],
|
|---|
| 77 | 'deft' => "true",
|
|---|
| 78 | 'reqd' => "no" },
|
|---|
| 79 | { 'name' => "isCluster",
|
|---|
| 80 | 'desc' => "Will the import be run on a cluster (multiple computers) or not (single computer - possibly multiple processors)",
|
|---|
| 81 | 'type' => "enum",
|
|---|
| 82 | 'list' => [{'name' => "true", 'desc' => "{common.true}"},
|
|---|
| 83 | {'name' => "false", 'desc' => "{common.false}"}],
|
|---|
| 84 | 'deft' => "false",
|
|---|
| 85 | 'reqd' => "no" },
|
|---|
| 86 | { 'name' => "separateIO",
|
|---|
| 87 | 'desc' => "copy and process the file locally (good for segregating IO cost)",
|
|---|
| 88 | 'type' => "enum",
|
|---|
| 89 | 'list' => [{'name' => "true", 'desc' => "{common.true}"},
|
|---|
| 90 | {'name' => "false", 'desc' => "{common.false}"}],
|
|---|
| 91 | 'deft' => "false",
|
|---|
| 92 | 'reqd' => "no" },
|
|---|
| 93 | { 'name' => "fixedCore",
|
|---|
| 94 | 'desc' => "Restrict the execution of Handbrake to a single core (0 = no restriction, > 0 use value-1'th core)",
|
|---|
| 95 | 'type' => "int",
|
|---|
| 96 | 'deft' => "0",
|
|---|
| 97 | 'reqd' => "no" },
|
|---|
| 98 | ];
|
|---|
| 99 |
|
|---|
| 100 | my $options = { 'name' => "BasicVideoPlugin",
|
|---|
| 101 | 'desc' => "",
|
|---|
| 102 | 'abstract' => "no",
|
|---|
| 103 | 'inherits' => "yes",
|
|---|
| 104 | 'args' => $arguments };
|
|---|
| 105 |
|
|---|
| 106 | sub new
|
|---|
| 107 | {
|
|---|
| 108 | my ($class) = shift (@_);
|
|---|
| 109 | my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
|
|---|
| 110 | push(@$pluginlist, $class);
|
|---|
| 111 |
|
|---|
| 112 | push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
|---|
| 113 | push(@{$hashArgOptLists->{"OptList"}},$options);
|
|---|
| 114 | my $self = new BasePlugin($pluginlist, $inputargs, $hashArgOptLists);
|
|---|
| 115 | return bless $self, $class;
|
|---|
| 116 | }
|
|---|
| 117 |
|
|---|
| 118 | sub get_default_process_exp
|
|---|
| 119 | {
|
|---|
| 120 | return '(?i)\.ts$';
|
|---|
| 121 | }
|
|---|
| 122 |
|
|---|
| 123 | sub get_oid_hash_type
|
|---|
| 124 | {
|
|---|
| 125 | my $self = shift (@_);
|
|---|
| 126 | return "hash_on_ga_xml";
|
|---|
| 127 | }
|
|---|
| 128 |
|
|---|
| 129 | sub process
|
|---|
| 130 | {
|
|---|
| 131 | my $self = shift (@_);
|
|---|
| 132 | my ($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
|
|---|
| 133 |
|
|---|
| 134 | print STDERR "[A:" . time() . "] SimpleVideoPlugin processing: " . $file . "\n";
|
|---|
| 135 |
|
|---|
| 136 | # - I have to add some text (yay, back to needing dummy text) otherwise the
|
|---|
| 137 | # DocumentText formatting is ignored (?!?)
|
|---|
| 138 | my $topsection = $doc_obj->get_top_section();
|
|---|
| 139 | $doc_obj->add_utf8_text($topsection, "This is dummy text");
|
|---|
| 140 |
|
|---|
| 141 |
|
|---|
| 142 | $file =~ /[\/]?([^\/]+)\.(?:ts)$/;
|
|---|
| 143 | my $filename = $1;
|
|---|
| 144 |
|
|---|
| 145 | # Optional date metadata (available on raw ReplayMe recordings)
|
|---|
| 146 | if ($filename =~ /(\d\d\d\d)-(\d\d)-(\d\d)/)
|
|---|
| 147 | {
|
|---|
| 148 | my $date = $1 . $2 . $3;
|
|---|
| 149 | $filename =~ s/[^a-z0-9]+/_/ig;
|
|---|
| 150 | $filename =~ s/^_+|_+$//g;
|
|---|
| 151 | $doc_obj->add_utf8_metadata($topsection,"Date",$date);
|
|---|
| 152 | }
|
|---|
| 153 |
|
|---|
| 154 | # Special Case: HDFS *only* supported by separateIO flag (you need to move
|
|---|
| 155 | # the file out of HDFS to local filespace to allow MediaInfo and Handbrake
|
|---|
| 156 | # to be run on it.
|
|---|
| 157 | my $separate_io = $self->{'separateIO'};
|
|---|
| 158 | if (&util::isHDFS($file))
|
|---|
| 159 | {
|
|---|
| 160 | $separate_io = 'true';
|
|---|
| 161 | }
|
|---|
| 162 |
|
|---|
| 163 | my $process_dir = $ENV{'GSDLCOLLECTDIR'};
|
|---|
| 164 | # If we are in a cluster, then we don't want to be writing all the logs
|
|---|
| 165 | # etc to the shared file system. Instead, we write to the tmp drive
|
|---|
| 166 | if ($separate_io eq 'true')
|
|---|
| 167 | {
|
|---|
| 168 | $process_dir = &util::filename_cat('/tmp', 'gsimport-' . $filename);
|
|---|
| 169 | if (!&util::dir_exists($process_dir))
|
|---|
| 170 | {
|
|---|
| 171 | mkdir($process_dir, 0775);
|
|---|
| 172 | }
|
|---|
| 173 | }
|
|---|
| 174 | my $logs_dir = &util::filename_cat($process_dir, "logs");
|
|---|
| 175 | if (!&util::dir_exists($logs_dir))
|
|---|
| 176 | {
|
|---|
| 177 | mkdir($logs_dir, 0775);
|
|---|
| 178 | }
|
|---|
| 179 | my $convert_log_path = &util::filename_cat($logs_dir, 'convert-' . $filename . '.log');
|
|---|
| 180 | my $pass_log_path = &util::filename_cat($logs_dir, 'convert-' . $filename . '-pass');
|
|---|
| 181 | my $tmp_dir = &util::filename_cat($process_dir, "cached");
|
|---|
| 182 | if (!&util::dir_exists($tmp_dir))
|
|---|
| 183 | {
|
|---|
| 184 | mkdir($tmp_dir, 0775);
|
|---|
| 185 | }
|
|---|
| 186 | $tmp_dir = &util::filename_cat($tmp_dir, $filename);
|
|---|
| 187 | if (!&util::dir_exists($tmp_dir))
|
|---|
| 188 | {
|
|---|
| 189 | mkdir($tmp_dir, 0775);
|
|---|
| 190 | }
|
|---|
| 191 |
|
|---|
| 192 | # If we are separating IO, then we also start by copying the file to
|
|---|
| 193 | # the process directory (local tmp) as well
|
|---|
| 194 | my $ivideo_path = &util::filename_cat($base_dir, $file);
|
|---|
| 195 | if ($separate_io eq 'true')
|
|---|
| 196 | {
|
|---|
| 197 | print STDERR "[B1:" . time() . "] Creating local copy of file\n";
|
|---|
| 198 | my $local_ivideo_path = &util::filename_cat($process_dir, $filename . ".ts");
|
|---|
| 199 | &util::cp($ivideo_path, $local_ivideo_path);
|
|---|
| 200 | $ivideo_path = $local_ivideo_path;
|
|---|
| 201 | print STDERR "[B2:" . time() . "] Complete\n";
|
|---|
| 202 | }
|
|---|
| 203 |
|
|---|
| 204 | # 1. Use MediaInfo to extract important metadata
|
|---|
| 205 | print STDERR "[C1:" . time() . "] Extracting metadata\n";
|
|---|
| 206 | print " - Extracting metadata using MediaInfo\n";
|
|---|
| 207 | my $mi_metadata = $self->getMetadata($ivideo_path);
|
|---|
| 208 | $doc_obj->add_utf8_metadata($topsection,"Format", 'multimedia (' . $mi_metadata->{'General'}->{'Format'} . ')');
|
|---|
| 209 | $doc_obj->set_metadata_element($topsection,"FileSize",$mi_metadata->{'General'}->{'File_size'});
|
|---|
| 210 | $doc_obj->add_utf8_metadata($topsection,"Duration",$mi_metadata->{'General'}->{'Duration'});
|
|---|
| 211 | if (defined $mi_metadata->{'Video'}->{'Format_Info'} && defined $mi_metadata->{'Video'}->{'Format'})
|
|---|
| 212 | {
|
|---|
| 213 | $doc_obj->add_utf8_metadata($topsection,"VideoFormat",$mi_metadata->{'Video'}->{'Format_Info'} . ' (' . $mi_metadata->{'Video'}->{'Format'} . ')');
|
|---|
| 214 | }
|
|---|
| 215 | if (defined $mi_metadata->{'Audio'}->{'Format_Info'} && defined $mi_metadata->{'Audio'}->{'Format'})
|
|---|
| 216 | {
|
|---|
| 217 | $doc_obj->add_utf8_metadata($topsection,"AudioFormat",$mi_metadata->{'Audio'}->{'Format_Info'} . ' (' . $mi_metadata->{'Audio'}->{'Format'} . ')');
|
|---|
| 218 | }
|
|---|
| 219 | $doc_obj->add_utf8_metadata($topsection,"Width",$mi_metadata->{'Video'}->{'Width'});
|
|---|
| 220 | $doc_obj->add_utf8_metadata($topsection,"Height",$mi_metadata->{'Video'}->{'Height'});
|
|---|
| 221 | print STDERR "[C2:" . time() . "] Complete\n";
|
|---|
| 222 |
|
|---|
| 223 | # 2. Convert into FLV, reprocess to make seekable, and associate
|
|---|
| 224 | # - generate a path for our temporary converted video file
|
|---|
| 225 | print STDERR "[D1:" . time() . "] Converting video to streamble format\n";
|
|---|
| 226 | my $ovideo_path = &util::filename_cat($tmp_dir, 'gsv.mp4');
|
|---|
| 227 | if (&util::file_exists($ovideo_path))
|
|---|
| 228 | {
|
|---|
| 229 | print " - Found existing converted video in cache\n";
|
|---|
| 230 | }
|
|---|
| 231 | else
|
|---|
| 232 | {
|
|---|
| 233 | # - first conversion pass
|
|---|
| 234 | print " - Convert using Handbrake\n";
|
|---|
| 235 | my $streaming_HQ_size = $self->{'streamingHQsize'};
|
|---|
| 236 | my $streaming_HQ_VideoBitrate = $self->{'streamingHQVideoBitrate'};
|
|---|
| 237 | my $streaming_HQ_AudioBitrate = $self->{'streamingHQAudioBitrate'};
|
|---|
| 238 | my $deinterlace = $self->{'videoDeinterlacingFilter'};
|
|---|
| 239 | my $video_processing_parameters;
|
|---|
| 240 | if (!$streaming_HQ_size || $streaming_HQ_size eq "fullsize")
|
|---|
| 241 | {
|
|---|
| 242 | $video_processing_parameters = "--strict-anamorphic";
|
|---|
| 243 | }
|
|---|
| 244 | else
|
|---|
| 245 | {
|
|---|
| 246 | $video_processing_parameters = "-w $streaming_HQ_size --loose-anamorphic";
|
|---|
| 247 | }
|
|---|
| 248 | if ($deinterlace eq "true")
|
|---|
| 249 | {
|
|---|
| 250 | $video_processing_parameters .= " --decomb";
|
|---|
| 251 | }
|
|---|
| 252 | # Default MenCoder options for x264
|
|---|
| 253 | my $mencoder_options = 'ref=2:bframes=2:subq=6:mixed-refs=0:weightb=0:8x8dct=0:trellis=0';
|
|---|
| 254 | my $is_cluster = $self->{'isCluster'};
|
|---|
| 255 | my $is_parallel = $self->{'isParallel'};
|
|---|
| 256 | # If we are parallel processing on a single (presumably) multicore computer
|
|---|
| 257 | # then we need to limit the number of threads (and hence CPUs) HandBrake
|
|---|
| 258 | # will utilize in order to emulate true parallel processing (otherwise the
|
|---|
| 259 | # first thread to get to HandBrake conversion will take up most the CPUs
|
|---|
| 260 | # causing all other threads to wait anyway). It will interesting to test
|
|---|
| 261 | # whether parallel processing or serial processing (with HandBrake parallel
|
|---|
| 262 | # processing) is faster. *update* threads=1 *only* controls the encoding and
|
|---|
| 263 | # several other parts of Handbrake can run parallel (demuxing etc). I've
|
|---|
| 264 | # had to include a 'taskset' command to truely make Handbrake serial
|
|---|
| 265 | if ($is_parallel eq 'true' && $is_cluster eq 'false')
|
|---|
| 266 | {
|
|---|
| 267 | $mencoder_options .= ':threads=1';
|
|---|
| 268 | }
|
|---|
| 269 | # Banish HandbrakeCLI to the (fixedCore-1)'th CPU if necessary
|
|---|
| 270 | my $cmd = '';
|
|---|
| 271 | if (defined $self->{'fixedCore'} && $self->{'fixedCore'} > 0)
|
|---|
| 272 | {
|
|---|
| 273 | $cmd .= 'taskset -c ' . ($self->{'fixedCore'} - 1) . ' ';
|
|---|
| 274 | }
|
|---|
| 275 | $cmd .= 'HandBrakeCLI -i "' . $ivideo_path . '" -t 1 -c 1 -f mp4 -O -o "' . $ovideo_path . '" ' . $video_processing_parameters . ' -e x264 -b ' . $streaming_HQ_VideoBitrate . ' -a 1 -E faac -6 dpl2 -R Auto -B ' . $streaming_HQ_AudioBitrate . ' -D 0.0 -x ' . $mencoder_options . ' > "' . $convert_log_path . '" 2>&1';
|
|---|
| 276 | print "[DEBUG: " . $cmd . "]\n";
|
|---|
| 277 | `$cmd`;
|
|---|
| 278 | }
|
|---|
| 279 | if (!&util::file_exists($ovideo_path))
|
|---|
| 280 | {
|
|---|
| 281 | die("Fatal Error! Failed to convert video: " . $ovideo_path . "\nReason:" . $! . "\n");
|
|---|
| 282 | }
|
|---|
| 283 | print STDERR "[D2:" . time() . "] Complete\n";
|
|---|
| 284 |
|
|---|
| 285 | # 3. Extract keyframes using hive
|
|---|
| 286 | print STDERR "[E1:" . time() . "] Extract keyframes\n";
|
|---|
| 287 | my $oshots_path = &util::filename_cat($tmp_dir, 'shots.xml');
|
|---|
| 288 | if (&util::file_exists($oshots_path))
|
|---|
| 289 | {
|
|---|
| 290 | print " - Found existing keyframe images in cache\n";
|
|---|
| 291 | }
|
|---|
| 292 | else
|
|---|
| 293 | {
|
|---|
| 294 | print " - Generating keyframe images using Hive2\n";
|
|---|
| 295 | my $cmd = 'hive2_ffmpegsvn -o "' . $oshots_path . '" -k "' . $tmp_dir . '" "' . $ovideo_path . '" >> "' . $convert_log_path . '" 2>&1';
|
|---|
| 296 | ###print "[cmd: " . $cmd . "]\n";
|
|---|
| 297 | `$cmd`;
|
|---|
| 298 | }
|
|---|
| 299 | if (!&util::file_exists($oshots_path))
|
|---|
| 300 | {
|
|---|
| 301 | die("Fatal Error! Failed to extract keyframe images: " . $oshots_path . "\nReason:" . $! . "\n");
|
|---|
| 302 | }
|
|---|
| 303 | print STDERR "[E2:" . time() . "] Complete\n";
|
|---|
| 304 |
|
|---|
| 305 |
|
|---|
| 306 | # 4. Associate files (copies back to shared space if IO separated)
|
|---|
| 307 | print STDERR "[F1:" . time() . "] Associate derived files to doc_obj\n";
|
|---|
| 308 | # - associate streamable video
|
|---|
| 309 | $doc_obj->associate_file($ovideo_path,'gsv.mp4','video/mp4',$topsection);
|
|---|
| 310 | # - associate all of the JPGs found in the temp directory
|
|---|
| 311 | opendir(my $dh, $tmp_dir);
|
|---|
| 312 | my @shots = readdir($dh);
|
|---|
| 313 | closedir($dh);
|
|---|
| 314 | my $thumbnail = 0;
|
|---|
| 315 | foreach my $shot (sort @shots)
|
|---|
| 316 | {
|
|---|
| 317 | my $shot_path = &util::filename_cat($tmp_dir, $shot);
|
|---|
| 318 | if ($shot =~ /.jpg$/)
|
|---|
| 319 | {
|
|---|
| 320 | if (!$thumbnail)
|
|---|
| 321 | {
|
|---|
| 322 | $doc_obj->add_utf8_metadata($topsection,"Thumbnail",$shot);
|
|---|
| 323 | $thumbnail = 1;
|
|---|
| 324 | }
|
|---|
| 325 | $doc_obj->add_utf8_metadata($topsection,"Keyframe",$shot);
|
|---|
| 326 | $doc_obj->associate_file($shot_path,$shot,"image/jpeg",$topsection);
|
|---|
| 327 | }
|
|---|
| 328 | }
|
|---|
| 329 | print STDERR "[F2:" . time() . "] Complete\n";
|
|---|
| 330 |
|
|---|
| 331 | # 5. Done! Cleanup.
|
|---|
| 332 | print STDERR "[G:" . time() . "] SimpleVideoPlugin: Complete!\n";
|
|---|
| 333 | return 1;
|
|---|
| 334 | }
|
|---|
| 335 |
|
|---|
| 336 | sub getMetadata
|
|---|
| 337 | {
|
|---|
| 338 | my ($self, $ivideo_path) = @_;
|
|---|
| 339 | my $cmd = 'mediainfo --Output=XML "' . $ivideo_path . '" 2>&1';
|
|---|
| 340 | ###print "Cmd: " . $cmd . "\n";
|
|---|
| 341 | my $metadata_xml = `$cmd`;
|
|---|
| 342 | my @lines = split(/\r?\n/, $metadata_xml);
|
|---|
| 343 | my $metadata = {'Unknown'=>{}};
|
|---|
| 344 | my $metadata_type = 'Unknown';
|
|---|
| 345 | foreach my $line (@lines)
|
|---|
| 346 | {
|
|---|
| 347 | if ($line =~ /<track type="(.+)">/)
|
|---|
| 348 | {
|
|---|
| 349 | $metadata_type = $1;
|
|---|
| 350 | if (!defined $metadata->{$metadata_type})
|
|---|
| 351 | {
|
|---|
| 352 | $metadata->{$metadata_type} = {};
|
|---|
| 353 | }
|
|---|
| 354 | }
|
|---|
| 355 | elsif ($line =~ /<([^>]+)>(.+)<\/[^>]+>/)
|
|---|
| 356 | {
|
|---|
| 357 | my $field = $1;
|
|---|
| 358 | my $value = $2;
|
|---|
| 359 | $metadata->{$metadata_type}->{$field} = $value;
|
|---|
| 360 | }
|
|---|
| 361 | }
|
|---|
| 362 | return $metadata;
|
|---|
| 363 | }
|
|---|
| 364 |
|
|---|
| 365 | 1;
|
|---|
| 366 |
|
|---|
| 367 |
|
|---|
| 368 |
|
|---|
| 369 |
|
|---|
| 370 |
|
|---|
| 371 |
|
|---|
| 372 |
|
|---|
| 373 |
|
|---|
| 374 |
|
|---|
| 375 |
|
|---|
| 376 |
|
|---|
