API tools faq
paste
Advertisement
Guest User

IMDB Scraper

a guest
Jun 3rd, 2010
260
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 9.74 KB | None | 0 0
raw download clone embed print report
  1. <?xml version="1.0" encoding="UTF-8"?>
  2. <scraper framework="1.1" date="2010-02-24" name="IMDb.com" content="movies" thumb="imdb.png" language="en">
  3. <include>common/imdb.xml</include>
  4. <include>common/tmdb.xml</include>
  5. <include>common/movieposterdb.xml</include>
  6. <include>common/impa.xml</include>
  7. <include>common/dtrailer.xml</include>
  8. <GetSettings dest="3">
  9. <RegExp input="$$5" output="&lt;settings&gt;\1&lt;/settings&gt;" dest="3">
  10. <RegExp input="$$1" output="&lt;setting label=&quot;Enable full cast credits&quot; type=&quot;bool&quot; id=&quot;fullcredits&quot; default=&quot;false&quot;&gt;&lt;/setting&gt;" dest="5">
  11. <expression/>
  12. </RegExp>
  13. <RegExp input="$$1" output="&lt;setting label=&quot;Enable fanart from themoviedb.org&quot; type=&quot;bool&quot; id=&quot;fanart&quot; default=&quot;true&quot;&gt;&lt;/setting&gt;" dest="5+">
  14. <expression/>
  15. </RegExp>
  16. <RegExp input="$$1" output="&lt;setting label=&quot;Enable posters from themoviedb.org&quot; type=&quot;bool&quot; id=&quot;tmdbthumbs&quot; default=&quot;true&quot;&gt;&lt;/setting&gt;" dest="5+">
  17. <expression/>
  18. </RegExp>
  19. <RegExp input="$$1" output="&lt;setting label=&quot;Enable posters from IMPAwards&quot; type=&quot;bool&quot; id=&quot;impawards&quot; default=&quot;false&quot;&gt;&lt;/setting&gt;" dest="5+">
  20. <expression/>
  21. </RegExp>
  22. <RegExp input="$$1" output="&lt;setting label=&quot;Enable posters from MoviePosterDB&quot; type=&quot;bool&quot; id=&quot;movieposterdb&quot; default=&quot;false&quot;&gt;&lt;/setting&gt;" dest="5+">
  23. <expression/>
  24. </RegExp>
  25. <RegExp input="$$1" output="&lt;setting label=&quot;Enable IMDb trailers&quot; type=&quot;bool&quot; id=&quot;trailer&quot; default=&quot;true&quot;&gt;&lt;/setting&gt;" dest="5+">
  26. <expression/>
  27. </RegExp>
  28. <RegExp input="$$1" output="&lt;setting label=&quot;Enable trailers from Dtrailer.com&quot; type=&quot;bool&quot; id=&quot;dtrailer&quot; default=&quot;false&quot;&gt;&lt;/setting&gt;" dest="5+">
  29. <expression/>
  30. </RegExp>
  31. <RegExp input="$$1" output="&lt;setting label=&quot;IMDb Poster &amp;amp; Actor Thumb(s) Size&quot; type=&quot;labelenum&quot; values=&quot;192|256|384|512|1024&quot; id=&quot;imdbscale&quot; default=&quot;512&quot;&gt;&lt;/setting&gt;" dest="5+">
  32. <expression/>
  33. </RegExp>
  34. <expression noclean="1"/>
  35. </RegExp>
  36. </GetSettings>
  37. <NfoUrl dest="3">
  38. <RegExp input="$$1" output="&lt;url&gt;http://www.\1/title/tt\2/&lt;/url&gt;&lt;id&gt;tt\2&lt;/id&gt;" dest="3">
  39. <expression clear="yes" noclean="1">(imdb.com)/Title\?([0-9]*)</expression>
  40. </RegExp>
  41. <RegExp input="$$1" output="&lt;url&gt;http://www.\1\2/&lt;/url&gt;&lt;id&gt;tt\2&lt;/id&gt;" dest="3+">
  42. <expression noclean="1">(imdb.com/title/tt)([0-9]*)</expression>
  43. </RegExp>
  44. </NfoUrl>
  45. <CreateSearchUrl dest="3" SearchStringEncoding="iso-8859-1">
  46. <RegExp input="$$1" output="&lt;url&gt;http://akas.imdb.com/find?s=tt;q=\1$$4&lt;/url&gt;" dest="3">
  47. <RegExp input="$$2" output="%20(\1)" dest="4">
  48. <expression clear="yes">(.+)</expression>
  49. </RegExp>
  50. <expression noclean="1"/>
  51. </RegExp>
  52. </CreateSearchUrl>
  53. <GetSearchResults dest="8">
  54. <RegExp input="$$5" output="&lt;?xml version=&quot;1.0&quot; encoding=&quot;iso-8859-1&quot; standalone=&quot;yes&quot;?&gt;&lt;results&gt;\1&lt;/results&gt;" dest="8">
  55. <RegExp input="$$1" output="\1" dest="7">
  56. <expression clear="yes">/title/([t0-9]*)/faq</expression>
  57. </RegExp>
  58. <RegExp input="$$1" output="&lt;entity&gt;&lt;title&gt;\1&lt;/title&gt;&lt;year&gt;\2&lt;/year&gt;&lt;url&gt;http://akas.imdb.com/title/$$7/&lt;/url&gt;&lt;id&gt;$$7&lt;/id&gt;&lt;/entity&gt;" dest="5">
  59. <expression clear="yes" noclean="1">&lt;meta name=&quot;title&quot; content=&quot;([^&quot;]*) \(([0-9]*)\)</expression>
  60. </RegExp>
  61. <RegExp input="$$1" output="\1" dest="4">
  62. <expression noclean="1">(&gt;&lt;a href=&quot;/title.*)</expression>
  63. </RegExp>
  64. <RegExp input="$$4" output="&lt;entity&gt;&lt;title&gt;\2&lt;/title&gt;&lt;year&gt;\3&lt;/year&gt;&lt;url&gt;http://akas.imdb.com/title/\1/&lt;/url&gt;&lt;id&gt;\1&lt;/id&gt;&lt;/entity&gt;" dest="5+">
  65. <expression repeat="yes" noclean="1,2">&gt;&lt;a href=&quot;/title/([t0-9]*)/[^&gt;]*&gt;([^&lt;]*)&lt;/a&gt; *\(([0-9]*)</expression>
  66. </RegExp>
  67. <expression clear="yes" noclean="1"/>
  68. </RegExp>
  69. </GetSearchResults>
  70. <GetDetails dest="3">
  71. <RegExp input="$$5" output="&lt;details&gt;\1&lt;/details&gt;" dest="3">
  72. <RegExp input="$$2" output="&lt;id&gt;\1&lt;/id&gt;" dest="5">
  73. <expression/>
  74. </RegExp>
  75. <RegExp input="$$1" output="&lt;title&gt;\1&lt;/title&gt;" dest="5+">
  76. <expression trim="1" noclean="1">&lt;h1&gt;([^&lt;]*)</expression>
  77. </RegExp>
  78. <RegExp input="$$1" output="&lt;year&gt;\1&lt;/year&gt;" dest="5+">
  79. <expression>a href=&quot;/year/([0-9]*)</expression>
  80. </RegExp>
  81. <RegExp input="$$1" output="&lt;top250&gt;\1&lt;/top250&gt;" dest="5+">
  82. <expression>Top 250: #([0-9]*)&lt;/a&gt;</expression>
  83. </RegExp>
  84. <RegExp input="$$1" output="&lt;mpaa&gt;\1&lt;/mpaa&gt;" dest="5+">
  85. <expression>MPAA&lt;/a&gt;:&lt;/h5&gt;\n&lt;div class=&quot;info-content&quot;&gt;\n(.[^&lt;]*)</expression>
  86. </RegExp>
  87. <RegExp input="$$1" output="&lt;certification&gt;\1 \3&lt;/certification&gt;" dest="5+">
  88. <expression repeat="yes">&lt;a href=&quot;/search/title\?certificates=[^&quot;]*&quot;&gt;([^&lt;]*)&lt;/a&gt;[^&lt;]*(&lt;i&gt;([^&lt;]*)&lt;/i&gt;)?</expression>
  89. </RegExp>
  90. <RegExp input="$$1" output="&lt;tagline&gt;\1&lt;/tagline&gt;" dest="5+">
  91. <expression>&lt;h5&gt;Tagline:&lt;/h5&gt;\n&lt;div class="info-content"&gt;([^&lt;]*)</expression>
  92. </RegExp>
  93. <RegExp input="$$1" output="&lt;runtime&gt;\1&lt;/runtime&gt;" dest="5+">
  94. <expression trim="1">&lt;h5&gt;Runtime:&lt;/h5&gt;[^0-9]*([^&lt;]*)</expression>
  95. </RegExp>
  96. <RegExp input="$$1" output="&lt;rating&gt;\1&lt;/rating&gt;&lt;votes&gt;\2&lt;/votes&gt;" dest="5+">
  97. <expression>&lt;b&gt;([0-9.]+)/10&lt;/b&gt;[^&lt;]*&lt;a href=&quot;ratings&quot; class=&quot;tn15more&quot;&gt;([0-9,]+) votes&lt;/a&gt;</expression>
  98. </RegExp>
  99. <RegExp input="$$1" output="&lt;genre&gt;\1&lt;/genre&gt;" dest="5+">
  100. <expression repeat="yes">&quot;/Sections/Genres/[^/]*/&quot;&gt;([^&lt;]*)&lt;/a&gt;</expression>
  101. </RegExp>
  102. <RegExp input="$$1" output="&lt;studio&gt;\1&lt;/studio&gt;" dest="5+">
  103. <expression repeat="yes">&quot;/company/[^/]*/&quot;&gt;([^&lt;]*)&lt;/a&gt;</expression>
  104. </RegExp>
  105. <RegExp input="$$1" output="&lt;outline&gt;\1&lt;/outline&gt;&lt;plot&gt;\1&lt;/plot&gt;" dest="5+">
  106. <expression>Plot:&lt;/h5&gt;\n&lt;div class=&quot;info-content&quot;&gt;\n(.*?) \| &lt;a class=&quot;tn15more</expression>
  107. </RegExp>
  108. <RegExp input="$$2" output="&lt;url function=&quot;GetIMDBPlot&quot;&gt;$$3plotsummary&lt;/url&gt;" dest="5+">
  109. <expression/>
  110. </RegExp>
  111. <RegExp conditional="!fullcredits" input="$$2" output="&lt;url cache=&quot;$$2-credits.html&quot; function=&quot;GetIMDBCast&quot;&gt;$$3&lt;/url&gt;" dest="5+">
  112. <expression/>
  113. </RegExp>
  114. <RegExp conditional="!fullcredits" input="$$2" output="&lt;url cache=&quot;$$2-credits.html&quot; function=&quot;GetIMDBDirectors&quot;&gt;$$3&lt;/url&gt;" dest="5+">
  115. <expression/>
  116. </RegExp>
  117. <RegExp conditional="!fullcredits" input="$$2" output="&lt;url cache=&quot;$$2-credits.html&quot; function=&quot;GetIMDBWriters&quot;&gt;$$3&lt;/url&gt;" dest="5+">
  118. <expression/>
  119. </RegExp>
  120. <RegExp conditional="fullcredits" input="$$2" output="&lt;url cache=&quot;$$2-fullcredits.html&quot; function=&quot;GetIMDBCast&quot;&gt;$$3fullcredits&lt;/url&gt;" dest="5+">
  121. <expression/>
  122. </RegExp>
  123. <RegExp conditional="fullcredits" input="$$2" output="&lt;url cache=&quot;$$2-fullcredits.html&quot; function=&quot;GetIMDBDirectors&quot;&gt;$$3fullcredits&lt;/url&gt;" dest="5+">
  124. <expression/>
  125. </RegExp>
  126. <RegExp conditional="fullcredits" input="$$2" output="&lt;url cache=&quot;$$2-fullcredits.html&quot; function=&quot;GetIMDBWriters&quot;&gt;$$3fullcredits&lt;/url&gt;" dest="5+">
  127. <expression/>
  128. </RegExp>
  129. <RegExp conditional="tmdbthumbs" input="$$2" output="&lt;url function=&quot;GetTMDBThumbsById&quot;&gt;$$3&lt;/url&gt;" dest="5+">
  130. <expression/>
  131. </RegExp>
  132. <RegExp conditional="impawards" input="$$2" output="&lt;url cache=&quot;$$2-posters.html&quot; function=&quot;GetIMPALink&quot;&gt;$$3posters&lt;/url&gt;" dest="5+">
  133. <expression/>
  134. </RegExp>
  135. <RegExp conditional="movieposterdb" input="$$1" output="&lt;url function=&quot;GetMoviePosterDBLink&quot;&gt;http://www.movieposterdb.com/browse/search?type=movies&amp;amp;query=\1&lt;/url&gt;" dest="5+">
  136. <expression>/title/tt([t0-9]*)/faq</expression>
  137. </RegExp>
  138. <RegExp conditional="trailer" input="$$1" output="&lt;url function=&quot;GetIMDBTrailer&quot;&gt;http://akas.imdb.com/video/imdb/vi\1/player&lt;/url&gt;" dest="5+">
  139. <expression>/vi([0-9]*)/</expression>
  140. </RegExp>
  141. <RegExp conditional="dtrailer" input="$$6" output="&lt;url function=&quot;GetDTrailerLink&quot;&gt;http://en.dtrailer.com/movies/search/\1&lt;/url&gt;" dest="5+">
  142. <RegExp input="$$4" output="\1-" dest="6">
  143. <RegExp input="$$1" output="\1" dest="4">
  144. <expression trim="1" noclean="1">&lt;h1&gt;([^&lt;]*)</expression>
  145. </RegExp>
  146. <expression repeat="yes">([a-zA-Z0-9]+)</expression>
  147. </RegExp>
  148. <expression/>
  149. </RegExp>
  150. <RegExp input="$$2" output="&lt;url cache=&quot;$$2-posters.html&quot; function=&quot;GetIMDBThumbs&quot;&gt;$$3posters&lt;/url&gt;" dest="5+">
  151. <expression/>
  152. </RegExp>
  153. <RegExp conditional="fanart" input="$$2" output="&lt;url function=&quot;GetTMDBFanartById&quot;&gt;$$3&lt;/url&gt;" dest="5+">
  154. <expression/>
  155. </RegExp>
  156. <expression noclean="1"/>
  157. </RegExp>
  158. </GetDetails>
  159. </scraper>
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement
Public Pastes
Advertisement
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!