debian/README.source: Change 'markup'.

[youtubedl] / youtube_dl / downloader / f4m.py

diff --git a/youtube_dl/downloader/f4m.py b/youtube_dl/downloader/f4m.py
index 688e086eb0536c55ef184ae68fa09a6ffb41462d..8dd3c2eeb3154b98e48cf6a58fe50ca50943dada 100644 (file)
--- a/youtube_dl/downloader/f4m.py
+++ b/youtube_dl/downloader/f4m.py
@@ -1,13 +1,12 @@
 from __future__ import division, unicode_literals
 
 from __future__ import division, unicode_literals
 

-import base64

 import io
 import itertools
 import io
 import itertools

-import os

 import time
 
 from .fragment import FragmentFD
 from ..compat import (
 import time
 
 from .fragment import FragmentFD
 from ..compat import (

+    compat_b64decode,

     compat_etree_fromstring,
     compat_urlparse,
     compat_urllib_error,
     compat_etree_fromstring,
     compat_urlparse,
     compat_urllib_error,


@@ -16,9 +15,7 @@ from ..compat import (
     compat_struct_unpack,
 )
 from ..utils import (
     compat_struct_unpack,
 )
 from ..utils import (

-    encodeFilename,

     fix_xml_ampersands,
     fix_xml_ampersands,

-    sanitize_open,

     xpath_text,
 )
 
     xpath_text,
 )
 


@@ -241,13 +238,22 @@ def write_metadata_tag(stream, metadata):
 
 
 def remove_encrypted_media(media):
 
 
 def remove_encrypted_media(media):

-    return list(filter(lambda e: 'drmAdditionalHeaderId' not in e.attrib and
-                                 'drmAdditionalHeaderSetId' not in e.attrib,
+    return list(filter(lambda e: 'drmAdditionalHeaderId' not in e.attrib
+                                 and 'drmAdditionalHeaderSetId' not in e.attrib,

                        media))
 
 
                        media))
 
 

-def _add_ns(prop):
-    return '{http://ns.adobe.com/f4m/1.0}%s' % prop
+def _add_ns(prop, ver=1):
+    return '{http://ns.adobe.com/f4m/%d.0}%s' % (ver, prop)
+
+
+def get_base_url(manifest):
+    base_url = xpath_text(
+        manifest, [_add_ns('baseURL'), _add_ns('baseURL', 2)],
+        'base URL', default=None)
+    if base_url:
+        base_url = base_url.strip()
+    return base_url

 
 
 class F4mFD(FragmentFD):
 
 
 class F4mFD(FragmentFD):


@@ -261,8 +267,8 @@ class F4mFD(FragmentFD):
         media = doc.findall(_add_ns('media'))
         if not media:
             self.report_error('No media found')
         media = doc.findall(_add_ns('media'))
         if not media:
             self.report_error('No media found')

-        for e in (doc.findall(_add_ns('drmAdditionalHeader')) +
-                  doc.findall(_add_ns('drmAdditionalHeaderSet'))):
+        for e in (doc.findall(_add_ns('drmAdditionalHeader'))
+                  + doc.findall(_add_ns('drmAdditionalHeaderSet'))):

             # If id attribute is missing it's valid for all media nodes
             # without drmAdditionalHeaderId or drmAdditionalHeaderSetId attribute
             if 'id' not in e.attrib:
             # If id attribute is missing it's valid for all media nodes
             # without drmAdditionalHeaderId or drmAdditionalHeaderSetId attribute
             if 'id' not in e.attrib:


@@ -306,7 +312,7 @@ class F4mFD(FragmentFD):
             boot_info = self._get_bootstrap_from_url(bootstrap_url)
         else:
             bootstrap_url = None
             boot_info = self._get_bootstrap_from_url(bootstrap_url)
         else:
             bootstrap_url = None

-            bootstrap = base64.b64decode(node.text.encode('ascii'))
+            bootstrap = compat_b64decode(node.text)

             boot_info = read_bootstrap_info(bootstrap)
         return boot_info, bootstrap_url
 
             boot_info = read_bootstrap_info(bootstrap)
         return boot_info, bootstrap_url
 


@@ -318,8 +324,8 @@ class F4mFD(FragmentFD):
         urlh = self.ydl.urlopen(self._prepare_url(info_dict, man_url))
         man_url = urlh.geturl()
         # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
         urlh = self.ydl.urlopen(self._prepare_url(info_dict, man_url))
         man_url = urlh.geturl()
         # Some manifests may be malformed, e.g. prosiebensat1 generated manifests

-        # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244
-        # and https://github.com/rg3/youtube-dl/issues/7823)
+        # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244
+        # and https://github.com/ytdl-org/youtube-dl/issues/7823)

         manifest = fix_xml_ampersands(urlh.read().decode('utf-8', 'ignore')).strip()
 
         doc = compat_etree_fromstring(manifest)
         manifest = fix_xml_ampersands(urlh.read().decode('utf-8', 'ignore')).strip()
 
         doc = compat_etree_fromstring(manifest)


@@ -333,17 +339,17 @@ class F4mFD(FragmentFD):
             rate, media = list(filter(
                 lambda f: int(f[0]) == requested_bitrate, formats))[0]
 
             rate, media = list(filter(
                 lambda f: int(f[0]) == requested_bitrate, formats))[0]
 

-        base_url = compat_urlparse.urljoin(man_url, media.attrib['url'])
+        # Prefer baseURL for relative URLs as per 11.2 of F4M 3.0 spec.
+        man_base_url = get_base_url(doc) or man_url
+
+        base_url = compat_urlparse.urljoin(man_base_url, media.attrib['url'])

         bootstrap_node = doc.find(_add_ns('bootstrapInfo'))
         bootstrap_node = doc.find(_add_ns('bootstrapInfo'))

-        # From Adobe F4M 3.0 spec:
-        # The <baseURL> element SHALL be the base URL for all relative
-        # (HTTP-based) URLs in the manifest. If <baseURL> is not present, said
-        # URLs should be relative to the location of the containing document.
-        boot_info, bootstrap_url = self._parse_bootstrap_node(bootstrap_node, man_url)
+        boot_info, bootstrap_url = self._parse_bootstrap_node(
+            bootstrap_node, man_base_url)

         live = boot_info['live']
         metadata_node = media.find(_add_ns('metadata'))
         if metadata_node is not None:
         live = boot_info['live']
         metadata_node = media.find(_add_ns('metadata'))
         if metadata_node is not None:

-            metadata = base64.b64decode(metadata_node.text.encode('ascii'))
+            metadata = compat_b64decode(metadata_node.text)

         else:
             metadata = None
 
         else:
             metadata = None
 


@@ -366,17 +372,21 @@ class F4mFD(FragmentFD):
 
         dest_stream = ctx['dest_stream']
 
 
         dest_stream = ctx['dest_stream']
 

-        write_flv_header(dest_stream)
-        if not live:
-            write_metadata_tag(dest_stream, metadata)
+        if ctx['complete_frags_downloaded_bytes'] == 0:
+            write_flv_header(dest_stream)
+            if not live:
+                write_metadata_tag(dest_stream, metadata)

 
         base_url_parsed = compat_urllib_parse_urlparse(base_url)
 
         self._start_frag_download(ctx)
 
 
         base_url_parsed = compat_urllib_parse_urlparse(base_url)
 
         self._start_frag_download(ctx)
 

-        frags_filenames = []
+        frag_index = 0

         while fragments_list:
             seg_i, frag_i = fragments_list.pop(0)
         while fragments_list:
             seg_i, frag_i = fragments_list.pop(0)

+            frag_index += 1
+            if frag_index <= ctx['fragment_index']:
+                continue

             name = 'Seg%d-Frag%d' % (seg_i, frag_i)
             query = []
             if base_url_parsed.query:
             name = 'Seg%d-Frag%d' % (seg_i, frag_i)
             query = []
             if base_url_parsed.query:


@@ -386,17 +396,10 @@ class F4mFD(FragmentFD):
             if info_dict.get('extra_param_to_segment_url'):
                 query.append(info_dict['extra_param_to_segment_url'])
             url_parsed = base_url_parsed._replace(path=base_url_parsed.path + name, query='&'.join(query))
             if info_dict.get('extra_param_to_segment_url'):
                 query.append(info_dict['extra_param_to_segment_url'])
             url_parsed = base_url_parsed._replace(path=base_url_parsed.path + name, query='&'.join(query))

-            frag_filename = '%s-%s' % (ctx['tmpfilename'], name)

             try:
             try:

-                success = ctx['dl'].download(frag_filename, {
-                    'url': url_parsed.geturl(),
-                    'http_headers': info_dict.get('http_headers'),
-                })
+                success, down_data = self._download_fragment(ctx, url_parsed.geturl(), info_dict)

                 if not success:
                     return False
                 if not success:
                     return False

-                (down, frag_sanitized) = sanitize_open(frag_filename, 'rb')
-                down_data = down.read()
-                down.close()

                 reader = FlvReader(down_data)
                 while True:
                     try:
                 reader = FlvReader(down_data)
                 while True:
                     try:


@@ -406,17 +409,13 @@ class F4mFD(FragmentFD):
                             # In tests, segments may be truncated, and thus
                             # FlvReader may not be able to parse the whole
                             # chunk. If so, write the segment as is
                             # In tests, segments may be truncated, and thus
                             # FlvReader may not be able to parse the whole
                             # chunk. If so, write the segment as is

-                            # See https://github.com/rg3/youtube-dl/issues/9214
+                            # See https://github.com/ytdl-org/youtube-dl/issues/9214

                             dest_stream.write(down_data)
                             break
                         raise
                     if box_type == b'mdat':
                             dest_stream.write(down_data)
                             break
                         raise
                     if box_type == b'mdat':

-                        dest_stream.write(box_data)
+                        self._append_fragment(ctx, box_data)

                         break
                         break

-                if live:
-                    os.remove(encodeFilename(frag_sanitized))
-                else:
-                    frags_filenames.append(frag_sanitized)

             except (compat_urllib_error.HTTPError, ) as err:
                 if live and (err.code == 404 or err.code == 410):
                     # We didn't keep up with the live window. Continue
             except (compat_urllib_error.HTTPError, ) as err:
                 if live and (err.code == 404 or err.code == 410):
                     # We didn't keep up with the live window. Continue


@@ -436,7 +435,4 @@ class F4mFD(FragmentFD):
 
         self._finish_frag_download(ctx)
 
 
         self._finish_frag_download(ctx)
 

-        for frag_file in frags_filenames:
-            os.remove(encodeFilename(frag_file))
-

         return True
         return True