[downloader/dash:hls] Respect --fragment-retries and --skip-unavailable-fragments...
authorSergey M․ <dstftw@gmail.com>
Fri, 26 Aug 2016 21:55:55 +0000 (04:55 +0700)
committerSergey M․ <dstftw@gmail.com>
Sat, 3 Sep 2016 10:51:48 +0000 (17:51 +0700)
youtube_dl/downloader/dash.py
youtube_dl/downloader/hls.py

index 8bbab9dbc596c659db622fe9910d0ae90018a598..cbcee324df957c36cdcd00ef7e49a4c30bf0d949 100644 (file)
@@ -38,6 +38,7 @@ class DashSegmentsFD(FragmentFD):
         segments_filenames = []
 
         fragment_retries = self.params.get('fragment_retries', 0)
+        skip_unavailable_fragments = self.params.get('skip_unavailable_fragments', True)
 
         def append_url_to_file(target_url, tmp_filename, segment_name):
             target_filename = '%s-%s' % (tmp_filename, segment_name)
@@ -52,19 +53,20 @@ class DashSegmentsFD(FragmentFD):
                     down.close()
                     segments_filenames.append(target_sanitized)
                     break
-                except (compat_urllib_error.HTTPError, ) as err:
+                except compat_urllib_error.HTTPError:
                     # YouTube may often return 404 HTTP error for a fragment causing the
                     # whole download to fail. However if the same fragment is immediately
                     # retried with the same request data this usually succeeds (1-2 attemps
                     # is usually enough) thus allowing to download the whole file successfully.
-                    # So, we will retry all fragments that fail with 404 HTTP error for now.
-                    if err.code != 404:
-                        raise
-                    # Retry fragment
+                    # To be future-proof we will retry all fragments that fail with any
+                    # HTTP error.
                     count += 1
                     if count <= fragment_retries:
                         self.report_retry_fragment(segment_name, count, fragment_retries)
             if count > fragment_retries:
+                if skip_unavailable_fragments:
+                    self.report_skip_fragment(segment_name)
+                    return
                 self.report_error('giving up after %s fragment retries' % fragment_retries)
                 return False
 
index baaff44d5a49de5fec84d4181fa9a4becbc35e9c..7412620a571cdcd9bdca261388d38f0ffc9f075b 100644 (file)
@@ -13,6 +13,7 @@ from .fragment import FragmentFD
 from .external import FFmpegFD
 
 from ..compat import (
+    compat_urllib_error,
     compat_urlparse,
     compat_struct_pack,
 )
@@ -83,6 +84,10 @@ class HlsFD(FragmentFD):
 
         self._prepare_and_start_frag_download(ctx)
 
+        fragment_retries = self.params.get('fragment_retries', 0)
+        skip_unavailable_fragments = self.params.get('skip_unavailable_fragments', True)
+        test = self.params.get('test', False)
+
         extra_query = None
         extra_param_to_segment_url = info_dict.get('extra_param_to_segment_url')
         if extra_param_to_segment_url:
@@ -99,15 +104,37 @@ class HlsFD(FragmentFD):
                         line
                         if re.match(r'^https?://', line)
                         else compat_urlparse.urljoin(man_url, line))
-                    frag_filename = '%s-Frag%d' % (ctx['tmpfilename'], i)
+                    frag_name = 'Frag%d' % i
+                    frag_filename = '%s-%s' % (ctx['tmpfilename'], frag_name)
                     if extra_query:
                         frag_url = update_url_query(frag_url, extra_query)
-                    success = ctx['dl'].download(frag_filename, {'url': frag_url})
-                    if not success:
+                    count = 0
+                    while count <= fragment_retries:
+                        try:
+                            success = ctx['dl'].download(frag_filename, {'url': frag_url})
+                            if not success:
+                                return False
+                            down, frag_sanitized = sanitize_open(frag_filename, 'rb')
+                            frag_content = down.read()
+                            down.close()
+                            break
+                        except compat_urllib_error.HTTPError:
+                            # Unavailable (possibly temporary) fragments may be served.
+                            # First we try to retry then either skip or abort.
+                            # See https://github.com/rg3/youtube-dl/issues/10165,
+                            # https://github.com/rg3/youtube-dl/issues/10448).
+                            count += 1
+                            if count <= fragment_retries:
+                                self.report_retry_fragment(frag_name, count, fragment_retries)
+                    if count > fragment_retries:
+                        if skip_unavailable_fragments:
+                            i += 1
+                            media_sequence += 1
+                            self.report_skip_fragment(frag_name)
+                            continue
+                        self.report_error(
+                            'giving up after %s fragment retries' % fragment_retries)
                         return False
-                    down, frag_sanitized = sanitize_open(frag_filename, 'rb')
-                    frag_content = down.read()
-                    down.close()
                     if decrypt_info['METHOD'] == 'AES-128':
                         iv = decrypt_info.get('IV') or compat_struct_pack('>8xq', media_sequence)
                         frag_content = AES.new(
@@ -115,7 +142,7 @@ class HlsFD(FragmentFD):
                     ctx['dest_stream'].write(frag_content)
                     frags_filenames.append(frag_sanitized)
                     # We only download the first fragment during the test
-                    if self.params.get('test', False):
+                    if test:
                         break
                     i += 1
                     media_sequence += 1