Support catchall for sub language without country

[gradha]

When the --sub-lang switch is specified with a value of en it doesn't download subs labelled with a country code like en-US. See this session log:

  $ youtube-dl  -U
  Updating to version 2014.01.17.2...
  Updated youtube-dl. Restart youtube-dl to use the new version.
  $ youtube-dl  --skip-download --write-sub --sub-lang en WR9-GRQfEkU
  [youtube] Setting language
  [youtube] WR9-GRQfEkU: Downloading webpage
  [youtube] WR9-GRQfEkU: Downloading video info webpage
  [youtube] WR9-GRQfEkU: Extracting video information
  WARNING: no closed captions found in the specified language "en"
  [youtube] WR9-GRQfEkU: Encrypted signatures detected.
  $ youtube-dl  --skip-download --list-subs WR9-GRQfEkU
  [youtube] Setting language
  [youtube] WR9-GRQfEkU: Downloading webpage
  [youtube] WR9-GRQfEkU: Downloading video info webpage
  [youtube] WR9-GRQfEkU: Extracting video information
  [youtube] WR9-GRQfEkU: Looking for automatic captions
  [youtube] WR9-GRQfEkU: Downloading XML
  WARNING: Video doesn't have automatic captions
  [youtube] WR9-GRQfEkU: Available subtitles for video: en-US
  [youtube] WR9-GRQfEkU: Available automatic captions for video: 
  $

It would be great if failing a perfect match the program downloaded whatever first entry matched partially the user specified value, changing the warning to:

WARNING: no closed captions found in the specified language "en", found partial match for "en-US"

Or maybe add the partial match behaviour as an additional switch if it breaks behaviour people depend on (aka: specifying a value and failing even if partial matches are found).

[gradha]

Untested prototype algorithm follows:

diff --git a/youtube-dl b/youtube-dl
index e3eb877..1e06aaa 100755
Binary files a/youtube-dl and b/youtube-dl differ
diff --git a/youtube_dl/extractor/subtitles.py b/youtube_dl/extractor/subtitles.py
index 4b4c523..7ab94e6 100644
--- a/youtube_dl/extractor/subtitles.py
+++ b/youtube_dl/extractor/subtitles.py
@@ -51,8 +51,13 @@ class SubtitlesInfoExtractor(InfoExtractor):
             sub_lang_list = {}
             for sub_lang in requested_langs:
                 if not sub_lang in available_subs_list:
-                    self._downloader.report_warning(u'no closed captions found in the specified language "%s"' % sub_lang)
-                    continue
+                    partial_dict = dict([(x[:len(sub_lang)], x) for x in reversed(available_subs_list)])
+                    if sub_lang in partial_dict:
+                        self._downloader.report_warning(u'no closed captions found in the specified language "%s", found partial match for "%s"' % (sub_lang, sub_lang))
+                        sub_lang = partial_dict[sub_lang]
+                    else:
+                        self._downloader.report_warning(u'no closed captions found in the specified language "%s"' % sub_lang)
+                        continue
                 sub_lang_list[sub_lang] = available_subs_list[sub_lang]

         subtitles = {}