fix The Daily Show/The Colbert Report

src/main/resources/scripts/youtube_dl_concat.groovy

@@ -1,3 +1,5 @@
+import groovyx.net.http.URIBuilder
+
 // videofeed.Web,Comedy Central=http://extechops.net/full-episode-feeds/daily-show.rss
 // videofeed.Web,Comedy Central=http://extechops.net/full-episode-feeds/colbert-report.rss
 
@@ -6,16 +8,59 @@ script (INIT) {
     def FILENAME = 'pmsencoder_ffmpeg_concat_%d_%d.txt'
     def TEMP_FOLDER = pms.getConfiguration().getTempFolder()
 
+    // XXX not sure if this is needed for the previously-used domains (without the .cc)
+    // or even if they're still used
+    profile ('Comedy Central: /episodes/ -> /full-episodes/', stopOnMatch: false) {
+        pattern {
+            // this is a workaround for a youtube-dl bug (or at least a missing feature)
+            match { YOUTUBE_DL_PATH }
+            domains([ 'thedailyshow.cc.com', 'thecolbertreport.cc.com' ])
+        }
+
+        /*
+         * XXX youtube-dl doesn't grok Comedy Central pages whose paths start with /episodes, but does handle
+         * those that start with /full-episodes, so replace the former with the latter:
+         *
+         *     from: http://thedailyshow.com/episodes/wqf1x5/april-2--2014---samuel-l--jackson
+         *     to:   http://thedailyshow.com/full-episodes/wqf1x5/april-2--2014---samuel-l--jackson
+         */
+        action {
+            /*
+             * XXX this would be much clearer/simpler if path was (also) exposed as a List e.g.:
+             *
+             *     def u = new URIBuilder(uri)
+             *
+             *     if (u.steps[0] == 'episodes') {
+             *         u.steps[0] = 'full-episodes'
+             *         uri = u.toString()
+             *     }
+             */
+            def u = new URIBuilder(uri)
+            def steps = u.path.split('/')
+
+            // path starts with a forward slash (and split preserves it), so steps is e.g.:
+            // [ '', 'episodes', 'wqf1x5', 'april-2--2014---samuel-l--jackson' ]
+            if (steps[1] == 'episodes') {
+                steps[1] = 'full-episodes'
+                u.path = steps.join('/')
+                uri = u.toString()
+            }
+        }
+    }
+
     profile ('YouTube-DL Concat') {
         pattern {
             match { YOUTUBE_DL_PATH }
-            domains([ 'thedailyshow.com', 'colbertnation.com' ])
+            domains([ 'thedailyshow.com', 'thedailyshow.cc.com', 'colbertnation.com', 'thecolbertreport.cc.com' ])
         }
 
         action {
-            // doesn't use cmd.exe, so no need to quote the URI
+            // this command doesn't use cmd.exe, so no need to quote the URI
             def command = YOUTUBE_DL_PATH + [ '-g', uri ]
-            def uris = command.execute().text.readLines(); logger.debug("URIs: ${uris.inspect()}")
+            def uris = command.execute().text.readLines()
+
+            logger.debug("URIs: ${uris.inspect()}")
+
             def files = uris.collect({ "file '$it'" }).join(SEPARATOR)
             def filename = String.format(FILENAME, Thread.currentThread().getId(), System.currentTimeMillis())
             def file = new File(TEMP_FOLDER, filename); logger.trace("concat file: ${file.getAbsolutePath()}")