/Assets/scrapers/metadata.imdb.com/imdb.xml
XML | 246 lines | 246 code | 0 blank | 0 comment | 0 complexity | 9283fd5c66bf0abcc5691359ff82d2bf MD5 | raw file
Possible License(s): GPL-2.0
- <?xml version="1.0" encoding="UTF-8"?>
- <scraper framework="1.1" date="2011-10-28">
- <NfoUrl dest="3">
- <RegExp input="$$1" output="<url>http://akas.imdb.com/title/tt\1/</url><id>tt\1</id>" dest="3">
- <expression clear="yes" noclean="1">imdb....?/Title\?([0-9]*)</expression>
- </RegExp>
- <RegExp input="$$1" output="<url>http://akas.imdb.com/title/tt\1/</url><id>tt\1</id>" dest="3+">
- <expression noclean="1">imdb....?/title/tt([0-9]*)</expression>
- </RegExp>
- </NfoUrl>
- <CreateSearchUrl dest="3" SearchStringEncoding="iso-8859-1">
- <RegExp input="$$1" output="<url>http://akas.imdb.com/find?s=tt;q=\1$$4</url>" dest="3">
- <RegExp input="$$2" output="%20(\1)" dest="4">
- <expression clear="yes">(.+)</expression>
- </RegExp>
- <expression noclean="1"/>
- </RegExp>
- </CreateSearchUrl>
- <GetSearchResults dest="8">
- <RegExp input="$$5" output="<?xml version="1.0" encoding="iso-8859-1" standalone="yes"?><results>\1</results>" dest="8">
- <RegExp input="$$1" output="\1" dest="7">
- <expression clear="yes">/title/([t0-9]*)/(combined|faq|releaseinfo|vote)</expression>
- </RegExp>
- <RegExp input="$$1" output="<entity><title>\1</title><year>\2</year><url cache="$$7-main.html">http://akas.imdb.com/title/$$7/</url><id>$$7</id></entity>" dest="5">
- <expression clear="yes" noclean="1"><meta name="title" content="(?:&#x22;)?([^"]*?)(?:&#x22;)? \([^\(]*?([0-9]{4})\)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="4">
- <expression noclean="1">(><a href="/title.*)</expression>
- </RegExp>
- <RegExp input="$$4" output="<entity><title>\2</title><year>\3</year><url cache="\1-main.html">http://akas.imdb.com/title/\1/</url><id>\1</id></entity>" dest="5+">
- <expression repeat="yes" noclean="1,2">><a href="/title/([t0-9]*)/[^>]*>(?:&#x22;)?([^<]*?)(?:&#x22;)?</a> *\([^\(]*?([0-9]{4})</expression>
- </RegExp>
- <expression clear="yes" noclean="1"/>
- </RegExp>
- </GetSearchResults>
- <GetDetails dest="3">
- <RegExp input="$$5" output="<details>\1</details>" dest="3">
- <RegExp input="$$2" output="<id>\1</id>" dest="5">
- <expression/>
- </RegExp>
- <RegExp input="$$1" output="<originaltitle>\2</originaltitle>" dest="5+">
- <expression><meta name="title" content="(IMDb - )?(?:&#x22;)?([^"]*?)(?:&#x22;)? \([^\(]*?([0-9]{4})\)</expression>
- </RegExp>
- <RegExp input="$$2" output="<chain function="GetIMDBAKATitlesById">\1</chain>" dest="5+">
- <expression>(tt[t0-9]*)</expression>
- </RegExp>
- <RegExp input="$$1" output="<year>\2</year>" dest="5+">
- <expression><meta name="title" content="([^"]*?) \([^\(]*?([0-9]{4})\)</expression>
- </RegExp>
- <RegExp input="$$1" output="<top250>\1</top250>" dest="5+">
- <expression>Top 250 #([0-9]*)</</expression>
- </RegExp>
- <RegExp input="$$1" output="<mpaa>\1</mpaa>" dest="5+">
- <expression>MPAA</a>\)</h4>\n?<span itemprop="contentRating">([^<]*)</expression>
- </RegExp>
- <RegExp input="$$1" output="<tagline>\1</tagline>" dest="5+">
- <expression><h4[^>]*?>Taglines:</h4>\n?([^<]*)</expression>
- </RegExp>
- <RegExp input="$$1" output="<runtime>\1</runtime>" dest="5+">
- <expression trim="1"><h4[^>]*?>Runtime:</h4>[^0-9]*([0-9]+)</expression>
- </RegExp>
- <RegExp input="$$2" output="<chain function="GetIMDBRatingById">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp input="$$2" output="<chain function="GetIMDBGenresById">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp input="$$1" output="<country>\1</country>" dest="5+">
- <expression repeat="yes">"/country/[^>]+>([^<]+)</a></expression>
- </RegExp>
- <RegExp input="$$1" output="<studio>\1</studio>" dest="5+">
- <expression>"/company/[^>]+>([^<]+)</a></expression>
- </RegExp>
- <RegExp input="$$1" output="<outline>\1</outline><plot>\1</plot>" dest="5+">
- <expression trim="1"><p>\n<p itemprop="description">([^<]+)</p></expression>
- </RegExp>
- <RegExp input="$$2" output="<chain function="GetIMDBPlotById">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp conditional="!fullcredits" input="$$2" output="<chain function="GetIMDBScaledCastById">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp conditional="!fullcredits" input="$$2" output="<chain function="GetIMDBDirectorsById">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp conditional="!fullcredits" input="$$2" output="<chain function="GetIMDBWritersById">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp conditional="fullcredits" input="$$2" output="<chain function="GetIMDBScaledFullCastById">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp conditional="fullcredits" input="$$2" output="<chain function="GetIMDBFullDirectorsById">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp conditional="fullcredits" input="$$2" output="<chain function="GetIMDBFullWritersById">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp conditional="tmdbthumbs" input="$$2" output="<chain function="GetTMDBThumbsByIdChain">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp conditional="impawards" input="$$2" output="<chain function="GetIMPAThumbsById">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp conditional="movieposterdb" input="$$2" output="<chain function="GetMoviePosterDBThumbs">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp conditional="fanart" input="$$2" output="<chain function="GetTMDBFanartByIdChain">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <RegExp conditional="ytrailer" input="$$1" output="<chain function="GetYoutubeTrailer">$$6=$$9</chain>" dest="5+">
- <RegExp input="$$1" output="\2" dest="6">
- <expression><meta name="title" content="(IMDb - )?(?:&#x22;)?([^"]*?)(?:&#x22;)? \([^\(]*?([0-9]{4})\)</expression>
- </RegExp>
- <RegExp input="$$1" output="\2" dest="9">
- <expression><meta name="title" content="([^"]*?) \([^\(]*?([0-9]{4})\)</expression>
- </RegExp>
- <expression/>
- </RegExp>
- <RegExp input="$INFO[TrailerQ]" output="<chain function="GetHDTrailersnet480p">$$6</chain>" dest="5+">
- <RegExp input="$$1" output="\1" dest="6">
- <expression><meta name="title" content="(?:&#x22;)?([^"]*?)(?:&#x22;)? \([^\(]*?([0-9]{4})\)</expression>
- </RegExp>
- <expression>480p</expression>
- </RegExp>
- <RegExp input="$INFO[TrailerQ]" output="<chain function="GetHDTrailersnet720p">$$6</chain>" dest="5+">
- <RegExp input="$$1" output="\1" dest="6">
- <expression><meta name="title" content="(?:&#x22;)?([^"]*?)(?:&#x22;)? \([^\(]*?([0-9]{4})\)</expression>
- </RegExp>
- <expression>720p</expression>
- </RegExp>
- <RegExp input="$INFO[TrailerQ]" output="<chain function="GetHDTrailersnet1080p">$$6</chain>" dest="5+">
- <RegExp input="$$1" output="\1" dest="6">
- <expression><meta name="title" content="(?:&#x22;)?([^"]*?)(?:&#x22;)? \([^\(]*?([0-9]{4})\)</expression>
- </RegExp>
- <expression>1080p</expression>
- </RegExp>
- <RegExp input="$$2" output="<chain function="GetIMDBThumbsById">$$2</chain>" dest="5+">
- <expression/>
- </RegExp>
- <expression noclean="1"/>
- </RegExp>
- </GetDetails>
- <GetIMDBAKATitlesById dest="5">
- <RegExp input="$$1" output="<details><url cache="$$1-combined.html" function="ParseIMDBAKATitles">http://akas.imdb.com/title/$$1/combined</url></details>" dest="5">
- <expression noclean="1" />
- </RegExp>
- </GetIMDBAKATitlesById>
- <ParseIMDBAKATitles dest="5">
- <RegExp input="$$2" output="<details><title>\1</title></details>" dest="5">
- <RegExp input="$$1" output="\1" dest="2">
- <expression><h1>([^<]*)</expression>
- </RegExp>
- <RegExp input="$$10" output="\1" dest="4">
- <RegExp input="$$1" output="\1" dest="9">
- <expression><meta name="title" content="(?:&#x22;)?([^"]*?)(?:&#x22;)? \([^\(]*?([0-9]{4})\)</expression>
- </RegExp>
- <RegExp input="$$9" output="\1" dest="10">
- <expression />
- </RegExp>
- <RegExp input="$$1" output="\1" dest="11">
- <expression clear="yes">>\s*?"([^<]+)"[^<]+[<em>]?[^"]+Hong Kong [<em>][^"]+English</expression>
- </RegExp>
- <RegExp input="$$11" output="\1" dest="10">
- <expression>(.+)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="11">
- <expression clear="yes">>\s*?"([^<]+)"[^<]+[<em>]?[^"]+((Canada)\s(<em>)?\((English|imdb))</expression>
- </RegExp>
- <RegExp input="$$11" output="\1" dest="10">
- <expression>(.+)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="11">
- <expression clear="yes">>\s*?"([^<]+)"[^<]+[<em>]?[^"]+(UK(<em>)?<br>)</expression>
- </RegExp>
- <RegExp input="$$11" output="\1" dest="10">
- <expression>(.+)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="11">
- <expression clear="yes">>\s*?"([^<]+)"[^"]+International\s(<em>)?\(English title\)(</em>)?(,|<)( |b)</expression>
- </RegExp>
- <RegExp input="$$11" output="\1" dest="10">
- <expression>(.+)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="11">
- <expression clear="yes">>\s*?"([^<]+)"[^"]+International\s(<em>)?\(English title\)(</em>)? (<em>)?\(imdb</expression>
- </RegExp>
- <RegExp input="$$11" output="\1" dest="10">
- <expression>(.+)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="12">
- <expression clear="yes"><a href="/country/[^>]+>(UK</a></div>)</expression>
- </RegExp>
- <RegExp input="$$12" output="$$9" dest="10">
- <expression>(.+)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="12">
- <expression clear="yes"><a href="/country/[^>]+>(USA</a></div>)</expression>
- </RegExp>
- <RegExp input="$$12" output="$$9" dest="10">
- <expression>(.+)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="11">
- <expression clear="yes">>\s*?"([^<]+)"[^<]+[<em>]?[^"]+(USA(<em>)?<br>)</expression>
- </RegExp>
- <RegExp input="$$11" output="\1" dest="10">
- <expression>(.+)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="11">
- <expression clear="yes">>\s*?"([^<]+)"[^<]+[<em>]?[^"]+((USA)\s(<em>)?\((English|imdb))</expression>
- </RegExp>
- <RegExp input="$$11" output="\1" dest="10">
- <expression>(.+)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="11">
- <expression clear="yes">>\s*?"([^<]+)"[^<]+[<em>]?[^"]+((USA)\s(<em>)?\((new title))</expression>
- </RegExp>
- <RegExp input="$$11" output="\1" dest="10">
- <expression>(.+)</expression>
- </RegExp>
- <expression noclean="1" />
- </RegExp>
- <RegExp input="$$4" output="\1" dest="2">
- <expression>(.+)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="4">
- <expression>>\s*?"([^<]+)"[^<]+[<em>]?[^"]+$INFO[akatitles]</expression>
- </RegExp>
- <RegExp input="$$4" output="\1" dest="2">
- <expression>(.+)</expression>
- </RegExp>
- <RegExp input="$$1" output="\1" dest="5">
- <expression><a href="/country/[^>]+>($INFO[akatitles])</expression>
- </RegExp>
- <RegExp input="$$5" output="$$9" dest="4">
- <expression>($INFO[akatitles])</expression>
- </RegExp>
- <RegExp input="$INFO[akatitles]" output="$$9" dest="4">
- <expression>Keep Original</expression>
- </RegExp>
- <RegExp input="$$4" output="\1" dest="2">
- <expression>(.+)</expression>
- </RegExp>
- <expression noclean="1" />
- </RegExp>
- </ParseIMDBAKATitles>
- </scraper>