X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fdownloader%2Fhttp.py;h=56840e026cc710a7c709da13f637e62842ccfd84;hb=5c2266df4b9aeb7881ed8c026a038e2a25e43734;hp=ceacb8522b79ec4e4d3fbef83e242ed457be4a7c;hpb=6800d3372f35e08dcc4d34d06601815bf0cb0a3d;p=youtube-dl diff --git a/youtube_dl/downloader/http.py b/youtube_dl/downloader/http.py index ceacb8522..56840e026 100644 --- a/youtube_dl/downloader/http.py +++ b/youtube_dl/downloader/http.py @@ -4,25 +4,20 @@ import errno import os import socket import time +import re from .common import FileDownloader -from .dash import DashSegmentsFD -from ..compat import ( - compat_urllib_request, - compat_urllib_error, -) +from ..compat import compat_urllib_error from ..utils import ( ContentTooShortError, encodeFilename, sanitize_open, + sanitized_Request, ) class HttpFD(FileDownloader): def real_download(self, filename, info_dict): - if info_dict.get('initialization_url') and list(filter(None, info_dict.get('segment_urls', []))): - return DashSegmentsFD(self.ydl, self.params).real_download(filename, info_dict) - url = info_dict['url'] tmpfilename = self.temp_name(filename) stream = None @@ -32,8 +27,8 @@ class HttpFD(FileDownloader): add_headers = info_dict.get('http_headers') if add_headers: headers.update(add_headers) - basic_request = compat_urllib_request.Request(url, None, headers) - request = compat_urllib_request.Request(url, None, headers) + basic_request = sanitized_Request(url, None, headers) + request = sanitized_Request(url, None, headers) is_test = self.params.get('test', False) @@ -61,6 +56,24 @@ class HttpFD(FileDownloader): # Establish connection try: data = self.ydl.urlopen(request) + # When trying to resume, Content-Range HTTP header of response has to be checked + # to match the value of requested Range HTTP header. This is due to a webservers + # that don't support resuming and serve a whole file with no Content-Range + # set in response despite of requested Range (see + # https://github.com/rg3/youtube-dl/issues/6057#issuecomment-126129799) + if resume_len > 0: + content_range = data.headers.get('Content-Range') + if content_range: + content_range_m = re.search(r'bytes (\d+)-', content_range) + # Content-Range is present and matches requested Range, resume is possible + if content_range_m and resume_len == int(content_range_m.group(1)): + break + # Content-Range is either not present or invalid. Assuming remote webserver is + # trying to send the whole file, resume is not possible, so wiping the local file + # and performing entire redownload + self.report_unable_to_resume() + resume_len = 0 + open_mode = 'wb' break except (compat_urllib_error.HTTPError, ) as err: if (err.code < 500 or err.code >= 600) and err.code != 416: