Advertisement
Guest User

imdb_olympia.xml

a guest
Nov 14th, 2012
320
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
XML 21.73 KB | None | 0 0
  1. <?xml version="1.0" encoding="utf-8"?><scraper framework="1,1" date="2011-03-26" name="imdb_olympia_v1.5" content="movies" thumb="imdb.png" language="en">
  2.     <GetSettings dest="3">
  3.         <RegExp input="$$5" output="&lt;settings&gt;\1&lt;/settings&gt;" dest="3">
  4.             <RegExp input="$$1" output="&lt;setting label=&quot;Title Language From&quot; type=&quot;labelenum&quot; values=&quot;Keep Original|USA / International|Argentina|Austria|Belgium|Brazil|Bulgaria|Canada|China|Colombia|Chile|Croatia|Czech Republic|Denmark|Finland|France|Germany|Greece|Hong Kong|Hungary|Iceland|India|Israel|Italy|Japan|Mexico|Netherlands|Norway|Pakistan|Poland|Portugal|Romania|Russia|Serbia|Singapore|Slovenia|Spain|Sweden|Switzerland|Thailand|Turkey|Uruguay|Venezuela&quot; id=&quot;akatitles&quot; default=&quot;Keep Original&quot;&gt;&lt;/setting&gt;" dest="5">
  5.                 <expression />
  6.             </RegExp>
  7.             <RegExp input="$$1" output="&lt;setting label=&quot;HD-Trailers.net Quality&quot; type=&quot;labelenum&quot; values=&quot;none|480p|720p|1080p&quot; id=&quot;TrailerQ&quot; default=&quot;480p&quot;&gt;&lt;/setting&gt;" dest="5+">
  8.                 <expression />
  9.             </RegExp>
  10.             <RegExp input="$$1" output="&lt;setting label=&quot;Get Plot from TMDb&quot; type=&quot;bool&quot; id=&quot;tmdbplot&quot; default=&quot;false&quot;&gt;&lt;/setting&gt;" dest="5+">
  11.                 <expression />
  12.             </RegExp>
  13.             <RegExp input="$$1" output="&lt;setting label=&quot;TMDb Plot Language&quot; type=&quot;labelenum&quot; values=&quot;da|fi|nl|de|it|es|fr|pl|hu|el|tr|ru|he|ja|pt|zh|cs|sl|hr|ko|en|sv|no&quot; id=&quot;plotlanguage&quot; default=&quot;en&quot;&gt;&lt;/setting&gt;" dest="5+">
  14.                 <expression />
  15.             </RegExp>
  16.             <expression noclean="1" />
  17.         </RegExp>
  18.     </GetSettings>
  19.     <NfoUrl dest="3">
  20.         <RegExp input="$$1" output="&lt;url&gt;http://akas.imdb.com/title/tt\1/&lt;/url&gt;&lt;id&gt;tt\1&lt;/id&gt;" dest="3">
  21.             <expression clear="yes" noclean="1">imdb....?/Title\?([0-9]*)</expression>
  22.         </RegExp>
  23.         <RegExp input="$$1" output="&lt;url&gt;http://akas.imdb.com/title/tt\1/&lt;/url&gt;&lt;id&gt;tt\1&lt;/id&gt;" dest="3+">
  24.             <expression noclean="1">imdb....?/title/tt([0-9]*)</expression>
  25.         </RegExp>
  26.     </NfoUrl>
  27.     <CreateSearchUrl dest="3" SearchStringEncoding="iso-8859-1">
  28.         <RegExp input="$$1" output="&lt;url&gt;http://akas.imdb.com/find?s=tt;q=\1$$4&lt;/url&gt;" dest="3">
  29.             <RegExp input="$$2" output="%20(\1)" dest="4">
  30.                 <expression clear="yes">(.+)</expression>
  31.             </RegExp>
  32.             <expression noclean="1"/>
  33.         </RegExp>
  34.     </CreateSearchUrl>
  35.     <GetSearchResults dest="8">
  36.         <RegExp input="$$5" output="&lt;?xml version=&quot;1.0&quot; encoding=&quot;iso-8859-1&quot; standalone=&quot;yes&quot;?&gt;&lt;results&gt;\1&lt;/results&gt;" dest="8">
  37.             <RegExp input="$$1" output="\1" dest="7">
  38.                 <expression clear="yes">/title/([t0-9]*)/(combined|faq|releaseinfo|vote)</expression>
  39.             </RegExp>
  40.             <RegExp input="$$1" output="&lt;entity&gt;&lt;title&gt;\2&lt;/title&gt;&lt;year&gt;\3&lt;/year&gt;&lt;url cache=&quot;$$7-main.html&quot;&gt;http://akas.imdb.com/title/$$7/&lt;/url&gt;&lt;id&gt;$$7&lt;/id&gt;&lt;/entity&gt;" dest="5">
  41.                 <expression clear="yes" noclean="2">&lt;meta name=&quot;title&quot; content=&quot;(IMDb - )*(?:&amp;#x22;)?([^&quot;]*?)(?:&amp;#x22;)? \([^\(]*?([0-9]{4})\)</expression>
  42.             </RegExp>
  43.             <RegExp input="$$1" output="\1" dest="4">
  44.                 <expression noclean="1">(&gt;&lt;a href=&quot;/title.*)</expression>
  45.             </RegExp>
  46.             <RegExp input="$$4" output="&lt;entity&gt;&lt;title&gt;\2&lt;/title&gt;&lt;year&gt;\3&lt;/year&gt;&lt;url cache=&quot;\1-main.html&quot;&gt;http://akas.imdb.com/title/\1/&lt;/url&gt;&lt;id&gt;\1&lt;/id&gt;&lt;/entity&gt;" dest="5+">
  47.                 <expression repeat="yes" noclean="1,2">&gt;&lt;a href=&quot;/title/([t0-9]*)/[^&gt;]*&gt;(?:&amp;#x22;)?([^&lt;]*?)(?:&amp;#x22;)?&lt;/a&gt; *\([^\(]*?([0-9]{4})</expression>
  48.             </RegExp>
  49.             <expression clear="yes" noclean="1"/>
  50.         </RegExp>
  51.     </GetSearchResults>
  52.     <GetDetails dest="3">
  53.         <RegExp input="$$5" output="&lt;details&gt;\1&lt;/details&gt;" dest="3">
  54.             <RegExp input="$$2" output="&lt;id&gt;\1&lt;/id&gt;" dest="5">
  55.                 <expression/>
  56.             </RegExp>
  57.             <RegExp input="$$1" output="\2" dest="18">
  58.                 <expression>&lt;meta name=&quot;title&quot; content=&quot;(IMDb - )*(?:&amp;#x22;)?([^&quot;]*?)(?:&amp;#x22;)? \([^\(]*?([0-9]{4})\)</expression>
  59.             </RegExp>
  60.             <RegExp input="$$18" output="&lt;sorttitle&gt;\1&lt;/sorttitle&gt;&lt;originaltitle&gt;\1&lt;/originaltitle&gt;" dest="5+">
  61.                 <expression/>
  62.             </RegExp>
  63.             <RegExp input="$$17" output="\1" dest="19">
  64.                 <RegExp input="$$18" output="\1-" dest="17">
  65.                     <expression repeat="yes" trim="1">([\w]+)</expression>
  66.                 </RegExp>
  67.                 <expression noclean="1" />
  68.             </RegExp>
  69.             <RegExp input="$$2" output="&lt;url cache=&quot;$$2-combined.html&quot; function=&quot;ParseIMDBAKATitles&quot;&gt;http://akas.imdb.com/title/$$2/combined&lt;/url&gt;" dest="5+">
  70.                 <expression noclean="1" />
  71.             </RegExp>
  72.             <RegExp input="$$1" output="&lt;year&gt;\2&lt;/year&gt;" dest="5+">
  73.                 <expression>&lt;meta name=&quot;title&quot; content=&quot;([^&quot;]*?) \([^\(]*?([0-9]{4})\)</expression>
  74.             </RegExp>
  75.             <RegExp input="$$1" output="&lt;top250&gt;\1&lt;/top250&gt;" dest="5+">
  76.                 <expression>Top 250 #([0-9]*)&lt;/</expression>
  77.             </RegExp>
  78.             <RegExp input="$$1" output="&lt;mpaa&gt;\1&lt;/mpaa&gt;" dest="5+">
  79.                 <expression>MPAA&lt;/a&gt;\)&lt;/h4&gt;\n?&lt;span itemprop=&quot;contentRating&quot;&gt;([^&lt;]*)</expression>
  80.             </RegExp>
  81.             <RegExp input="$$1" output="&lt;tagline&gt;\1&lt;/tagline&gt;" dest="5+">
  82.                 <expression>&lt;h4[^&gt;]*?&gt;Taglines:&lt;/h4&gt;\n?([^&lt;]*)</expression>
  83.             </RegExp>
  84.             <RegExp input="$$1" output="&lt;runtime&gt;\1&lt;/runtime&gt;" dest="5+">
  85.                 <expression trim="1">&lt;h4[^&gt;]*?&gt;Runtime:&lt;/h4&gt;[^0-9]*([0-9]+)</expression>
  86.             </RegExp>
  87.             <RegExp input="$$1" output="&lt;rating&gt;\1&lt;/rating&gt;&lt;votes&gt;\2&lt;/votes&gt;" dest="5+">
  88.                 <expression>itemprop=&quot;ratingValue&quot;&gt;([0-9.]+).*?&gt;.*?title=&quot;([0-9,]+)\s</expression>
  89.             </RegExp>
  90.             <RegExp input="$$1" output="&lt;genre&gt;\1&lt;/genre&gt;" dest="5+">
  91.                 <expression repeat="yes">&lt;a href="/genre/([^"]+)</expression>
  92.             </RegExp>
  93.             <RegExp input="$$1" output="&lt;country&gt;\1&lt;/country&gt;" dest="5+">
  94.                 <expression repeat="yes">&lt;a href="/country/[^&gt;]+&gt;([^&lt;]+)</expression>
  95.             </RegExp>
  96.             <RegExp input="$$1" output="&lt;studio&gt;\1&lt;/studio&gt;" dest="5+">
  97.                 <expression>"/company/[^&gt;]+&gt;([^&lt;]+)&lt;/a&gt;</expression>
  98.             </RegExp>
  99.             <RegExp input="$$1" output="&lt;plot&gt;\1&lt;/plot&gt;" dest="5+">
  100.                 <expression trim="1">&lt;h2&gt;Storyline&lt;/h2&gt;\n+&lt;p&gt;([^&lt;]+)</expression>
  101.             </RegExp>
  102.             <RegExp input="$$1" output="&lt;outline&gt;\1&lt;/outline&gt;" dest="5+">
  103.                 <expression trim="1">itemprop=&quot;description&quot;&gt;([^&lt;]+)&lt;</expression>
  104.             </RegExp>
  105.             <RegExp input="$$1" output="&lt;premiered&gt;\1&lt;/premiered&gt;" dest="5+">
  106.                 <expression trim="1">itemprop=&quot;datePublished&quot; datetime=&quot;.{10,}?&quot;&gt;([^&lt;]+)&lt;</expression>
  107.             </RegExp>
  108.             <RegExp input="$$2" output="&lt;url cache=&quot;$$2-combined.html&quot; function=&quot;ParseIMDBCertification&quot;&gt;http://akas.imdb.com/title/$$2/combined&lt;/url&gt;" dest="5+">
  109.                 <expression noclean="1" />
  110.             </RegExp>
  111.             <RegExp input="$$2" output="&lt;url cache=&quot;$$2-main.html&quot; function=&quot;ParseIMDBCast&quot;&gt;http://akas.imdb.com/title/$$2/&lt;/url&gt;" dest="5+">
  112.                 <expression noclean="1" />
  113.             </RegExp>
  114.             <RegExp input="$$2" output="&lt;url cache=&quot;$$2-main.html&quot; function=&quot;ParseIMDBDirectors&quot;&gt;http://akas.imdb.com/title/$$2/&lt;/url&gt;" dest="5+">
  115.                 <expression noclean="1" />
  116.             </RegExp>
  117.             <RegExp input="$$2" output="&lt;url cache=&quot;$$2-main.html&quot; function=&quot;ParseIMDBWriters&quot;&gt;http://akas.imdb.com/title/$$2/&lt;/url&gt;" dest="5+">
  118.                 <expression noclean="1" />
  119.             </RegExp>
  120.             <RegExp input="$$2" output="&lt;url function=&quot;ParseTMDBFanart&quot; cache=&quot;tmdb-images-$$2.xml&quot;&gt;http://api.themoviedb.org/2.1/Movie.getImages/en/xml/57983e31fb435df4df77afb854740ea9/$$2&lt;/url&gt;" dest="5+">
  121.                 <expression/>
  122.             </RegExp>
  123.             <RegExp input="$$2" output="&lt;url function=&quot;ParseTMDBThumbs&quot; cache=&quot;tmdb-images-$$2.xml&quot;&gt;http://api.themoviedb.org/2.1/Movie.imdbLookup/en/xml/57983e31fb435df4df77afb854740ea9/$$2&lt;/url&gt;" dest="5+">
  124.                 <expression/>
  125.             </RegExp>
  126.             <RegExp conditional="tmdbplot" input="$$2" output="&lt;url function=&quot;GetTMDBPlot&quot;&gt;http://api.themoviedb.org/2.1/Movie.imdbLookup/en/xml/57983e31fb435df4df77afb854740ea9/$$2&lt;/url&gt;" dest="5+">
  127.                 <expression/>
  128.             </RegExp>
  129.             <RegExp input="$INFO[TrailerQ]" output="&lt;url function=&quot;Get480pTrailer&quot;&gt;http://www.hd-trailers.net/movie/$$19&lt;/url&gt;" dest="5+">
  130.                 <expression>480p</expression>
  131.             </RegExp>
  132.             <RegExp input="$INFO[TrailerQ]" output="&lt;url function=&quot;Get720pTrailer&quot;&gt;http://www.hd-trailers.net/movie/$$19&lt;/url&gt;" dest="5+">
  133.                 <expression>720p</expression>
  134.             </RegExp>
  135.             <RegExp input="$INFO[TrailerQ]" output="&lt;url function=&quot;Get1080pTrailer&quot;&gt;http://www.hd-trailers.net/movie/$$19&lt;/url&gt;" dest="5+">
  136.                 <expression>1080p</expression>
  137.             </RegExp>
  138.             <expression noclean="1"/>
  139.         </RegExp>
  140.     </GetDetails>
  141.     <ParseIMDBAKATitles dest="5">
  142.         <RegExp input="$$2" output="&lt;details&gt;&lt;sorttitle&gt;\1&lt;/sorttitle&gt;&lt;title&gt;\1&lt;/title&gt;&lt;/details&gt;" dest="5">
  143.             <RegExp input="$$1" output="\1" dest="2">
  144.                 <expression>&lt;h1&gt;([^&lt;]*)</expression>
  145.             </RegExp>
  146.             <RegExp input="$$10" output="\1" dest="4">
  147.                 <RegExp input="$$1" output="\1" dest="9">
  148.                     <expression>&lt;meta name=&quot;title&quot; content=&quot;(?:&amp;#x22;)?([^&quot;]*?)(?:&amp;#x22;)? \([^\(]*?([0-9]{4})\)</expression>
  149.                 </RegExp>
  150.                 <RegExp input="$$9" output="\1" dest="10">
  151.                     <expression />
  152.                 </RegExp>
  153.                 <RegExp input="$$1" output="\1" dest="11">
  154.                     <expression clear="yes">&gt;\s*?&quot;([^&lt;]+)&quot;[^&lt;]+[&lt;em&gt;]?[^&quot;]+Hong Kong [&lt;em&gt;][^&quot;]+English</expression>
  155.                 </RegExp>
  156.                 <RegExp input="$$11" output="\1" dest="10">
  157.                     <expression>(.+)</expression>
  158.                 </RegExp>
  159.                 <RegExp input="$$1" output="\1" dest="11">
  160.                     <expression clear="yes">&gt;\s*?&quot;([^&lt;]+)&quot;[^&lt;]+[&lt;em&gt;]?[^&quot;]+((Canada)\s(&lt;em&gt;)?\((English|imdb))</expression>
  161.                 </RegExp>
  162.                 <RegExp input="$$11" output="\1" dest="10">
  163.                     <expression>(.+)</expression>
  164.                 </RegExp>
  165.                 <RegExp input="$$1" output="\1" dest="11">
  166.                     <expression clear="yes">&gt;\s*?&quot;([^&lt;]+)&quot;[^&lt;]+[&lt;em&gt;]?[^&quot;]+(UK(&lt;em&gt;)?&lt;br&gt;)</expression>
  167.                 </RegExp>
  168.                 <RegExp input="$$11" output="\1" dest="10">
  169.                     <expression>(.+)</expression>
  170.                 </RegExp>
  171.                 <RegExp input="$$1" output="\1" dest="11">
  172.                     <expression clear="yes">&gt;\s*?&quot;([^&lt;]+)&quot;[^&quot;]+International\s(&lt;em&gt;)?\(English title\)(&lt;/em&gt;)?(,|&lt;)( |b)</expression>
  173.                 </RegExp>
  174.                 <RegExp input="$$11" output="\1" dest="10">
  175.                     <expression>(.+)</expression>
  176.                 </RegExp>
  177.                 <RegExp input="$$1" output="\1" dest="11">
  178.                     <expression clear="yes">&gt;\s*?&quot;([^&lt;]+)&quot;[^&quot;]+International\s(&lt;em&gt;)?\(English title\)(&lt;/em&gt;)? (&lt;em&gt;)?\(imdb</expression>
  179.                 </RegExp>
  180.                 <RegExp input="$$11" output="\1" dest="10">
  181.                     <expression>(.+)</expression>
  182.                 </RegExp>
  183.                 <RegExp input="$$1" output="\1" dest="12">
  184.                     <expression clear="yes">&lt;a href=&quot;/country/[^&gt;]+&gt;(UK&lt;/a&gt;&lt;/div&gt;)</expression>
  185.                 </RegExp>
  186.                 <RegExp input="$$12" output="$$9" dest="10">
  187.                     <expression>(.+)</expression>
  188.                 </RegExp>
  189.                 <RegExp input="$$1" output="\1" dest="12">
  190.                     <expression clear="yes">&lt;a href=&quot;/country/[^&gt;]+&gt;(USA&lt;/a&gt;&lt;/div&gt;)</expression>
  191.                 </RegExp>
  192.                 <RegExp input="$$12" output="$$9" dest="10">
  193.                     <expression>(.+)</expression>
  194.                 </RegExp>
  195.                 <RegExp input="$$1" output="\1" dest="11">
  196.                     <expression clear="yes">&gt;\s*?&quot;([^&lt;]+)&quot;[^&lt;]+[&lt;em&gt;]?[^&quot;]+(USA(&lt;em&gt;)?&lt;br&gt;)</expression>
  197.                 </RegExp>
  198.                 <RegExp input="$$11" output="\1" dest="10">
  199.                     <expression>(.+)</expression>
  200.                 </RegExp>
  201.                 <RegExp input="$$1" output="\1" dest="11">
  202.                     <expression clear="yes">&gt;\s*?&quot;([^&lt;]+)&quot;[^&lt;]+[&lt;em&gt;]?[^&quot;]+((USA)\s(&lt;em&gt;)?\((English|imdb))</expression>
  203.                 </RegExp>
  204.                 <RegExp input="$$11" output="\1" dest="10">
  205.                     <expression>(.+)</expression>
  206.                 </RegExp>
  207.                 <RegExp input="$$1" output="\1" dest="11">
  208.                     <expression clear="yes">&gt;\s*?&quot;([^&lt;]+)&quot;[^&lt;]+[&lt;em&gt;]?[^&quot;]+((USA)\s(&lt;em&gt;)?\((new title))</expression>
  209.                 </RegExp>
  210.                 <RegExp input="$$11" output="\1" dest="10">
  211.                     <expression>(.+)</expression>
  212.                 </RegExp>
  213.                 <expression noclean="1" />
  214.             </RegExp>
  215.             <RegExp input="$$4" output="\1" dest="2">
  216.                 <expression>(.+)</expression>
  217.             </RegExp>
  218.             <RegExp input="$$1" output="\1" dest="4">
  219.                 <expression>&gt;\s*?&quot;([^&lt;]+)&quot;[^&lt;]+[&lt;em&gt;]?[^&quot;]+$INFO[akatitles]</expression>
  220.             </RegExp>
  221.             <RegExp input="$$4" output="\1" dest="2">
  222.                 <expression>(.+)</expression>
  223.             </RegExp>
  224.             <RegExp input="$$1" output="\1" dest="5">
  225.                 <expression>&gt;&lt;a href=&quot;/country/[^&gt;]+&gt;($INFO[akatitles]&lt;/a&gt;&lt;/div&gt;)</expression>
  226.             </RegExp>
  227.             <RegExp input="$$5" output="$$9" dest="4">
  228.                 <expression>($INFO[akatitles])</expression>
  229.             </RegExp>
  230.             <RegExp input="$INFO[akatitles]" output="$$9" dest="4">
  231.                 <expression>Keep Original</expression>
  232.             </RegExp>
  233.             <RegExp input="$$4" output="\1" dest="2">
  234.                 <expression>(.+)</expression>
  235.             </RegExp>
  236.             <expression noclean="1" />
  237.         </RegExp>
  238.     </ParseIMDBAKATitles>
  239.     <ParseIMDBCertification dest="5">
  240.         <RegExp input="$$2" output="&lt;details&gt;\1&lt;/details&gt;" dest="5">
  241.             <RegExp input="$$1" output="&lt;certification&gt;\1&lt;/certification&gt;" dest="2+">
  242.                 <expression repeat="yes">certificates=[^&quot;]*&quot;&gt;([^&lt;]*)</expression>
  243.             </RegExp>
  244.             <expression noclean="1" />
  245.         </RegExp>
  246.     </ParseIMDBCertification>
  247.     <ParseIMDBCast dest="5">
  248.         <RegExp input="$$2" output="&lt;details&gt;\1&lt;/details&gt;" dest="5">
  249.             <RegExp input="$$1" output="\1" dest="6">
  250.                 <expression noclean="1">&lt;table class=&quot;cast_list&quot;&gt;(.*?)&lt;/table&gt;</expression>
  251.             </RegExp>
  252.             <RegExp input="$$6" output="&lt;actor&gt;&lt;thumb&gt;\2_SX512_SY512_\3&lt;/thumb&gt;&lt;name&gt;\1&lt;/name&gt;&lt;role&gt;\5&lt;/role&gt;&lt;/actor&gt;" dest="7">
  253.                 <expression repeat="yes" clear="yes" trim="3,4" noclean="1,2">&lt;noscript&gt;&lt;img \n[^a]*alt=&quot;([^&quot;]*)&quot;[^&quot;]*&quot;[^&quot;]*&quot;[^s]*src=&quot;(?:([^&quot;]*\.)[^&quot;]*(\.jpg))[^&gt;]*.*?ter"&gt;[^&gt;]*&gt;\n\s*(&lt;[^&gt;]*&gt;)?([^&lt;\(]*)?</expression>
  254.             </RegExp>
  255.             <RegExp input="$$6" output="&lt;actor&gt;&lt;thumb&gt;&lt;/thumb&gt;&lt;name&gt;\1&lt;/name&gt;&lt;role&gt;\3&lt;/role&gt;&lt;/actor&gt;" dest="7+">
  256.                 <expression repeat="yes" trim="1,2" noclean="1,2">&lt;img \n[^a]*alt=&quot;([^&quot;]*)&quot;[^&quot;]*&quot;[^&quot;]*&quot;[^s]*src=&quot;(?:[^&quot;]*\.[^&quot;]*\.png)&quot; class=&quot;&quot; /&gt;*.*?ter"&gt;[^&gt;]*&gt;\n\s*(&lt;[^&gt;]*&gt;)?([^&lt;\(]*)?</expression>
  257.             </RegExp>
  258.             <RegExp input="$$7" output="&lt;actor&gt;&lt;thumb&gt;\1&lt;/thumb&gt;\2&lt;/actor&gt;" dest="2+">
  259.                 <expression repeat="yes" clear="yes" noclean="1,2,3">&lt;actor&gt;&lt;thumb&gt;(?:(http.*?)|_SX[0-9]+_SY[0-9]+_)?&lt;/thumb&gt;(.*?)&lt;/actor&gt;</expression>
  260.             </RegExp>
  261.             <expression noclean="1" />
  262.         </RegExp>
  263.     </ParseIMDBCast>
  264.     <ParseIMDBDirectors dest="5">
  265.         <RegExp input="$$2" output="&lt;details&gt;\1&lt;/details&gt;" dest="5">
  266.             <RegExp input="$$1" output="\1" dest="6">
  267.                 <expression clear="yes" noclean="1">Director[s]?.*?&lt;/h4&gt;(.*?)&lt;div class</expression>
  268.             </RegExp>
  269.             <RegExp input="$$6" output="&lt;director&gt;\1&lt;/director&gt;" dest="2+">
  270.                 <expression repeat="yes" fixchars="1" clear="yes">&lt;a  href=&quot;/name[^&gt;]*&gt;([^&lt;]*)&lt;</expression>
  271.             </RegExp>
  272.             <expression noclean="1" />
  273.         </RegExp>
  274.     </ParseIMDBDirectors>
  275.     <ParseIMDBWriters dest="5">
  276.         <RegExp input="$$2" output="&lt;details&gt;\1&lt;/details&gt;" dest="5">
  277.             <RegExp input="$$1" output="\1" dest="6">
  278.                 <expression noclean="1">Writer[s]?.*?&lt;/h4&gt;(.*?)&lt;div class</expression>
  279.             </RegExp>
  280.             <RegExp input="$$6" output="&lt;credits&gt;\1&lt;/credits&gt;" dest="2+">
  281.                 <expression repeat="yes" fixchars="1">&lt;a  href=&quot;/name[^&gt;]*&gt;([^&lt;]*)&lt;</expression>
  282.             </RegExp>
  283.             <expression noclean="1" />
  284.         </RegExp>
  285.     </ParseIMDBWriters>
  286.     <ParseTMDBFanart dest="5">
  287.         <RegExp input="$$2" output="&lt;details&gt;&lt;fanart&gt;\1&lt;/fanart&gt;&lt;/details&gt;" dest="5">
  288.             <RegExp input="$$1" output="&lt;thumb&gt;\1&lt;/thumb&gt;" dest="3">
  289.                 <expression noclean="1" repeat="yes">&lt;backdrop[^&gt;]*&gt;(.*?)&lt;/backdrop&gt;</expression>
  290.             </RegExp>
  291.             <RegExp input="$$3" output="&lt;thumb preview=&quot;\2&quot;&gt;\1\3&lt;/thumb&gt;" dest="4">
  292.                 <expression noclean="1,3" repeat="yes">&lt;thumb&gt;(.*?)url=&quot;([^&quot;]*)&quot; size=&quot;poster&quot;(.*?)&lt;/thumb&gt;</expression>
  293.             </RegExp>
  294.             <RegExp input="$$4" output="&lt;thumb\1&gt;\2&lt;/thumb&gt;" dest="2">
  295.                 <expression repeat="yes" noclean="1">&lt;thumb([^&gt;]*)&gt;.*?url=&quot;([^&quot;]*)&quot; size=&quot;original&quot;.*?&lt;/thumb&gt;</expression>
  296.             </RegExp>
  297.             <expression noclean="1">(.+)</expression>
  298.         </RegExp>
  299.     </ParseTMDBFanart>
  300.     <ParseTMDBThumbs dest="5">
  301.         <RegExp input="$$2" output="&lt;details&gt;\1&lt;/details&gt;" dest="5">
  302.             <RegExp input="$$1" output="&lt;thumb&gt;\1&lt;/thumb&gt;" dest="3">
  303.                 <expression noclean="1" repeat="yes">&lt;poster[^&gt;]*&gt;(.*?)&lt;/poster&gt;</expression>
  304.             </RegExp>
  305.             <RegExp input="$$3" output="&lt;thumb preview=&quot;\2&quot;&gt;\1\3&lt;/thumb&gt;" dest="4">
  306.                 <expression noclean="1,3" repeat="yes">&lt;thumb&gt;(.*?)url=&quot;([^&quot;]*)&quot; size=&quot;mid&quot;(.*?)&lt;/thumb&gt;</expression>
  307.             </RegExp>
  308.             <RegExp input="$$4" output="&lt;thumb\1&gt;\2&lt;/thumb&gt;" dest="2">
  309.                 <expression repeat="yes" noclean="1">&lt;thumb([^&gt;]*)&gt;.*?url=&quot;([^&quot;]*)&quot; size=&quot;original&quot;.*?&lt;/thumb&gt;</expression>
  310.             </RegExp>
  311.             <expression noclean="1">(.+)</expression>
  312.         </RegExp>
  313.     </ParseTMDBThumbs>
  314.     <GetTMDBPlot dest="5">
  315.         <RegExp input="$$1" output="&lt;details&gt;&lt;url function=&quot;ParseTMDBPlot&quot; cache=&quot;tmdb-info-$INFO[plotlanguage]-\1.xml&quot;&gt;http://api.themoviedb.org/2.1/Movie.getInfo/$INFO[plotlanguage]/xml/57983e31fb435df4df77afb854740ea9/\1&lt;/url&gt;&lt;/details&gt;" dest="5">
  316.             <expression noclean="1">&lt;id&gt;([^&lt;]*)&lt;/id&gt;</expression>
  317.         </RegExp>
  318.     </GetTMDBPlot>
  319.     <ParseTMDBPlot dest="5">
  320.         <RegExp input="$$2" output="&lt;details&gt;\1&lt;/details&gt;" dest="5">
  321.             <RegExp input="$$1" output="&lt;plot&gt;\1&lt;/plot&gt;" dest="2">
  322.                 <expression noclean="1">&lt;overview&gt;([^&lt;]*)&lt;/overview&gt;</expression>
  323.             </RegExp>
  324.             <RegExp input="$$1" output="&lt;outline&gt;\1.&lt;/outline&gt;" dest="2+">
  325.                 <expression noclean="1">&lt;overview&gt;(.{180,}?)[.?!]</expression>
  326.             </RegExp>
  327.             <expression noclean="1">(.+)</expression>
  328.         </RegExp>
  329.     </ParseTMDBPlot>
  330.     <Get480pTrailer dest="5">
  331.         <!--Fetch 480p trailer link from hdtrailers.net-->
  332.         <RegExp input="$$1" output="&lt;details&gt;&lt;trailer&gt;http://playlist.yahoo.com/makeplaylist.dll?sdm=web&amp;pt=rd&amp;sid=\1&lt;/trailer&gt;&lt;/details&gt;" dest="5">
  333.             <expression noclean="1">&quot;&gt;Trailer&lt;/span&gt;&lt;/td&gt;.*?sid=(\d*)&quot; rel=&quot;lightbox\[res480p</expression>
  334.         </RegExp>
  335.     </Get480pTrailer>
  336.     <Get720pTrailer dest="5">
  337.         <RegExp input="$$6" output="&lt;details&gt;&lt;trailer&gt;\1&lt;/trailer&gt;&lt;/details&gt;" dest="5">
  338.             <!--Fallback to fetch 480p trailer link if 720p is not available-->
  339.             <RegExp input="$$1" output="http://playlist.yahoo.com/makeplaylist.dll?sdm=web&amp;pt=rd&amp;sid=\1" dest="6">
  340.                 <expression noclean="1">&quot;&gt;Trailer&lt;/span&gt;&lt;/td&gt;.*?sid=(\d*)&quot; rel=&quot;lightbox\[res480p</expression>
  341.             </RegExp>
  342.             <!--Fetch 720p trailer link from hdtrailers.net-->
  343.             <RegExp input="$$1" output="http://playlist.yahoo.com/makeplaylist.dll?sdm=web&amp;pt=rd&amp;sid=\1" dest="6">
  344.                 <expression noclean="1">&quot;&gt;Trailer&lt;/span&gt;&lt;/td&gt;.*?sid=(\d*)&quot; rel=&quot;lightbox\[res720p</expression>
  345.             </RegExp>
  346.             <expression noclean="1" />
  347.         </RegExp>
  348.     </Get720pTrailer>
  349.     <Get1080pTrailer dest="5">
  350.         <RegExp input="$$6" output="&lt;details&gt;&lt;trailer&gt;\1&lt;/trailer&gt;&lt;/details&gt;" dest="5">
  351.             <!--Fallback to fetch 480p trailer link if 720p and 1080p are not available-->
  352.             <RegExp input="$$1" output="http://playlist.yahoo.com/makeplaylist.dll?sdm=web&amp;pt=rd&amp;sid=\1" dest="6">
  353.                 <expression noclean="1">&quot;&gt;Trailer&lt;/span&gt;&lt;/td&gt;.*?sid=(\d*)&quot; rel=&quot;lightbox\[res480p</expression>
  354.             </RegExp>
  355.             <!--Fallback to fetch 720p trailer link if 1080p is not available-->
  356.             <RegExp input="$$1" output="http://playlist.yahoo.com/makeplaylist.dll?sdm=web&amp;pt=rd&amp;sid=\1" dest="6">
  357.                 <expression noclean="1">&quot;&gt;Trailer&lt;/span&gt;&lt;/td&gt;.*?sid=(\d*)&quot; rel=&quot;lightbox\[res720p</expression>
  358.             </RegExp>
  359.             <!--Fetch 1080p trailer link from hdtrailers.net-->
  360.             <RegExp input="$$1" output="http://playlist.yahoo.com/makeplaylist.dll?sdm=web&amp;pt=rd&amp;sid=\1" dest="6">
  361.                 <expression noclean="1">&quot;&gt;Trailer&lt;/span&gt;&lt;/td&gt;.*?sid=(\d*)&quot; rel=&quot;lightbox\[res1080p</expression>
  362.             </RegExp>
  363.             <expression noclean="1" />
  364.         </RegExp>
  365.     </Get1080pTrailer>
  366. </scraper>
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement