| 1 | #!/usr/bin/perl -w |
|---|
| 2 | |
|---|
| 3 | my $myprogname = 'dog'; |
|---|
| 4 | my $progname = 'shepherd'; |
|---|
| 5 | my $version = '0.3.1'; |
|---|
| 6 | |
|---|
| 7 | # 'dog' |
|---|
| 8 | # "Shepherd" |
|---|
| 9 | # A wrapper for various Aussie TV guide data grabbers |
|---|
| 10 | # |
|---|
| 11 | # Use --help for command-line options. |
|---|
| 12 | # |
|---|
| 13 | # Shepherd is an attempt to reconcile many different tv_grab_au scripts and |
|---|
| 14 | # make one cohesive reliable data set. It works by calling a series of |
|---|
| 15 | # scripts that grab data from a large variety of sources, and then |
|---|
| 16 | # analysing the resulting XML data sets and determining which of the many |
|---|
| 17 | # is the most reliable. |
|---|
| 18 | |
|---|
| 19 | # Shepherd runs in 4 passes: |
|---|
| 20 | # pass 1: (app/shepherd) Checks that all components are up-to-date, auto- |
|---|
| 21 | # updates if not. |
|---|
| 22 | # Passes control onto shepherd |
|---|
| 23 | # pass 2: (engine/dog) calls grabbers to fill in missing data |
|---|
| 24 | # pass 3: (engine/dog) calls reconciler to reconcile overlapping data |
|---|
| 25 | # and normalize programme titles to our preferred title |
|---|
| 26 | # pass 4: (engine/dog) calls postprocessors to postprocess data |
|---|
| 27 | # (e.g. flag HDTV programmes, augment with IMDb etc.) |
|---|
| 28 | |
|---|
| 29 | # Changelog: |
|---|
| 30 | # 0.3.1 : split 'dog' into its own engine |
|---|
| 31 | |
|---|
| 32 | BEGIN { *CORE::GLOBAL::die = \&my_die; } |
|---|
| 33 | |
|---|
| 34 | use strict; |
|---|
| 35 | no strict 'refs'; |
|---|
| 36 | |
|---|
| 37 | use LWP::UserAgent; |
|---|
| 38 | use Getopt::Long; |
|---|
| 39 | use Data::Dumper; |
|---|
| 40 | use XMLTV; |
|---|
| 41 | use POSIX qw(strftime mktime); |
|---|
| 42 | use Date::Manip; |
|---|
| 43 | use Algorithm::Diff; |
|---|
| 44 | use List::Compare; |
|---|
| 45 | |
|---|
| 46 | # --------------------------------------------------------------------------- |
|---|
| 47 | # --- Global Variables |
|---|
| 48 | # --------------------------------------------------------------------------- |
|---|
| 49 | |
|---|
| 50 | my @options = @ARGV; |
|---|
| 51 | |
|---|
| 52 | # By default, Shepherd runs from ~/.shepherd/. If it's not run as a user, |
|---|
| 53 | # it will try /opt/shepherd/ instead. |
|---|
| 54 | my $CWD = ($ENV{HOME} ? $ENV{HOME} . "/." : "/opt/") . $progname; |
|---|
| 55 | -d $CWD or mkdir $CWD or die "Cannot create directory $CWD: $!"; |
|---|
| 56 | chdir($CWD); |
|---|
| 57 | |
|---|
| 58 | #### analyzer settings #### |
|---|
| 59 | # the following thresholds are used to control whether we keep calling grabbers or |
|---|
| 60 | # not. |
|---|
| 61 | |
|---|
| 62 | my %policy; |
|---|
| 63 | $policy{timeslot_size} = (5 * 60); # 5 minute slots |
|---|
| 64 | |
|---|
| 65 | # PEAK timeslots - |
|---|
| 66 | # between 4.30pm and 11.30pm every day, only allow a maximum of |
|---|
| 67 | # 15 minutes "programming data" missing |
|---|
| 68 | # if there is more than this, we will continue asking grabbers for more |
|---|
| 69 | # programming on this channel |
|---|
| 70 | $policy{peak_max_missing} = 15*60; # up to 15 mins max allowed missing |
|---|
| 71 | $policy{peak_start} = (16*(60*60))+(30*60); # 4.30pm |
|---|
| 72 | $policy{peak_stop} = (23*(60*60))+(30*60); # 11.30pm |
|---|
| 73 | |
|---|
| 74 | # NON-PEAK timeslots - |
|---|
| 75 | # between midnight and 6.15am every day, only allow up to 6 hours missing |
|---|
| 76 | # if there is more than this, we will continue asking grabbers for more |
|---|
| 77 | # programming on this channel |
|---|
| 78 | $policy{nonpeak_max_missing} = 6*(60*60); # up to 6 hours can be missing |
|---|
| 79 | $policy{nonpeak_start} = 0; # midnight |
|---|
| 80 | $policy{nonpeak_stop} = (6*(60*60))+(15*60); # 6.15am |
|---|
| 81 | |
|---|
| 82 | # all other timeslots - (6.15am-4.30pm, 11.30pm-midnight) |
|---|
| 83 | # allow up to 60 minutes maximum missing programming |
|---|
| 84 | $policy{other_max_missing} = 60*60; # up to 60 mins max allowed missing |
|---|
| 85 | |
|---|
| 86 | # if a postprocessor failed 5 times in a row, automatically disable it |
|---|
| 87 | $policy{postprocessor_disable_failure_threshold} = 5; |
|---|
| 88 | |
|---|
| 89 | #### end analyzer section #### |
|---|
| 90 | |
|---|
| 91 | my $opt = { }; |
|---|
| 92 | my $pref_title_source; |
|---|
| 93 | my $mirror_site; |
|---|
| 94 | my $debug = 0; |
|---|
| 95 | my $components = { }; |
|---|
| 96 | my $gscore; |
|---|
| 97 | my $region; |
|---|
| 98 | my $channels; |
|---|
| 99 | my $opt_channels; |
|---|
| 100 | my $config_file = "$CWD/$progname.conf"; |
|---|
| 101 | my $channels_file = "$CWD/channels.conf"; |
|---|
| 102 | my $days = 7; |
|---|
| 103 | my $missing; |
|---|
| 104 | my $timeslice; |
|---|
| 105 | my $grabbed; |
|---|
| 106 | my $gmt_offset; |
|---|
| 107 | my $grabber_found_all_data; |
|---|
| 108 | my $writer; |
|---|
| 109 | my $components_used = ""; |
|---|
| 110 | |
|---|
| 111 | # postprocessing |
|---|
| 112 | my $langs = [ 'en' ]; |
|---|
| 113 | my $plugin_data = { }; |
|---|
| 114 | my $channel_data = { }; |
|---|
| 115 | my $reconciler_found_all_data; |
|---|
| 116 | my $input_postprocess_file = ""; |
|---|
| 117 | |
|---|
| 118 | # OBSOLETE: will be removed |
|---|
| 119 | my $preferred; |
|---|
| 120 | my $title_translation_table; |
|---|
| 121 | my $pref_order; |
|---|
| 122 | |
|---|
| 123 | # --------------------------------------------------------------------------- |
|---|
| 124 | # --- Setup |
|---|
| 125 | # --------------------------------------------------------------------------- |
|---|
| 126 | |
|---|
| 127 | # Any options Shepherd doesn't understand, we'll pass to the grabber(s) |
|---|
| 128 | Getopt::Long::Configure(qw/pass_through/); |
|---|
| 129 | &get_command_line_options; |
|---|
| 130 | |
|---|
| 131 | if ($opt->{version}) { print "$version\n"; exit 0; } # This is horrible |
|---|
| 132 | $| = 1; |
|---|
| 133 | print ucfirst($myprogname) . " v$version\n\n"; |
|---|
| 134 | |
|---|
| 135 | exit 0 if ($opt->{version}); |
|---|
| 136 | die "run tv_grab_au --help for details.\n" if ($opt->{help}); |
|---|
| 137 | |
|---|
| 138 | &read_config_file; |
|---|
| 139 | &read_channels_file; |
|---|
| 140 | |
|---|
| 141 | # --------------------------------------------------------------------------- |
|---|
| 142 | # --- Go! |
|---|
| 143 | # --------------------------------------------------------------------------- |
|---|
| 144 | |
|---|
| 145 | &calc_date_range; |
|---|
| 146 | &grab_data; |
|---|
| 147 | &reconcile_data; |
|---|
| 148 | &postprocess_data; |
|---|
| 149 | &output_data; |
|---|
| 150 | &write_config_file; |
|---|
| 151 | |
|---|
| 152 | # --------------------------------------------------------------------------- |
|---|
| 153 | # --- Subroutines |
|---|
| 154 | # --------------------------------------------------------------------------- |
|---|
| 155 | |
|---|
| 156 | # ----------------------------------------- |
|---|
| 157 | # Subs: Grabbing |
|---|
| 158 | # ----------------------------------------- |
|---|
| 159 | |
|---|
| 160 | sub grab_data |
|---|
| 161 | { |
|---|
| 162 | my $used_grabbers = 0; |
|---|
| 163 | |
|---|
| 164 | printf "\nGrabber stage.\n"; |
|---|
| 165 | |
|---|
| 166 | &analyze_plugin_data("",1); |
|---|
| 167 | |
|---|
| 168 | while (my $grabber = choose_grabber()) |
|---|
| 169 | { |
|---|
| 170 | $grabber_found_all_data = 0; |
|---|
| 171 | $used_grabbers++; |
|---|
| 172 | |
|---|
| 173 | $components->{$grabber}->{laststatus} = "unknown"; |
|---|
| 174 | |
|---|
| 175 | printf "\nSHEPHERD: Using grabber: (%d) %s\n", $used_grabbers, $grabber; |
|---|
| 176 | |
|---|
| 177 | my $output = "$CWD/grabbers/$grabber/output.xmltv"; |
|---|
| 178 | |
|---|
| 179 | my $comm = "$CWD/grabbers/$grabber/$grabber " . |
|---|
| 180 | "--region $region " . |
|---|
| 181 | "--output $output"; |
|---|
| 182 | |
|---|
| 183 | # Category 1 grabbers (i.e. slow ones) are requested to only fetch the timeslice |
|---|
| 184 | # that we need. Category 2 grabbers are requested to get everything, since there's |
|---|
| 185 | # very little cost in grabbing that extra data, and we can use it in the reconciler |
|---|
| 186 | # to verify that everything looks OK. |
|---|
| 187 | if (query_config($grabber, 'category') == 1) |
|---|
| 188 | { |
|---|
| 189 | printf "$grabber is Category 1: grabbing timeslice.\n" if ($debug); |
|---|
| 190 | |
|---|
| 191 | record_requested_chandays($grabber, $timeslice); |
|---|
| 192 | |
|---|
| 193 | if ($timeslice->{start} != 0) |
|---|
| 194 | { |
|---|
| 195 | $comm .= " " . |
|---|
| 196 | query_config($grabber, 'option_days_offset') . |
|---|
| 197 | " " . |
|---|
| 198 | $timeslice->{start}; |
|---|
| 199 | } |
|---|
| 200 | |
|---|
| 201 | my $n = $timeslice->{stop} + 1; |
|---|
| 202 | if ($timeslice->{start} != 0 |
|---|
| 203 | and |
|---|
| 204 | !query_config($grabber, 'option_offset_eats_days')) |
|---|
| 205 | { |
|---|
| 206 | $n -= $timeslice->{start}; |
|---|
| 207 | } |
|---|
| 208 | $comm .= " " . |
|---|
| 209 | query_config($grabber, 'option_days') . |
|---|
| 210 | " " . |
|---|
| 211 | $n; |
|---|
| 212 | |
|---|
| 213 | # Write a temporary channels file specifying only the channels we want |
|---|
| 214 | my $tmpchans; |
|---|
| 215 | foreach (@{$timeslice->{chans}}) |
|---|
| 216 | { |
|---|
| 217 | $tmpchans->{$_} = $channels->{$_}; |
|---|
| 218 | } |
|---|
| 219 | my $tmpcf = "$CWD/channels.conf.tmp"; |
|---|
| 220 | write_file($tmpcf, 'temporary channels', [ $tmpchans ], [ 'channels' ]); |
|---|
| 221 | $comm .= " --channels_file $tmpcf"; |
|---|
| 222 | } |
|---|
| 223 | else |
|---|
| 224 | { |
|---|
| 225 | printf "$grabber is category 2: grabbing everything.\n" if ($debug); |
|---|
| 226 | $comm .= " --days $days" if ($days); |
|---|
| 227 | $comm .= " --offset $opt->{offset}" if ($opt->{offset}); |
|---|
| 228 | $comm .= " --channels_file $channels_file"; |
|---|
| 229 | } |
|---|
| 230 | $comm .= " --debug" if ($debug); |
|---|
| 231 | $comm .= " @ARGV" if (@ARGV); |
|---|
| 232 | |
|---|
| 233 | my $retval = 0; |
|---|
| 234 | if ((defined $opt->{dontcallgrabbers}) && ($opt->{dontcallgrabbers})) { |
|---|
| 235 | printf "SHEPHERD: not calling grabber because of --dontcallgrabbers option, but will instead use existing $output\n"; |
|---|
| 236 | printf "SHEPHERD: would have called: $comm\n" if ($debug); |
|---|
| 237 | } else { |
|---|
| 238 | printf "SHEPHERD: Excuting command: $comm\n"; |
|---|
| 239 | chdir "$CWD/grabbers/$grabber/"; |
|---|
| 240 | $retval = call_prog($comm); |
|---|
| 241 | chdir $CWD; |
|---|
| 242 | } |
|---|
| 243 | |
|---|
| 244 | if ($retval != 0) { |
|---|
| 245 | printf "grabber returned with non-zero return code $retval: assuming it failed.\n"; |
|---|
| 246 | next; |
|---|
| 247 | } |
|---|
| 248 | |
|---|
| 249 | # soak up the data we just collected |
|---|
| 250 | &soak_up_data($grabber, $output, "grabber"); |
|---|
| 251 | $components->{$grabber}->{laststatus} = $plugin_data->{$grabber}->{laststatus}; |
|---|
| 252 | $components->{$grabber}->{lastdata} = time if ($plugin_data->{$grabber}->{valid}); |
|---|
| 253 | |
|---|
| 254 | # check to see if we have all the data we want |
|---|
| 255 | $grabber_found_all_data = &analyze_plugin_data("analysis of all grabbers so far"); |
|---|
| 256 | |
|---|
| 257 | # Record what we grabbed from cacheable C1 grabbers |
|---|
| 258 | if (query_config($grabber, 'category') == 1 and query_config($grabber, 'cache')) |
|---|
| 259 | { |
|---|
| 260 | my $missing_before = convert_dayhash_to_list($missing); |
|---|
| 261 | my $missing_after = convert_dayhash_to_list(detect_missing_data()); |
|---|
| 262 | my $list = List::Compare->new($missing_before, $missing_after); |
|---|
| 263 | my @grabbed = $list->get_symmetric_difference(); |
|---|
| 264 | printf "Grabbed: " . join (', ', @grabbed) . ".\n" if ($debug); |
|---|
| 265 | record_cached($grabber, @grabbed); |
|---|
| 266 | write_config_file(); |
|---|
| 267 | } |
|---|
| 268 | |
|---|
| 269 | last if ($grabber_found_all_data); |
|---|
| 270 | } |
|---|
| 271 | |
|---|
| 272 | |
|---|
| 273 | if ($used_grabbers == 0) |
|---|
| 274 | { |
|---|
| 275 | printf "No valid grabbers installed/enabled!\n"; |
|---|
| 276 | return; |
|---|
| 277 | } |
|---|
| 278 | |
|---|
| 279 | unless ($grabber_found_all_data) |
|---|
| 280 | { |
|---|
| 281 | printf "SHEPHERD: Ran through ALL grabbers but still missing data!!! :(\n"; |
|---|
| 282 | return; |
|---|
| 283 | } |
|---|
| 284 | } |
|---|
| 285 | |
|---|
| 286 | # ----------------------------------------- |
|---|
| 287 | # Subs: Intelli-random grabber selection |
|---|
| 288 | # ----------------------------------------- |
|---|
| 289 | |
|---|
| 290 | sub choose_grabber |
|---|
| 291 | { |
|---|
| 292 | if (defined $gscore) # Reset score hash |
|---|
| 293 | { |
|---|
| 294 | foreach (keys %$gscore) |
|---|
| 295 | { |
|---|
| 296 | $gscore->{$_} = 0; |
|---|
| 297 | } |
|---|
| 298 | } |
|---|
| 299 | else # Create score hash |
|---|
| 300 | { |
|---|
| 301 | foreach (query_grabbers()) |
|---|
| 302 | { |
|---|
| 303 | unless ($components->{$_}->{disabled}) |
|---|
| 304 | { |
|---|
| 305 | $gscore->{$_} = 0; |
|---|
| 306 | if (query_config($_, 'category') == 1 and query_config($_, 'cache')) |
|---|
| 307 | { |
|---|
| 308 | $gscore->{$_ . ' [cache]'} = 0; |
|---|
| 309 | } |
|---|
| 310 | } |
|---|
| 311 | } |
|---|
| 312 | } |
|---|
| 313 | |
|---|
| 314 | $missing = detect_missing_data(); |
|---|
| 315 | $timeslice = find_best_timeslice(); |
|---|
| 316 | |
|---|
| 317 | if ($debug) |
|---|
| 318 | { |
|---|
| 319 | printf "Best timeslice: day%s of channels %s (%d chandays).\n", |
|---|
| 320 | ($timeslice->{start} == $timeslice->{stop} ? |
|---|
| 321 | " $timeslice->{start}" : |
|---|
| 322 | "s $timeslice->{start} - $timeslice->{stop}"), |
|---|
| 323 | join(', ', @{$timeslice->{chans}}), |
|---|
| 324 | $timeslice->{chandays}; |
|---|
| 325 | } |
|---|
| 326 | |
|---|
| 327 | my $total = score_grabbers(); |
|---|
| 328 | |
|---|
| 329 | if ($debug) |
|---|
| 330 | { |
|---|
| 331 | printf "Grabber selection:\n"; |
|---|
| 332 | foreach (sort { $gscore->{$b} <=> $gscore->{$a} } keys %$gscore) |
|---|
| 333 | { |
|---|
| 334 | next if ($_ =~ /\[cache\]/); |
|---|
| 335 | |
|---|
| 336 | my $score = $gscore->{$_}; |
|---|
| 337 | my $cscore = $gscore->{"$_ [cache]"}; |
|---|
| 338 | my $cstr = $cscore ? "(inc. $cscore cache pts)" : ""; |
|---|
| 339 | |
|---|
| 340 | if ($opt->{randomize}) |
|---|
| 341 | { |
|---|
| 342 | printf "%15s %6.1f%% %9s %s\n", |
|---|
| 343 | $_, |
|---|
| 344 | ($total ? 100* $score / $total : 0), |
|---|
| 345 | "$score pts", |
|---|
| 346 | $cstr; |
|---|
| 347 | } |
|---|
| 348 | else |
|---|
| 349 | { |
|---|
| 350 | printf "%15s %4s pts %s\n", |
|---|
| 351 | $_, |
|---|
| 352 | $score, |
|---|
| 353 | $cstr; |
|---|
| 354 | } |
|---|
| 355 | } |
|---|
| 356 | } |
|---|
| 357 | |
|---|
| 358 | return undef unless ($total); |
|---|
| 359 | |
|---|
| 360 | # Select a grabber |
|---|
| 361 | |
|---|
| 362 | # If the user has specified a pref_title_source -- i.e. he is |
|---|
| 363 | # transitioning from a known grabber -- then we make sure it |
|---|
| 364 | # has run at least once, to build the list of title translations. |
|---|
| 365 | if ($pref_title_source) |
|---|
| 366 | { |
|---|
| 367 | my @prefs = split(/,/, $pref_title_source); |
|---|
| 368 | foreach my $grabber (@prefs) |
|---|
| 369 | { |
|---|
| 370 | unless ($components->{$grabber}->{lastdata}) |
|---|
| 371 | { |
|---|
| 372 | printf "Need to build title translation list for transitional grabber $grabber.\n"; |
|---|
| 373 | return select_grabber($grabber, $gscore) if ($gscore->{$grabber}); |
|---|
| 374 | printf "WARNING: Can't run $grabber to build title translation list!\n"; |
|---|
| 375 | } |
|---|
| 376 | } |
|---|
| 377 | } |
|---|
| 378 | |
|---|
| 379 | # Either do it randomly based on scores, or just return the |
|---|
| 380 | # highest-scoring grabber, depending on whether --randomize has |
|---|
| 381 | # been used. |
|---|
| 382 | |
|---|
| 383 | my $r = int(rand($total)); |
|---|
| 384 | my $c = 0; |
|---|
| 385 | my $best; |
|---|
| 386 | |
|---|
| 387 | foreach my $grabber (keys %$gscore) |
|---|
| 388 | { |
|---|
| 389 | next if (!$gscore->{$grabber} or $grabber =~ /\[cache\]/); |
|---|
| 390 | if ($opt->{randomize}) |
|---|
| 391 | { |
|---|
| 392 | if ($r >= $c and $r < ($c + $gscore->{$grabber})) |
|---|
| 393 | { |
|---|
| 394 | return select_grabber($grabber, $gscore); |
|---|
| 395 | } |
|---|
| 396 | $c += $gscore->{$grabber}; |
|---|
| 397 | } |
|---|
| 398 | else |
|---|
| 399 | { |
|---|
| 400 | if (!$best or $gscore->{$grabber} > $gscore->{$best}) |
|---|
| 401 | { |
|---|
| 402 | $best = $grabber; |
|---|
| 403 | } |
|---|
| 404 | } |
|---|
| 405 | } |
|---|
| 406 | |
|---|
| 407 | if ($opt->{randomize} or !$best) |
|---|
| 408 | { |
|---|
| 409 | die "ERROR: failed to choose grabber."; |
|---|
| 410 | } |
|---|
| 411 | return select_grabber($best, $gscore); |
|---|
| 412 | } |
|---|
| 413 | |
|---|
| 414 | sub select_grabber |
|---|
| 415 | { |
|---|
| 416 | my ($grabber, $gscore) = @_; |
|---|
| 417 | |
|---|
| 418 | printf "Selected $grabber.\n" if ($debug); |
|---|
| 419 | if (query_config($grabber, 'category') == 2) |
|---|
| 420 | { |
|---|
| 421 | # We might want to run C1 grabbers multiple times |
|---|
| 422 | # to grab various timeslices, but not C2 grabbers, |
|---|
| 423 | # which should get everything at once. |
|---|
| 424 | delete $gscore->{$grabber}; |
|---|
| 425 | } |
|---|
| 426 | return $grabber; |
|---|
| 427 | } |
|---|
| 428 | |
|---|
| 429 | # Grabbers earn 1 point for each slot or chanday they can fill. |
|---|
| 430 | # This score is multiplied if the grabber: |
|---|
| 431 | # * is a category 2 grabber (i.e. fast/cheap) |
|---|
| 432 | # * is a category 1 grabber that has the data we want in a cache |
|---|
| 433 | # * can supply high-quality data |
|---|
| 434 | # Very low quality grabbers score 0 unless we need them; i.e. they're backups. |
|---|
| 435 | sub score_grabbers |
|---|
| 436 | { |
|---|
| 437 | my ($score, $total, $day, $catbonus, $dqbonus, $mult, $key); |
|---|
| 438 | |
|---|
| 439 | my $bestdq = 0; |
|---|
| 440 | |
|---|
| 441 | # Compare C2 grabbers against the raw missing file, because we'll get |
|---|
| 442 | # everything. But compare C1 grabbers against the timeslice, because we'll |
|---|
| 443 | # only ask them for a slice. This goes for the [cache] and regular C1s. |
|---|
| 444 | foreach my $grabber (keys %$gscore) |
|---|
| 445 | { |
|---|
| 446 | # for each slot, say whether we can fill it or not -- that is, |
|---|
| 447 | # whether we support this channel and this day #. |
|---|
| 448 | |
|---|
| 449 | my $hits = 0; |
|---|
| 450 | my $cat = query_config($grabber, 'category'); |
|---|
| 451 | my $dq = query_config($grabber, 'quality'); |
|---|
| 452 | |
|---|
| 453 | if ($cat == 1) |
|---|
| 454 | { |
|---|
| 455 | $key = cut_down_missing($grabber); |
|---|
| 456 | # printf "Grabber $grabber is Category 1: comparing capability to best timeslice.\n" if ($debug); |
|---|
| 457 | } |
|---|
| 458 | else |
|---|
| 459 | { |
|---|
| 460 | $key = $missing; |
|---|
| 461 | # printf "Grabber $grabber is Category 2: comparing capability to all wanted channels and days.\n" if ($debug); |
|---|
| 462 | } |
|---|
| 463 | |
|---|
| 464 | if ($grabber =~ /\[cache\]/) |
|---|
| 465 | { |
|---|
| 466 | $hits = find_cache_hits($grabber, $key); |
|---|
| 467 | } |
|---|
| 468 | else |
|---|
| 469 | { |
|---|
| 470 | foreach my $day (sort keys %$key) |
|---|
| 471 | { |
|---|
| 472 | my $val = supports_day($grabber, $day); |
|---|
| 473 | next unless ($val); |
|---|
| 474 | # printf "Day $day:" if ($debug); |
|---|
| 475 | foreach my $ch (@{$key->{$day}}) |
|---|
| 476 | { |
|---|
| 477 | if (supports_channel($grabber, $ch, $day)) |
|---|
| 478 | { |
|---|
| 479 | # printf " $ch" if ($debug); |
|---|
| 480 | $hits += $val; |
|---|
| 481 | } |
|---|
| 482 | } |
|---|
| 483 | # printf "\n" if $debug; |
|---|
| 484 | $hits = 1 if ($hits > 0 and $hits < 1); |
|---|
| 485 | } |
|---|
| 486 | } |
|---|
| 487 | |
|---|
| 488 | my $catbonus = 1; |
|---|
| 489 | $catbonus = 3 if ($cat == 2); |
|---|
| 490 | if ($grabber =~ /\[cache\]/) |
|---|
| 491 | { |
|---|
| 492 | # Bonus is on a sliding scale between 1 and 2 depending on |
|---|
| 493 | # % of required data in cache |
|---|
| 494 | $catbonus += $hits / $timeslice->{chandays}; |
|---|
| 495 | } |
|---|
| 496 | |
|---|
| 497 | my $dqbonus = 2 ** ($dq-1); |
|---|
| 498 | |
|---|
| 499 | my $mult = $dq ** $catbonus; |
|---|
| 500 | |
|---|
| 501 | my $score = int($hits * $mult); |
|---|
| 502 | |
|---|
| 503 | if ($debug) |
|---|
| 504 | { |
|---|
| 505 | my $str = sprintf "Grabber %s can supply %d chandays", |
|---|
| 506 | $grabber, $hits; |
|---|
| 507 | if ($hits) |
|---|
| 508 | { |
|---|
| 509 | $str .= sprintf " at x%.1f (cat: %d, DQ: %d): %d pts", |
|---|
| 510 | $mult, |
|---|
| 511 | $cat, |
|---|
| 512 | $dq, |
|---|
| 513 | $score; |
|---|
| 514 | } |
|---|
| 515 | printf "$str.\n"; |
|---|
| 516 | } |
|---|
| 517 | |
|---|
| 518 | $gscore->{$grabber} += $score; |
|---|
| 519 | $total += $score; |
|---|
| 520 | if ($grabber =~ /\[cache\]/) |
|---|
| 521 | { |
|---|
| 522 | $gscore->{query_name($grabber)} += $score; |
|---|
| 523 | } |
|---|
| 524 | |
|---|
| 525 | if ($score and $dq > $bestdq) |
|---|
| 526 | { |
|---|
| 527 | $bestdq = $dq; |
|---|
| 528 | } |
|---|
| 529 | } |
|---|
| 530 | |
|---|
| 531 | # Eliminate grabbers of data quality 1 if there are any better-quality |
|---|
| 532 | # alternatives when using randomize. |
|---|
| 533 | if ($opt->{randomize}) |
|---|
| 534 | { |
|---|
| 535 | foreach (keys %$gscore) |
|---|
| 536 | { |
|---|
| 537 | if ($gscore->{$_} |
|---|
| 538 | and |
|---|
| 539 | query_config($_, 'quality') == 1 |
|---|
| 540 | and |
|---|
| 541 | $bestdq > 1) |
|---|
| 542 | { |
|---|
| 543 | $total -= $gscore->{$_}; |
|---|
| 544 | $gscore->{$_} = 0; |
|---|
| 545 | printf "Zeroing grabber $_ due to low data quality.\n" if ($debug); |
|---|
| 546 | } |
|---|
| 547 | } |
|---|
| 548 | } |
|---|
| 549 | |
|---|
| 550 | return $total; |
|---|
| 551 | } |
|---|
| 552 | |
|---|
| 553 | # Return 1 if the grabber can provide data for this channel, else 0. |
|---|
| 554 | sub supports_channel |
|---|
| 555 | { |
|---|
| 556 | my ($grabber, $ch, $day) = @_; |
|---|
| 557 | |
|---|
| 558 | my $mdpc = query_config($grabber, 'max_days_per_chan'); |
|---|
| 559 | if ($mdpc) |
|---|
| 560 | { |
|---|
| 561 | if ($mdpc->{$ch}) |
|---|
| 562 | { |
|---|
| 563 | return ($mdpc->{$ch} > $day); |
|---|
| 564 | } |
|---|
| 565 | } |
|---|
| 566 | |
|---|
| 567 | my $channels_supported = query_config($grabber, 'channels'); |
|---|
| 568 | unless (defined $channels_supported) |
|---|
| 569 | { |
|---|
| 570 | printf "WARNING: Grabber $grabber has no channel support " . |
|---|
| 571 | "specified in config.\n"; |
|---|
| 572 | $channels_supported = ''; |
|---|
| 573 | } |
|---|
| 574 | |
|---|
| 575 | return 1 unless ($channels_supported); # Empty string means we support all |
|---|
| 576 | |
|---|
| 577 | $ch =~ s/ /_/g; |
|---|
| 578 | my $match = ($channels_supported =~ /\b$ch\b/); |
|---|
| 579 | my $exceptions = ($channels_supported =~/^-/); |
|---|
| 580 | return ($match != $exceptions); |
|---|
| 581 | } |
|---|
| 582 | |
|---|
| 583 | # Return 0 if the grabber can't provide data for this day, |
|---|
| 584 | # 1 if it can reliably, and 0.5 if it can unreliably. |
|---|
| 585 | # |
|---|
| 586 | # Note that a max_days of 7 means the grabber can retrieve data for |
|---|
| 587 | # today plus 6 days. |
|---|
| 588 | sub supports_day |
|---|
| 589 | { |
|---|
| 590 | my ($grabber, $day) = @_; |
|---|
| 591 | |
|---|
| 592 | return 0 unless ($day < query_config($grabber, 'max_days')); |
|---|
| 593 | return 0.5 if ($day >= query_config($grabber, 'max_reliable_days')); |
|---|
| 594 | return 1; |
|---|
| 595 | } |
|---|
| 596 | |
|---|
| 597 | sub find_cache_hits |
|---|
| 598 | { |
|---|
| 599 | my ($grabber, $key) = @_; |
|---|
| 600 | |
|---|
| 601 | $grabber = query_name($grabber); |
|---|
| 602 | |
|---|
| 603 | return 0 unless ($components->{$grabber}->{cached}); |
|---|
| 604 | |
|---|
| 605 | my $hits = 0; |
|---|
| 606 | |
|---|
| 607 | foreach my $day (keys %$key) |
|---|
| 608 | { |
|---|
| 609 | my $date = substr(DateCalc("today", "+ $day days"), 0, 8); |
|---|
| 610 | foreach my $ch (@{$key->{$day}}) |
|---|
| 611 | { |
|---|
| 612 | $hits++ if (grep(/^$date:$ch$/, @{$components->{$grabber}->{cached}})); |
|---|
| 613 | } |
|---|
| 614 | } |
|---|
| 615 | return $hits; |
|---|
| 616 | } |
|---|
| 617 | |
|---|
| 618 | # Build a dayhash of what channel/day data we're currently missing. |
|---|
| 619 | # I think granularity of one day is good for now; could possibly be |
|---|
| 620 | # made more fine-grained if we think grabbers will support that. |
|---|
| 621 | sub detect_missing_data |
|---|
| 622 | { |
|---|
| 623 | my $m = { }; |
|---|
| 624 | |
|---|
| 625 | my $chandays = 0; |
|---|
| 626 | foreach my $ch (keys %$channels) |
|---|
| 627 | { |
|---|
| 628 | # is this channel missing too much data? |
|---|
| 629 | unless ($channel_data->{$ch}->{analysis}->{data_ok}) { |
|---|
| 630 | # not ok - record which days are bad |
|---|
| 631 | foreach my $day (@{($channel_data->{$ch}->{analysis}->{day})}) { |
|---|
| 632 | push(@{($m->{($day->{num})})}, $ch) unless ($day->{day_ok}); |
|---|
| 633 | } |
|---|
| 634 | } |
|---|
| 635 | } |
|---|
| 636 | |
|---|
| 637 | foreach my $day (keys %$m) |
|---|
| 638 | { |
|---|
| 639 | $m->{$day} = [ sort @{$m->{$day}} ]; |
|---|
| 640 | $chandays += scalar(@{$m->{$day}}) if ($debug); |
|---|
| 641 | } |
|---|
| 642 | |
|---|
| 643 | if ($debug) |
|---|
| 644 | { |
|---|
| 645 | printf "Need data for days " . join(", ", sort keys %$m) . |
|---|
| 646 | " ($chandays chandays).\n"; |
|---|
| 647 | } |
|---|
| 648 | return $m; |
|---|
| 649 | } |
|---|
| 650 | |
|---|
| 651 | # Find the largest timeslice in the current $missing dayhash; i.e. |
|---|
| 652 | # something like "Days 4 - 6 of ABC and SBS." This works by iterating |
|---|
| 653 | # through the days and looking for overlaps where consecutive days |
|---|
| 654 | # want the same channels. |
|---|
| 655 | sub find_best_timeslice |
|---|
| 656 | { |
|---|
| 657 | my ($overlap, $a); |
|---|
| 658 | my $slice = { 'chandays' => 0 }; |
|---|
| 659 | |
|---|
| 660 | foreach my $day (0 .. $days-1) |
|---|
| 661 | { |
|---|
| 662 | consider_slice($slice, $day, $day, @{$missing->{$day}}); |
|---|
| 663 | $overlap = $missing->{$day}; |
|---|
| 664 | foreach my $nextday (($day + 1) .. $days-1) |
|---|
| 665 | { |
|---|
| 666 | last unless ($missing->{$nextday}); |
|---|
| 667 | $a = Algorithm::Diff::LCS($overlap, $missing->{$nextday}); |
|---|
| 668 | last unless ($a and @{$a}); |
|---|
| 669 | consider_slice($slice, $day, $nextday, @{$a}); |
|---|
| 670 | $overlap = $a; |
|---|
| 671 | } |
|---|
| 672 | } |
|---|
| 673 | return $slice; |
|---|
| 674 | } |
|---|
| 675 | |
|---|
| 676 | sub consider_slice |
|---|
| 677 | { |
|---|
| 678 | my ($slice, $startday, $stopday, @chans) = @_; |
|---|
| 679 | |
|---|
| 680 | my $challenger = ($stopday - $startday + 1) * scalar(@chans); |
|---|
| 681 | return unless ($challenger > $slice->{chandays}); |
|---|
| 682 | |
|---|
| 683 | # We have a winner! |
|---|
| 684 | $slice->{start} = $startday; |
|---|
| 685 | $slice->{stop} = $stopday; |
|---|
| 686 | $slice->{chans} = [ @chans ]; |
|---|
| 687 | $slice->{chandays} = $challenger; |
|---|
| 688 | } |
|---|
| 689 | |
|---|
| 690 | # Record what a cacheable C1 grabber has just retrieved for us, |
|---|
| 691 | # so we know next time that this data can be grabbed quickly. |
|---|
| 692 | sub record_cached |
|---|
| 693 | { |
|---|
| 694 | my ($grabber, @grabbed) = @_; |
|---|
| 695 | |
|---|
| 696 | printf "Recording cache for grabber $grabber.\n" if ($debug); |
|---|
| 697 | |
|---|
| 698 | my $gcache = $components->{$grabber}->{cached}; |
|---|
| 699 | $gcache = [ ] unless ($gcache); |
|---|
| 700 | my @newcache; |
|---|
| 701 | my $today = strftime("%Y%m%d", localtime); |
|---|
| 702 | |
|---|
| 703 | # remove old chandays |
|---|
| 704 | foreach my $chanday (@$gcache) |
|---|
| 705 | { |
|---|
| 706 | $chanday =~ /(\d+):(.*)/; |
|---|
| 707 | if ($1 >= $today) |
|---|
| 708 | { |
|---|
| 709 | push (@newcache, $chanday); |
|---|
| 710 | } |
|---|
| 711 | } |
|---|
| 712 | |
|---|
| 713 | # record new chandays |
|---|
| 714 | foreach my $chanday (@grabbed) |
|---|
| 715 | { |
|---|
| 716 | push (@newcache, $chanday) unless (grep(/^$chanday$/, @newcache)); |
|---|
| 717 | } |
|---|
| 718 | $components->{$grabber}->{cached} = [ @newcache ]; |
|---|
| 719 | } |
|---|
| 720 | |
|---|
| 721 | # Takes a dayhash and returns it as a list like this: |
|---|
| 722 | # ( "20061018:ABC", "20061018:Seven", ... ) |
|---|
| 723 | sub convert_dayhash_to_list |
|---|
| 724 | { |
|---|
| 725 | my $h = shift; |
|---|
| 726 | |
|---|
| 727 | my @ret; |
|---|
| 728 | foreach my $day (keys %$h) |
|---|
| 729 | { |
|---|
| 730 | my $date = substr(DateCalc("today", "+ $day days"), 0, 8); |
|---|
| 731 | foreach my $ch (@{$h->{$day}}) |
|---|
| 732 | { |
|---|
| 733 | push (@ret, "$date:$ch"); |
|---|
| 734 | } |
|---|
| 735 | } |
|---|
| 736 | @ret = sort @ret; |
|---|
| 737 | return \@ret; |
|---|
| 738 | } |
|---|
| 739 | |
|---|
| 740 | # If we're about to re-try a grabber, make sure that we're not asking |
|---|
| 741 | # it for the same data. That is, prevent a broken C1 grabber causing |
|---|
| 742 | # an infinite loop. |
|---|
| 743 | sub record_requested_chandays |
|---|
| 744 | { |
|---|
| 745 | my ($grabber, $slice) = @_; |
|---|
| 746 | |
|---|
| 747 | printf "Recording timeslice request; will not request these chandays " . |
|---|
| 748 | "from $grabber again.\n" if ($debug); |
|---|
| 749 | |
|---|
| 750 | my @requested; |
|---|
| 751 | for my $day ($slice->{start} .. $slice->{stop}) |
|---|
| 752 | { |
|---|
| 753 | foreach my $ch (@{$slice->{chans}}) |
|---|
| 754 | { |
|---|
| 755 | push @requested, "$day:$ch"; |
|---|
| 756 | } |
|---|
| 757 | } |
|---|
| 758 | if ($grabbed->{$grabber}) |
|---|
| 759 | { |
|---|
| 760 | push @{$grabbed->{$grabber}}, @requested; |
|---|
| 761 | } |
|---|
| 762 | else |
|---|
| 763 | { |
|---|
| 764 | $grabbed->{$grabber} = [ @requested ]; |
|---|
| 765 | } |
|---|
| 766 | } |
|---|
| 767 | |
|---|
| 768 | # If this grabber has been called previously, remove those chandays |
|---|
| 769 | # from the current request -- we don't want to ask it over and over |
|---|
| 770 | # for a timeslice that it has already failed to provide. |
|---|
| 771 | sub cut_down_missing |
|---|
| 772 | { |
|---|
| 773 | my $grabber = shift; |
|---|
| 774 | |
|---|
| 775 | $grabber = query_name($grabber); |
|---|
| 776 | my $dayhash = {}; |
|---|
| 777 | |
|---|
| 778 | # Take the timeslice and expand it to a dayhash, while pruning |
|---|
| 779 | # any chandays that have previously been requested from this |
|---|
| 780 | # grabber. |
|---|
| 781 | foreach my $day ($timeslice->{start} .. $timeslice->{stop}) |
|---|
| 782 | { |
|---|
| 783 | my @chans; |
|---|
| 784 | foreach my $ch (@{$timeslice->{chans}}) |
|---|
| 785 | { |
|---|
| 786 | unless ($grabbed->{$grabber} and grep(/$day:$ch/, @{$grabbed->{$grabber}})) |
|---|
| 787 | { |
|---|
| 788 | push (@chans, $ch) |
|---|
| 789 | } |
|---|
| 790 | } |
|---|
| 791 | $dayhash->{$day} = [ @chans ] if (@chans); |
|---|
| 792 | } |
|---|
| 793 | |
|---|
| 794 | return $dayhash; |
|---|
| 795 | } |
|---|
| 796 | |
|---|
| 797 | # ----------------------------------------- |
|---|
| 798 | # Subs: Analyzing data |
|---|
| 799 | # ----------------------------------------- |
|---|
| 800 | |
|---|
| 801 | # interpret xmltv data from this grabber/postprocessor |
|---|
| 802 | sub soak_up_data |
|---|
| 803 | { |
|---|
| 804 | my ($plugin, $output, $plugintype) = @_; |
|---|
| 805 | |
|---|
| 806 | if (! -r $output) { |
|---|
| 807 | printf "SHEPHERD: Warning: plugin '%s' output file '%s' does not exist\n",$plugin,$output; |
|---|
| 808 | return; |
|---|
| 809 | } |
|---|
| 810 | |
|---|
| 811 | my $this_plugin = $plugin_data->{$plugin}; |
|---|
| 812 | printf "SHEPHERD: Started parsing XMLTV from '%s' in '%s' .. any errors below are from parser:\n",$plugin,$output; |
|---|
| 813 | eval { $this_plugin->{xmltv} = XMLTV::parsefiles($output); }; |
|---|
| 814 | printf "SHEPHERD: Completed XMLTV parsing from '%s'\n",$plugin; |
|---|
| 815 | |
|---|
| 816 | if (!($this_plugin->{xmltv})) { |
|---|
| 817 | printf "WARNING: Plugin $plugin didn't seem to return any valid XMLTV!\n"; |
|---|
| 818 | return; |
|---|
| 819 | } |
|---|
| 820 | |
|---|
| 821 | $this_plugin->{valid} = 1; |
|---|
| 822 | $this_plugin->{output_filename} = $output; |
|---|
| 823 | $components_used .= " + ".$plugin."(v".$components->{$plugin}->{ver}.")"; |
|---|
| 824 | |
|---|
| 825 | my $xmltv = $this_plugin->{xmltv}; |
|---|
| 826 | my ($encoding, $credits, $chan, $progs) = @$xmltv; |
|---|
| 827 | $this_plugin->{total_duration} = 0; |
|---|
| 828 | $this_plugin->{programmes} = 0; |
|---|
| 829 | $this_plugin->{progs_with_invalid_date} = 0; # explicitly track unparsable dates |
|---|
| 830 | $this_plugin->{progs_with_unknown_channel} = 0; # explicitly track unknown channels |
|---|
| 831 | |
|---|
| 832 | my $seen_channels_with_data = 0; |
|---|
| 833 | |
|---|
| 834 | # |
|---|
| 835 | # first iterate through all programmes and see if there are any channels we don't know about |
|---|
| 836 | # |
|---|
| 837 | my %chan_xml_list; |
|---|
| 838 | foreach my $ch (sort keys %{$channels}) { |
|---|
| 839 | $chan_xml_list{($channels->{$ch})} = 1; |
|---|
| 840 | } |
|---|
| 841 | foreach my $prog (@$progs) { |
|---|
| 842 | if (!defined $chan_xml_list{($prog->{channel})}) { |
|---|
| 843 | $this_plugin->{progs_with_unknown_channel}++; |
|---|
| 844 | printf " - WARNING: plugin '%s' returned data for unknown channel '%s': ignored.\n",$plugin,$prog->{channel}; |
|---|
| 845 | $chan_xml_list{($prog->{channel})} = 1; # so we warn only once |
|---|
| 846 | } |
|---|
| 847 | } |
|---|
| 848 | |
|---|
| 849 | # iterate thru channels |
|---|
| 850 | foreach my $ch (sort keys %{$channels}) { |
|---|
| 851 | my $seen_progs_on_this_channel = 0; |
|---|
| 852 | |
|---|
| 853 | # iterate thru programmes per channel |
|---|
| 854 | foreach my $prog (@$progs) { |
|---|
| 855 | next if ($prog->{channel} ne $channels->{$ch}); |
|---|
| 856 | |
|---|
| 857 | my $t1 = &parse_xmltv_date($prog->{start}); |
|---|
| 858 | my $t2 = &parse_xmltv_date($prog->{stop}); |
|---|
| 859 | |
|---|
| 860 | if (!$t1 || !$t2) { |
|---|
| 861 | printf " - WARNING: plugin '%s' returned programme data with invalid timestamp format: \"%s\": can't parse.\n", |
|---|
| 862 | $plugin,(!$t1 ? $prog->{start} : $prog->{stop}) if (!$this_plugin->{progs_with_invalid_date}); |
|---|
| 863 | $this_plugin->{progs_with_invalid_date}++; |
|---|
| 864 | next; |
|---|
| 865 | } |
|---|
| 866 | |
|---|
| 867 | # store plugin-specific stats |
|---|
| 868 | $this_plugin->{programmes}++; |
|---|
| 869 | $this_plugin->{total_duration} += ($t2 - $t1); |
|---|
| 870 | $seen_progs_on_this_channel++; |
|---|
| 871 | $this_plugin->{earliest_data_seen} = $t1 if (!defined $this_plugin->{earliest_data_seen}); |
|---|
| 872 | $this_plugin->{earliest_data_seen} = $t1 if ($t1 < $this_plugin->{earliest_data_seen}); |
|---|
| 873 | $this_plugin->{latest_data_seen} = $t2 if (!defined $this_plugin->{latest_data_seen}); |
|---|
| 874 | $this_plugin->{latest_data_seen} = $t2 if ($t2 > $this_plugin->{latest_data_seen}); |
|---|
| 875 | |
|---|
| 876 | # store channel-specific stats |
|---|
| 877 | $channel_data->{$ch}->{programmes}++; |
|---|
| 878 | $channel_data->{$ch}->{total_duration} += ($t2 - $t1); |
|---|
| 879 | |
|---|
| 880 | # programme is outside the timeslots we are interested in. |
|---|
| 881 | next if ($t1 > $policy{endtime}); |
|---|
| 882 | next if ($t2 < $policy{starttime}); |
|---|
| 883 | |
|---|
| 884 | # store timeslot info |
|---|
| 885 | my $start_slotnum = 0; |
|---|
| 886 | $start_slotnum = int(($t1 - $policy{starttime}) / $policy{timeslot_size}) |
|---|
| 887 | if ($t1 >= $policy{starttime}); |
|---|
| 888 | |
|---|
| 889 | my $end_slotnum = ($policy{num_timeslots}-1); |
|---|
| 890 | $end_slotnum = int(($t2 - $policy{starttime}) / $policy{timeslot_size}) |
|---|
| 891 | if ($t2 < $policy{endtime}); |
|---|
| 892 | |
|---|
| 893 | # add this programme into the global timeslots table for this channel |
|---|
| 894 | foreach my $slotnum ($start_slotnum..$end_slotnum) { |
|---|
| 895 | $channel_data->{$ch}->{timeslots}[$slotnum]++; |
|---|
| 896 | } |
|---|
| 897 | } |
|---|
| 898 | |
|---|
| 899 | $seen_channels_with_data++ if ($seen_progs_on_this_channel > 0); |
|---|
| 900 | } |
|---|
| 901 | |
|---|
| 902 | # print some stats about what we saw! |
|---|
| 903 | printf "SHEPHERD: %s '%s' returned data for %d channels, %d programmes, %dd%02dh%02dm%02ds duration, %s%s\n", |
|---|
| 904 | ucfirst($plugintype), $plugin, $seen_channels_with_data, $this_plugin->{programmes}, |
|---|
| 905 | int($this_plugin->{total_duration} / 86400), # days |
|---|
| 906 | int(($this_plugin->{total_duration} % 86400) / 3600), # hours |
|---|
| 907 | int(($this_plugin->{total_duration} % 3600) / 60), # mins |
|---|
| 908 | int($this_plugin->{total_duration} % 60), # sec |
|---|
| 909 | (defined $this_plugin->{earliest_data_seen} ? (strftime "%a %e %b %H:%M - ", localtime($this_plugin->{earliest_data_seen})) : 'no data'), |
|---|
| 910 | (defined $this_plugin->{latest_data_seen} ? (strftime "%a %e %b %H:%M", localtime($this_plugin->{latest_data_seen})) : ''); |
|---|
| 911 | |
|---|
| 912 | $this_plugin->{laststatus} = sprintf "%dch/%dpr/%dhrs %s-%s", |
|---|
| 913 | $seen_channels_with_data, $this_plugin->{programmes}, |
|---|
| 914 | int($this_plugin->{total_duration} / 3600), |
|---|
| 915 | (defined $this_plugin->{earliest_data_seen} ? (strftime "%a%d%b", localtime($this_plugin->{earliest_data_seen})) : 'no'), |
|---|
| 916 | (defined $this_plugin->{latest_data_seen} ? (strftime "%a%d%b", localtime($this_plugin->{latest_data_seen})) : 'data'); |
|---|
| 917 | |
|---|
| 918 | $plugin_data->{$plugin} = $this_plugin; |
|---|
| 919 | } |
|---|
| 920 | |
|---|
| 921 | |
|---|
| 922 | # analyze grabber data - do we have all the data we want? |
|---|
| 923 | sub analyze_plugin_data |
|---|
| 924 | { |
|---|
| 925 | my ($analysistype,$quiet) = @_; |
|---|
| 926 | printf "SHEPHERD: $analysistype:\n" unless $quiet; |
|---|
| 927 | |
|---|
| 928 | my $total_channels = 0; |
|---|
| 929 | |
|---|
| 930 | my $overall_data_ok = 1; # until proven otherwise |
|---|
| 931 | |
|---|
| 932 | # iterate across each channel |
|---|
| 933 | foreach my $ch (sort keys %{$channels}) { |
|---|
| 934 | $total_channels++; |
|---|
| 935 | |
|---|
| 936 | my $data; |
|---|
| 937 | my $lastpol = ""; |
|---|
| 938 | $data->{data_ok} = 1; # unless proven otherwise |
|---|
| 939 | $data->{have} = 0; |
|---|
| 940 | $data->{missing} = 0; |
|---|
| 941 | |
|---|
| 942 | for my $slotnum (0..($policy{num_timeslots}-1)) { |
|---|
| 943 | my $bucket_start_offset = ($slotnum * $policy{timeslot_size}); |
|---|
| 944 | |
|---|
| 945 | # work out day number of when this bucket is. |
|---|
| 946 | # number from 0 onwards. (i.e. today=0). |
|---|
| 947 | # for a typical 7 day grabber this will actually mean 8 days of data (0-7) |
|---|
| 948 | # with days 0 and 7 truncated to half-days |
|---|
| 949 | my $day = int(($bucket_start_offset + $policy{first_bucket_offset}) / 86400); |
|---|
| 950 | |
|---|
| 951 | if (!defined $data->{day}->[$day]) { |
|---|
| 952 | $data->{day}->[$day]->{num} = $day; |
|---|
| 953 | $data->{day}->[$day]->{have} = 0; |
|---|
| 954 | $data->{day}->[$day]->{missing} = 0; |
|---|
| 955 | $data->{day}->[$day]->{missing_peak} = 0; |
|---|
| 956 | $data->{day}->[$day]->{missing_nonpeak} = 0; |
|---|
| 957 | $data->{day}->[$day]->{missing_other} = 0; |
|---|
| 958 | |
|---|
| 959 | $data->{day}->[$day]->{day_ok} = 1; # until proven otherwise |
|---|
| 960 | |
|---|
| 961 | # day changed, dump any 'already_missing' data |
|---|
| 962 | &dump_already_missing($data); |
|---|
| 963 | } |
|---|
| 964 | |
|---|
| 965 | # we have programming data for this bucket. great! process next bucket |
|---|
| 966 | if ((defined $channel_data->{$ch}->{timeslots}[$slotnum]) && |
|---|
| 967 | ($channel_data->{$ch}->{timeslots}[$slotnum] > 0)) { |
|---|
| 968 | |
|---|
| 969 | # if we have missing data queued up, push it now |
|---|
| 970 | &dump_already_missing($data); |
|---|
| 971 | |
|---|
| 972 | &dump_already_missing_period($data->{day}->[$day],$lastpol) |
|---|
| 973 | if ($lastpol ne ""); |
|---|
| 974 | |
|---|
| 975 | $data->{day}->[$day]->{have} += $policy{timeslot_size}; |
|---|
| 976 | $data->{have} += $policy{timeslot_size}; |
|---|
| 977 | next; |
|---|
| 978 | } |
|---|
| 979 | |
|---|
| 980 | # we don't have programming for this channel for this bucket |
|---|
| 981 | |
|---|
| 982 | # some grabbers take HOURS to run. if this bucket (missing data) is for |
|---|
| 983 | # a time period now in the past, then don't include it |
|---|
| 984 | next if (($bucket_start_offset + $policy{starttime}) < time); |
|---|
| 985 | |
|---|
| 986 | # work out the localtime of when this bucket is |
|---|
| 987 | my $bucket_seconds_offset = ($bucket_start_offset + $policy{first_bucket_offset}) % 86400; |
|---|
| 988 | |
|---|
| 989 | # store details of where we are missing data |
|---|
| 990 | if (!defined $data->{already_missing}) { |
|---|
| 991 | $data->{already_missing} = sprintf "#%d/%02d:%02d", |
|---|
| 992 | $day, |
|---|
| 993 | int($bucket_seconds_offset / 3600), |
|---|
| 994 | int(($bucket_seconds_offset % 3600) / 60); |
|---|
| 995 | } |
|---|
| 996 | $data->{already_missing_last} = $bucket_seconds_offset + $policy{timeslot_size} - 1; |
|---|
| 997 | |
|---|
| 998 | $data->{day}->[$day]->{missing} += $policy{timeslot_size}; |
|---|
| 999 | $data->{missing} += $policy{timeslot_size}; |
|---|
| 1000 | |
|---|
| 1001 | # work out what policy missing data for this bucket fits into |
|---|
| 1002 | my $pol; |
|---|
| 1003 | if (($bucket_seconds_offset >= $policy{peak_start}) && |
|---|
| 1004 | (($bucket_seconds_offset+$policy{timeslot_size}) <= $policy{peak_stop})) { |
|---|
| 1005 | $pol = "peak"; |
|---|
| 1006 | } elsif (($bucket_seconds_offset >= $policy{nonpeak_start}) && |
|---|
| 1007 | (($bucket_seconds_offset+$policy{timeslot_size}) <= $policy{nonpeak_stop})) { |
|---|
| 1008 | $pol = "nonpeak"; |
|---|
| 1009 | } else { |
|---|
| 1010 | $pol = "other"; |
|---|
| 1011 | } |
|---|
| 1012 | |
|---|
| 1013 | &dump_already_missing_period($data->{day}->[$day],$lastpol) |
|---|
| 1014 | if (($lastpol ne $pol) && ($lastpol ne "")); |
|---|
| 1015 | |
|---|
| 1016 | $lastpol = $pol; |
|---|
| 1017 | |
|---|
| 1018 | $data->{day}->[$day]->{"missing_".$pol} += $policy{timeslot_size}; |
|---|
| 1019 | |
|---|
| 1020 | $data->{day}->[$day]->{"already_missing_".$pol."_start"} = $bucket_seconds_offset |
|---|
| 1021 | if (!defined $data->{day}->[$day]->{"already_missing_".$pol."_start"}); |
|---|
| 1022 | $data->{day}->[$day]->{"already_missing_".$pol."_stop"} = $bucket_seconds_offset + $policy{timeslot_size} - 1; |
|---|
| 1023 | |
|---|
| 1024 | $data->{day}->[$day]->{day_ok} = 0 if ($data->{day}->[$day]->{missing_peak} > $policy{peak_max_missing}); |
|---|
| 1025 | $data->{day}->[$day]->{day_ok} = 0 if ($data->{day}->[$day]->{missing_nonpeak} > $policy{nonpeak_max_missing}); |
|---|
| 1026 | $data->{day}->[$day]->{day_ok} = 0 if ($data->{day}->[$day]->{missing_other} > $policy{other_max_missing}); |
|---|
| 1027 | $data->{data_ok} = 0 if ($data->{day}->[$day]->{day_ok} == 0); |
|---|
| 1028 | $overall_data_ok = 0 if ($data->{data_ok} == 0); |
|---|
| 1029 | } |
|---|
| 1030 | |
|---|
| 1031 | # finished all timeslots in this channel. |
|---|
| 1032 | # if we have missing data queued up, push it now |
|---|
| 1033 | &dump_already_missing($data); |
|---|
| 1034 | |
|---|
| 1035 | # fill in any last missing period data |
|---|
| 1036 | foreach my $day (@{($data->{day})}) { |
|---|
| 1037 | &dump_already_missing_period($day,"peak"); |
|---|
| 1038 | &dump_already_missing_period($day,"nonpeak"); |
|---|
| 1039 | &dump_already_missing_period($day,"other"); |
|---|
| 1040 | } |
|---|
| 1041 | |
|---|
| 1042 | my $statusstring = sprintf " > ch %s: %s programming: %s\n", |
|---|
| 1043 | $ch, pretty_duration($data->{have}), |
|---|
| 1044 | $data->{data_ok} ? "PASS (within thresholds)" : "FAIL, missing data over policy threshold:"; |
|---|
| 1045 | |
|---|
| 1046 | # display per-day missing data statistics |
|---|
| 1047 | foreach my $day (@{($data->{day})}) { |
|---|
| 1048 | unless ($day->{day_ok}) { |
|---|
| 1049 | $statusstring .= sprintf "\t".(strftime("%a %e %b",localtime($policy{starttime}+($day->{num}*86400)))).": "; |
|---|
| 1050 | |
|---|
| 1051 | # do we have any data for this day? |
|---|
| 1052 | $statusstring .= "peak ".join(", ",(@{($day->{missing_peak_table})})) |
|---|
| 1053 | if (($day->{missing_peak}) && ($day->{missing_peak} > $policy{peak_max_missing})); |
|---|
| 1054 | |
|---|
| 1055 | $statusstring .= sprintf "%snon-peak %s", |
|---|
| 1056 | ($day->{missing_peak} ? " / " : ""), |
|---|
| 1057 | join(", ",(@{($day->{missing_nonpeak_table})})) |
|---|
| 1058 | if (($day->{missing_nonpeak}) && ($day->{missing_nonpeak} > $policy{nonpeak_max_missing})); |
|---|
| 1059 | |
|---|
| 1060 | $statusstring .= sprintf "%sother %s", |
|---|
| 1061 | (($day->{missing_peak} + $day->{missing_nonpeak}) > 0 ? " / " : ""), |
|---|
| 1062 | join(", ",(@{($day->{missing_other_table})})) |
|---|
| 1063 | if (($day->{missing_other}) && ($day->{missing_other} > $policy{other_max_missing})); |
|---|
| 1064 | |
|---|
| 1065 | $statusstring .= "\n"; |
|---|
| 1066 | } |
|---|
| 1067 | } |
|---|
| 1068 | printf $statusstring unless $quiet; |
|---|
| 1069 | |
|---|
| 1070 | delete $channel_data->{$ch}->{analysis} if (defined $channel_data->{$ch}->{analysis}); |
|---|
| 1071 | $channel_data->{$ch}->{analysis} = $data; |
|---|
| 1072 | } |
|---|
| 1073 | |
|---|
| 1074 | printf " > OVERALL: %s\n", ($overall_data_ok ? "PASS" : "FAIL") unless $quiet; |
|---|
| 1075 | |
|---|
| 1076 | return $overall_data_ok; # return 1 for good, 0 for need more |
|---|
| 1077 | } |
|---|
| 1078 | |
|---|
| 1079 | # helper routine for filling in 'missing_all' array |
|---|
| 1080 | sub dump_already_missing |
|---|
| 1081 | { |
|---|
| 1082 | my $d = shift; |
|---|
| 1083 | if (defined $d->{already_missing}) { |
|---|
| 1084 | $d->{already_missing} .= sprintf "-%02d:%02d", |
|---|
| 1085 | int($d->{already_missing_last} / 3600), |
|---|
| 1086 | int(($d->{already_missing_last} % 3600) / 60) |
|---|
| 1087 | if (defined $d->{already_missing_last}); |
|---|
| 1088 | push(@{($d->{missing_all})}, $d->{already_missing}); |
|---|
| 1089 | delete $d->{already_missing}; |
|---|
| 1090 | delete $d->{already_missing_last}; |
|---|
| 1091 | } |
|---|
| 1092 | } |
|---|
| 1093 | |
|---|
| 1094 | # helper routine for filling in per-day missing data |
|---|
| 1095 | # specific to peak/nonpeak/other |
|---|
| 1096 | sub dump_already_missing_period |
|---|
| 1097 | { |
|---|
| 1098 | my ($d,$p) = @_; |
|---|
| 1099 | my $startvar = "already_missing_".$p."_start"; |
|---|
| 1100 | my $stopvar = "already_missing_".$p."_stop"; |
|---|
| 1101 | |
|---|
| 1102 | if (defined $d->{$startvar}) { |
|---|
| 1103 | push(@{($d->{"missing_".$p."_table"})}, |
|---|
| 1104 | sprintf "%02d:%02d-%02d:%02d", |
|---|
| 1105 | int($d->{$startvar} / 3600), |
|---|
| 1106 | int(($d->{$startvar} % 3600) / 60), |
|---|
| 1107 | int($d->{$stopvar} / 3600), |
|---|
| 1108 | int(($d->{$stopvar} % 3600) / 60)); |
|---|
| 1109 | delete $d->{$startvar}; |
|---|
| 1110 | delete $d->{$stopvar}; |
|---|
| 1111 | } |
|---|
| 1112 | } |
|---|
| 1113 | |
|---|
| 1114 | # given a duration (seconds), return it in a pretty "{days}d{hr}h{min}m" string |
|---|
| 1115 | # and indication of whether the duration is over its threshold or not |
|---|
| 1116 | sub pretty_duration |
|---|
| 1117 | { |
|---|
| 1118 | my ($d,$crit) = @_; |
|---|
| 1119 | my $s = ""; |
|---|
| 1120 | $s .= sprintf "%dd",int($d / (60*60*24)) if ($d >= (60*60*24)); |
|---|
| 1121 | $s .= sprintf "%dh",int(($d % (60*60*24)) / (60*60)) if (($d % (60*60*24)) >= (60*60)); |
|---|
| 1122 | $s .= sprintf "%dm",int(($d % (60*60)) / 60) if (($d % (60*60)) >= 60); |
|---|
| 1123 | $s .= "no" if ($s eq ""); |
|---|
| 1124 | |
|---|
| 1125 | if (defined $crit) { |
|---|
| 1126 | $s .= "[!]" if ($d > $crit); |
|---|
| 1127 | } |
|---|
| 1128 | return $s; |
|---|
| 1129 | } |
|---|
| 1130 | |
|---|
| 1131 | # work out date range we are expecting data to be in |
|---|
| 1132 | sub calc_date_range |
|---|
| 1133 | { |
|---|
| 1134 | # work out GMT offset - we only do this once |
|---|
| 1135 | if (!$gmt_offset) { |
|---|
| 1136 | # work out our gmt offset |
|---|
| 1137 | my @l = localtime(43200), my @g = gmtime(43200); |
|---|
| 1138 | $gmt_offset = (($l[2] - $g[2])*(60*60)) + (($l[1] - $g[1])*60); |
|---|
| 1139 | } |
|---|
| 1140 | |
|---|
| 1141 | $policy{starttime} = time; |
|---|
| 1142 | |
|---|
| 1143 | # set endtime as per $days less 1 day + hours left today |
|---|
| 1144 | $policy{endtime} = $policy{starttime} + ((60*60*24)*($days-1)) + (86400 - (($policy{starttime} + $gmt_offset) % 86400)); |
|---|
| 1145 | |
|---|
| 1146 | # normalize starttime to beginning of next bucket |
|---|
| 1147 | $policy{starttime} += ($policy{timeslot_size} - ($policy{starttime} % $policy{timeslot_size})); |
|---|
| 1148 | |
|---|
| 1149 | # work out how many seconds into a day our first bucket starts |
|---|
| 1150 | $policy{first_bucket_offset} = ($policy{starttime} + $gmt_offset) % 86400; |
|---|
| 1151 | |
|---|
| 1152 | # normalize endtime to end of previous bucket |
|---|
| 1153 | $policy{endtime} -= ($policy{endtime} % $policy{timeslot_size}); |
|---|
| 1154 | |
|---|
| 1155 | # if we are working with an --offset, apply it now. |
|---|
| 1156 | $policy{starttime} += (86400 * $opt->{offset}) if ($opt->{offset}); |
|---|
| 1157 | |
|---|
| 1158 | # work out number of buckets |
|---|
| 1159 | $policy{num_timeslots} = ($policy{endtime} - $policy{starttime}) / $policy{timeslot_size}; |
|---|
| 1160 | } |
|---|
| 1161 | |
|---|
| 1162 | |
|---|
| 1163 | # strptime type date parsing - BUT - if no timezone is present, treat time as being in localtime |
|---|
| 1164 | # rather than the various other perl implementation which treat it as being in UTC/GMT |
|---|
| 1165 | sub parse_xmltv_date |
|---|
| 1166 | { |
|---|
| 1167 | my $datestring = shift; |
|---|
| 1168 | my @t; # 0=sec,1=min,2=hour,3=day,4=month,5=year,6=wday,7=yday,8=isdst |
|---|
| 1169 | my $tz_offset = 0; |
|---|
| 1170 | |
|---|
| 1171 | if ($datestring =~ /^(\d{4})(\d{2})(\d{2})(\d{2})(\d{2})/) { |
|---|
| 1172 | ($t[5],$t[4],$t[3],$t[2],$t[1],$t[0]) = (int($1)-1900,int($2)-1,int($3),int($4),int($5),0); |
|---|
| 1173 | ($t[6],$t[7],$t[8]) = (-1,-1,-1); |
|---|
| 1174 | |
|---|
| 1175 | # if input data has a timezone offset, then offset by that |
|---|
| 1176 | if ($datestring =~ /\+(\d{2})(\d{2})/) { |
|---|
| 1177 | $tz_offset = $gmt_offset - (($1*(60*60)) + ($2*60)); |
|---|
| 1178 | } elsif ($datestring =~ /\-(\d{2})(\d{2})/) { |
|---|
| 1179 | $tz_offset = $gmt_offset + (($1*(60*60)) + ($2*60)); |
|---|
| 1180 | } |
|---|
| 1181 | |
|---|
| 1182 | my $e = mktime(@t); |
|---|
| 1183 | return ($e+$tz_offset) if ($e > 1); |
|---|
| 1184 | } |
|---|
| 1185 | return undef; |
|---|
| 1186 | } |
|---|
| 1187 | |
|---|
| 1188 | # ----------------------------------------- |
|---|
| 1189 | # Subs: Reconciling data |
|---|
| 1190 | # ----------------------------------------- |
|---|
| 1191 | |
|---|
| 1192 | # for all the data we have, try to pick the best bits! |
|---|
| 1193 | sub reconcile_data |
|---|
| 1194 | { |
|---|
| 1195 | printf "\nReconciling data:\n\n"; |
|---|
| 1196 | |
|---|
| 1197 | my $num_grabbers = 0; |
|---|
| 1198 | my $input_files = ""; |
|---|
| 1199 | my @input_file_list; |
|---|
| 1200 | |
|---|
| 1201 | printf "Preferred title preferences from '$pref_title_source'\n" |
|---|
| 1202 | if ((defined $pref_title_source) && |
|---|
| 1203 | ($plugin_data->{$pref_title_source}) && |
|---|
| 1204 | ($plugin_data->{$pref_title_source}->{valid})); |
|---|
| 1205 | |
|---|
| 1206 | printf "Preference for whose data we prefer as follows:\n"; |
|---|
| 1207 | foreach my $proggy (sort { $components->{$b}->{config}->{quality} <=> $components->{$a}->{config}->{quality} } query_grabbers()) { |
|---|
| 1208 | if ((!$components->{$proggy}->{disabled}) && ($plugin_data->{$proggy}) && ($plugin_data->{$proggy}->{valid})) { |
|---|
| 1209 | $num_grabbers++; |
|---|
| 1210 | printf " %d. %s (%s)\n", $num_grabbers, $proggy, $plugin_data->{$proggy}->{output_filename}; |
|---|
| 1211 | |
|---|
| 1212 | $input_files .= $plugin_data->{$proggy}->{output_filename}." "; |
|---|
| 1213 | push(@input_file_list,$plugin_data->{$proggy}->{output_filename}); |
|---|
| 1214 | } |
|---|
| 1215 | } |
|---|
| 1216 | |
|---|
| 1217 | if ($num_grabbers == 0) { |
|---|
| 1218 | die "Nothing to reconcile! There is no valid grabber data!\n"; |
|---|
| 1219 | } |
|---|
| 1220 | |
|---|
| 1221 | foreach my $reconciler (sort { $components->{$a} <=> $components->{$b} } query_reconcilers()) { |
|---|
| 1222 | next if ($components->{$reconciler}->{disabled}); |
|---|
| 1223 | next if (!$components->{$reconciler}->{ready}); |
|---|
| 1224 | |
|---|
| 1225 | $reconciler_found_all_data = &call_data_processor("reconciler",$reconciler,$input_files); |
|---|
| 1226 | |
|---|
| 1227 | if ((!$reconciler_found_all_data) && ($grabber_found_all_data)) { |
|---|
| 1228 | # urgh. this reconciler did a bad bad thing ... |
|---|
| 1229 | printf "SHEPHERD: XML data from reconciler $reconciler appears bogus, will try to use another reconciler\n"; |
|---|
| 1230 | } else { |
|---|
| 1231 | printf "SHEPHERD: Data from reconciler $reconciler looks good\n"; |
|---|
| 1232 | $input_postprocess_file = $plugin_data->{$reconciler}->{output_filename}; |
|---|
| 1233 | } |
|---|
| 1234 | |
|---|
| 1235 | last if ($input_postprocess_file ne ""); |
|---|
| 1236 | } |
|---|
| 1237 | |
|---|
| 1238 | if ($input_postprocess_file eq "") { |
|---|
| 1239 | # no reconcilers worked!! |
|---|
| 1240 | printf "SHEPHERD: WARNING: No reconcilers seemed to work! Falling back to concatenating the data together!\n"; |
|---|
| 1241 | |
|---|
| 1242 | my %w_args = (); |
|---|
| 1243 | $input_postprocess_file = "$CWD/input_preprocess.xmltv"; |
|---|
| 1244 | my $fh = new IO::File ">$input_postprocess_file" || die "could not open $input_postprocess_file for writing: $!\n"; |
|---|
| 1245 | %w_args = (OUTPUT => $fh); |
|---|
| 1246 | XMLTV::catfiles(\%w_args, @input_file_list); |
|---|
| 1247 | } |
|---|
| 1248 | } |
|---|
| 1249 | |
|---|
| 1250 | |
|---|
| 1251 | # ----------------------------------------- |
|---|
| 1252 | # Subs: Postprocessing |
|---|
| 1253 | # ----------------------------------------- |
|---|
| 1254 | |
|---|
| 1255 | sub postprocess_data |
|---|
| 1256 | { |
|---|
| 1257 | # for our first postprocessor, we feed it ALL of the XMLTV files we have |
|---|
| 1258 | # as each postprocessor runs, we feed in the output from the previous one |
|---|
| 1259 | # Shepherd checks the "completeness" of the data that comes out of a postprocessor & automatically |
|---|
| 1260 | # reverts back to the previous postprocessor if it was shown to be bad |
|---|
| 1261 | |
|---|
| 1262 | # first time around: feed in reconciled data ($input_postprocess_file) |
|---|
| 1263 | |
|---|
| 1264 | printf "\nPostprocessing stage:\n"; |
|---|
| 1265 | |
|---|
| 1266 | foreach my $postprocessor (sort { $components->{$a} <=> $components->{$b} } query_postprocessors()) { |
|---|
| 1267 | next if ($components->{$postprocessor}->{disabled}); |
|---|
| 1268 | next if (!$components->{$postprocessor}->{ready}); |
|---|
| 1269 | |
|---|
| 1270 | my $found_all_data = call_data_processor("postprocessor",$postprocessor,$input_postprocess_file); |
|---|
| 1271 | |
|---|
| 1272 | if ($found_all_data) { |
|---|
| 1273 | # accept what this postprocessor did to our output ... |
|---|
| 1274 | printf "SHEPHERD: accepting output from postprocessor $postprocessor, feeding it into next stage\n"; |
|---|
| 1275 | $input_postprocess_file = $plugin_data->{$postprocessor}->{output_filename}; |
|---|
| 1276 | delete $components->{$postprocessor}->{conescutive_failures} if (defined $components->{$postprocessor}->{conescutive_failures}); |
|---|
| 1277 | next; |
|---|
| 1278 | } |
|---|
| 1279 | |
|---|
| 1280 | # urgh. this postprocessor did a bad bad thing ... |
|---|
| 1281 | printf "SHEPHERD: XML data from postprocessor $postprocessor rejected, using XML from previous stage\n"; |
|---|
| 1282 | |
|---|
| 1283 | if (defined $components->{$postprocessor}->{conescutive_failures}) { |
|---|
| 1284 | $components->{$postprocessor}->{conescutive_failures}++; |
|---|
| 1285 | } else { |
|---|
| 1286 | $components->{$postprocessor}->{conescutive_failures} = 1; |
|---|
| 1287 | } |
|---|
| 1288 | printf "SHEPHERD: Postprocessor \"%s\" has now failed %d times in a row. %d more and it will be automatically disabled.\n", |
|---|
| 1289 | $postprocessor, |
|---|
| 1290 | $components->{$postprocessor}->{conescutive_failures}, |
|---|
| 1291 | ($policy{postprocessor_disable_failure_threshold} - $components->{$postprocessor}->{conescutive_failures}); |
|---|
| 1292 | |
|---|
| 1293 | if ($components->{$postprocessor}->{conescutive_failures} >= $policy{postprocessor_disable_failure_threshold}) { |
|---|
| 1294 | printf "SHEPHERD: Disabling Postprocessor \"$postprocessor\".\n"; |
|---|
| 1295 | $components->{$postprocessor}->{disabled} = 1; |
|---|
| 1296 | } |
|---|
| 1297 | } |
|---|
| 1298 | } |
|---|
| 1299 | |
|---|
| 1300 | |
|---|
| 1301 | # ----------------------------------------- |
|---|
| 1302 | # Subs: Postprocessing/Reconciler helpers |
|---|
| 1303 | # ----------------------------------------- |
|---|
| 1304 | |
|---|
| 1305 | sub call_data_processor |
|---|
| 1306 | { |
|---|
| 1307 | my ($data_processor_type, $data_processor_name, $input_files) = @_; |
|---|
| 1308 | |
|---|
| 1309 | $components->{$data_processor_name}->{lastdata} = time; |
|---|
| 1310 | $components->{$data_processor_name}->{laststatus} = "unknown"; |
|---|
| 1311 | |
|---|
| 1312 | printf "\nSHEPHERD: Using %s: %s\n",$data_processor_type,$data_processor_name; |
|---|
| 1313 | |
|---|
| 1314 | my $output = sprintf "%s/%ss/%s/output.xmltv",$CWD,$data_processor_type,$data_processor_name; |
|---|
| 1315 | my $comm = sprintf "%s/%ss/%s/%s",$CWD,$data_processor_type,$data_processor_name,$data_processor_name; |
|---|
| 1316 | $comm .= " --region $region" . |
|---|
| 1317 | " --channels_file $channels_file" . |
|---|
| 1318 | " --output $output"; |
|---|
| 1319 | $comm .= " --days $days" if ($days); |
|---|
| 1320 | $comm .= " --offset $opt->{offset}" if ($opt->{offset}); |
|---|
| 1321 | $comm .= " --debug" if ($debug); |
|---|
| 1322 | $comm .= " @ARGV" if (@ARGV); |
|---|
| 1323 | |
|---|
| 1324 | $comm .= " --preftitle ".$plugin_data->{$pref_title_source}->{output_filename} |
|---|
| 1325 | if (($data_processor_type eq "reconciler") && |
|---|
| 1326 | (defined $pref_title_source) && |
|---|
| 1327 | ($plugin_data->{$pref_title_source}) && |
|---|
| 1328 | ($plugin_data->{$pref_title_source}->{valid})); |
|---|
| 1329 | |
|---|
| 1330 | $comm .= " $input_files"; |
|---|
| 1331 | printf "SHEPHERD: Excuting command: $comm\n"; |
|---|
| 1332 | |
|---|
| 1333 | my $dir = sprintf "%s/%ss/%s/",$CWD,$data_processor_type,$data_processor_name; |
|---|
| 1334 | chdir $dir; |
|---|
| 1335 | my $retval = call_prog($comm); |
|---|
| 1336 | chdir $CWD; |
|---|
| 1337 | |
|---|
| 1338 | if ($retval != 0) { |
|---|
| 1339 | printf "$data_processor_type returned with non-zero return code $retval: assuming it failed.\n"; |
|---|
| 1340 | return 0; |
|---|
| 1341 | } |
|---|
| 1342 | |
|---|
| 1343 | # |
|---|
| 1344 | # soak up the data we just collected and check it |
|---|
| 1345 | # YES - these are the SAME routines we used in the previous 'grabber' phase |
|---|
| 1346 | # but the difference here is that we clear out our 'channel_data' beforehand |
|---|
| 1347 | # so we can independently analyze the impact of this postprocessor. |
|---|
| 1348 | # if it clearly returns bad data, don't use that data (go back one step) and |
|---|
| 1349 | # flag the postprocessor as having failed. after 3 consecutive failures, disable it |
|---|
| 1350 | # |
|---|
| 1351 | |
|---|
| 1352 | # clear out channel_data |
|---|
| 1353 | foreach my $ch (keys %{$channels}) { |
|---|
| 1354 | delete $channel_data->{$ch}; |
|---|
| 1355 | } |
|---|
| 1356 | |
|---|
| 1357 | # process and analyze it! |
|---|
| 1358 | &soak_up_data($data_processor_name, $output, $data_processor_type); |
|---|
| 1359 | my $have_all_data = &analyze_plugin_data("$data_processor_type $data_processor_name"); |
|---|
| 1360 | |
|---|
| 1361 | $components->{$data_processor_name}->{laststatus} = $plugin_data->{$data_processor_name}->{laststatus}; |
|---|
| 1362 | |
|---|
| 1363 | return $have_all_data; |
|---|
| 1364 | } |
|---|
| 1365 | |
|---|
| 1366 | |
|---|
| 1367 | sub output_data |
|---|
| 1368 | { |
|---|
| 1369 | # $input_postprocess_file contains our final output |
|---|
| 1370 | # send it to whereever --output told us to! |
|---|
| 1371 | |
|---|
| 1372 | my $output_filename = "$CWD/output.xmltv"; |
|---|
| 1373 | $output_filename = $opt->{output} if ($opt->{output}); |
|---|
| 1374 | |
|---|
| 1375 | my %writer_args = ( encoding => 'ISO-8859-1' ); |
|---|
| 1376 | my $fh = new IO::File(">$output_filename") || |
|---|
| 1377 | die "can't open $output_filename for writing: $!"; |
|---|
| 1378 | $writer_args{OUTPUT} = $fh; |
|---|
| 1379 | |
|---|
| 1380 | $writer = new XMLTV::Writer(%writer_args); |
|---|
| 1381 | $writer->start( { |
|---|
| 1382 | 'source-info-name' => "$progname v".$components->{$progname}->{ver}, |
|---|
| 1383 | 'generator-info-name' => |
|---|
| 1384 | $progname."(v".$components->{$progname}->{ver}.") + ". |
|---|
| 1385 | $myprogname."(v".$version.")".$components_used} ); |
|---|
| 1386 | |
|---|
| 1387 | XMLTV::parsefiles_callback(undef, undef, \&output_data_channel_cb, |
|---|
| 1388 | \&output_data_programme_cb, $input_postprocess_file); |
|---|
| 1389 | $writer->end(); |
|---|
| 1390 | |
|---|
| 1391 | printf "Final output stored in $output_filename.\n"; |
|---|
| 1392 | } |
|---|
| 1393 | |
|---|
| 1394 | sub output_data_channel_cb |
|---|
| 1395 | { |
|---|
| 1396 | my $c = shift; |
|---|
| 1397 | $writer->write_channel($c); |
|---|
| 1398 | } |
|---|
| 1399 | |
|---|
| 1400 | sub output_data_programme_cb |
|---|
| 1401 | { |
|---|
| 1402 | my $prog=shift; |
|---|
| 1403 | $writer->write_programme($prog); |
|---|
| 1404 | } |
|---|
| 1405 | |
|---|
| 1406 | # ----------------------------------------- |
|---|
| 1407 | # Subs: Utilities |
|---|
| 1408 | # ----------------------------------------- |
|---|
| 1409 | # |
|---|
| 1410 | |
|---|
| 1411 | sub query_grabbers |
|---|
| 1412 | { |
|---|
| 1413 | my ($conf, $val) = @_; |
|---|
| 1414 | return query_component_type('grabber',$conf,$val); |
|---|
| 1415 | } |
|---|
| 1416 | |
|---|
| 1417 | sub query_reconcilers |
|---|
| 1418 | { |
|---|
| 1419 | return query_component_type('reconciler'); |
|---|
| 1420 | } |
|---|
| 1421 | |
|---|
| 1422 | sub query_postprocessors |
|---|
| 1423 | { |
|---|
| 1424 | return query_component_type('postprocessor'); |
|---|
| 1425 | } |
|---|
| 1426 | |
|---|
| 1427 | sub query_component_type |
|---|
| 1428 | { |
|---|
| 1429 | my ($progtype,$conf,$val) = @_; |
|---|
| 1430 | |
|---|
| 1431 | my @ret = (); |
|---|
| 1432 | foreach (keys %$components) |
|---|
| 1433 | { |
|---|
| 1434 | if ($components->{$_}->{type} eq $progtype) { |
|---|
| 1435 | if (defined $conf) { |
|---|
| 1436 | push (@ret, $_) if (query_config($_,$conf) eq $val); |
|---|
| 1437 | } else { |
|---|
| 1438 | push (@ret, $_); |
|---|
| 1439 | } |
|---|
| 1440 | } |
|---|
| 1441 | } |
|---|
| 1442 | return @ret; |
|---|
| 1443 | } |
|---|
| 1444 | |
|---|
| 1445 | sub query_name |
|---|
| 1446 | { |
|---|
| 1447 | my $str = shift; |
|---|
| 1448 | if ($str =~ /(.*) \[cache\]/) |
|---|
| 1449 | { |
|---|
| 1450 | return $1; |
|---|
| 1451 | } |
|---|
| 1452 | return $str; |
|---|
| 1453 | } |
|---|
| 1454 | |
|---|
| 1455 | sub query_config |
|---|
| 1456 | { |
|---|
| 1457 | my ($grabber, $key) = @_; |
|---|
| 1458 | |
|---|
| 1459 | $grabber = query_name($grabber); |
|---|
| 1460 | return undef unless ($components->{$grabber}); |
|---|
| 1461 | return $components->{$grabber}->{config}->{$key}; |
|---|
| 1462 | } |
|---|
| 1463 | |
|---|
| 1464 | sub call_prog |
|---|
| 1465 | { |
|---|
| 1466 | my $prog = shift; |
|---|
| 1467 | if (!(open(PROG,"$prog|"))) { |
|---|
| 1468 | printf "warning: couldn't exec \"$prog\": $!\n"; |
|---|
| 1469 | return -1; |
|---|
| 1470 | } |
|---|
| 1471 | while(<PROG>) { |
|---|
| 1472 | print $_; |
|---|
| 1473 | } |
|---|
| 1474 | close(PROG); |
|---|
| 1475 | |
|---|
| 1476 | if ($? == -1) { |
|---|
| 1477 | printf "Failed to execute prog: $!\n"; |
|---|
| 1478 | return -1; |
|---|
| 1479 | } elsif ($? & 127) { |
|---|
| 1480 | printf "prog died with signal %d, %s coredump\n", |
|---|
| 1481 | ($? & 127), (($? & 128) ? "with" : "without"); |
|---|
| 1482 | return $?; |
|---|
| 1483 | } else { |
|---|
| 1484 | printf "prog exited with value %d\n", ($? >> 8) if ($debug or $?); |
|---|
| 1485 | return ($? >> 8); |
|---|
| 1486 | } |
|---|
| 1487 | } |
|---|
| 1488 | |
|---|
| 1489 | # ----------------------------------------- |
|---|
| 1490 | # Subs: Setup |
|---|
| 1491 | # ----------------------------------------- |
|---|
| 1492 | |
|---|
| 1493 | sub read_config_file |
|---|
| 1494 | { |
|---|
| 1495 | read_file($config_file, 'configuration'); |
|---|
| 1496 | } |
|---|
| 1497 | |
|---|
| 1498 | sub read_channels_file |
|---|
| 1499 | { |
|---|
| 1500 | read_file($channels_file, 'channels'); |
|---|
| 1501 | } |
|---|
| 1502 | |
|---|
| 1503 | sub read_file |
|---|
| 1504 | { |
|---|
| 1505 | my $fn = shift; |
|---|
| 1506 | my $name = shift; |
|---|
| 1507 | |
|---|
| 1508 | print "Reading $name file: $fn\n"; |
|---|
| 1509 | unless (-r $fn) |
|---|
| 1510 | { |
|---|
| 1511 | unless ($opt->{configure}) |
|---|
| 1512 | { |
|---|
| 1513 | print "\nNo $name file found.\n" . |
|---|
| 1514 | ucfirst($progname) . " must be configured: " . |
|---|
| 1515 | "configuring now.\n\n"; |
|---|
| 1516 | $opt->{'configure'} = 1; |
|---|
| 1517 | } |
|---|
| 1518 | return; |
|---|
| 1519 | } |
|---|
| 1520 | local (@ARGV, $/) = ($fn); |
|---|
| 1521 | no warnings 'all'; |
|---|
| 1522 | eval <>; |
|---|
| 1523 | if ($@ and !$opt->{configure}) |
|---|
| 1524 | { |
|---|
| 1525 | warn "\nERROR in $name file! Details:\n$@"; |
|---|
| 1526 | print "You may wish to CTRL-C and fix this.\n\nContinuing anyway in:"; |
|---|
| 1527 | foreach (1 .. 5) |
|---|
| 1528 | { |
|---|
| 1529 | print " " . (6 - $_); |
|---|
| 1530 | sleep 1; |
|---|
| 1531 | } |
|---|
| 1532 | print "\n"; |
|---|
| 1533 | } |
|---|
| 1534 | } |
|---|
| 1535 | |
|---|
| 1536 | sub write_config_file |
|---|
| 1537 | { |
|---|
| 1538 | write_file($config_file, 'configuration', |
|---|
| 1539 | [$region, $pref_title_source, $mirror_site, $components ], |
|---|
| 1540 | ["region", "pref_title_source", "mirror_site", "components" ]); |
|---|
| 1541 | } |
|---|
| 1542 | |
|---|
| 1543 | sub write_channels_file |
|---|
| 1544 | { |
|---|
| 1545 | write_file($channels_file, 'channels', |
|---|
| 1546 | [ $channels, $opt_channels ], |
|---|
| 1547 | [ 'channels', 'opt_channels' ]); |
|---|
| 1548 | } |
|---|
| 1549 | |
|---|
| 1550 | sub write_file |
|---|
| 1551 | { |
|---|
| 1552 | my ($fn, $name, $vars, $varnames) = @_; |
|---|
| 1553 | open (FN, ">$fn") or die "Can't write to $name file $fn: $!"; |
|---|
| 1554 | print FN Data::Dumper->Dump($vars, $varnames); |
|---|
| 1555 | close FN; |
|---|
| 1556 | print "Wrote $name file $fn.\n" if ($debug); |
|---|
| 1557 | } |
|---|
| 1558 | |
|---|
| 1559 | sub get_command_line_options |
|---|
| 1560 | { |
|---|
| 1561 | GetOptions($opt, qw( help |
|---|
| 1562 | dontcallgrabbers |
|---|
| 1563 | version |
|---|
| 1564 | debug |
|---|
| 1565 | days=i |
|---|
| 1566 | offset=i |
|---|
| 1567 | output=s |
|---|
| 1568 | randomize |
|---|
| 1569 | )); |
|---|
| 1570 | $debug = $opt->{debug}; |
|---|
| 1571 | $days = $opt->{days} if ($opt->{days}); |
|---|
| 1572 | } |
|---|
| 1573 | |
|---|
| 1574 | # ----------------------------------------- |
|---|
| 1575 | # Subs: override handlers for standard perl. |
|---|
| 1576 | # ----------------------------------------- |
|---|
| 1577 | |
|---|
| 1578 | # ugly hack. please don't try this at home kids! |
|---|
| 1579 | sub my_die { |
|---|
| 1580 | my ($arg,@rest) = @_; |
|---|
| 1581 | my ($pack,$file,$line,$sub) = caller(0); |
|---|
| 1582 | |
|---|
| 1583 | # check if we are in an eval() |
|---|
| 1584 | if ($^S) { |
|---|
| 1585 | printf " shepherd caught a die() within eval{} from file $file line $line\n"; |
|---|
| 1586 | } else { |
|---|
| 1587 | printf "\nDIE: line %d in file %s\n",$line,$file; |
|---|
| 1588 | if ($arg) { |
|---|
| 1589 | CORE::die($arg,@rest); |
|---|
| 1590 | } else { |
|---|
| 1591 | CORE::die(join("",@rest)); |
|---|
| 1592 | } |
|---|
| 1593 | } |
|---|
| 1594 | } |
|---|
