From: Sergey M․ Date: Sat, 15 Aug 2015 10:32:06 +0000 (+0600) Subject: Merge branch 'shahid' of https://github.com/remitamine/youtube-dl into remitamine... X-Git-Url: http://git.bitcoin.ninja/index.cgi?p=youtube-dl;a=commitdiff_plain;h=11bed5827dace09b5483b159476ce9f8c29d6078;hp=59e89e62d7b45554cef502dc4986f35618110679 Merge branch 'shahid' of https://github.com/remitamine/youtube-dl into remitamine-shahid --- diff --git a/AUTHORS b/AUTHORS index 531ec5767..d1693224e 100644 --- a/AUTHORS +++ b/AUTHORS @@ -132,3 +132,11 @@ George Brighton Remita Amine Aurélio A. Heckert Bernhard Minks +sceext +Zach Bruggeman +Tjark Saul +slangangular +Behrouz Abbasi +ngld +nyuszika7h +Shaun Walbridge diff --git a/README.md b/README.md index a2cc89cdb..542a7c26a 100644 --- a/README.md +++ b/README.md @@ -75,7 +75,7 @@ which means you can modify it, redistribute it or use it however you like. ## Video Selection: --playlist-start NUMBER Playlist video to start at (default is 1) --playlist-end NUMBER Playlist video to end at (default is last) - --playlist-items ITEM_SPEC Playlist video items to download. Specify indices of the videos in the playlist seperated by commas like: "--playlist-items 1,2,5,8" + --playlist-items ITEM_SPEC Playlist video items to download. Specify indices of the videos in the playlist separated by commas like: "--playlist-items 1,2,5,8" if you want to download videos indexed 1, 2, 5, 8 in the playlist. You can specify range: "--playlist-items 1-3,7,10-13", it will download the videos at index 1, 2, 3, 7, 10, 11, 12 and 13. --match-title REGEX Download only matching titles (regex or caseless sub-string) @@ -108,7 +108,7 @@ which means you can modify it, redistribute it or use it however you like. --playlist-reverse Download playlist videos in reverse order --xattr-set-filesize Set file xattribute ytdl.filesize with expected filesize (experimental) --hls-prefer-native Use the native HLS downloader instead of ffmpeg (experimental) - --external-downloader COMMAND Use the specified external downloader. Currently supports aria2c,curl,httpie,wget + --external-downloader COMMAND Use the specified external downloader. Currently supports aria2c,axel,curl,httpie,wget --external-downloader-args ARGS Give these arguments to the external downloader ## Filesystem Options: @@ -272,6 +272,7 @@ The `-o` option allows users to indicate a template for the output file names. T - `autonumber`: The sequence will be replaced by a five-digit number that will be increased with each download, starting at zero. - `playlist`: The name or the id of the playlist that contains the video. - `playlist_index`: The index of the video in the playlist, a five-digit number. + - `format_id`: The sequence will be replaced by the format code specified by `--format`. The current default template is `%(title)s-%(id)s.%(ext)s`. @@ -289,7 +290,7 @@ youtube-dl_test_video_.mp4 # A simple file name By default youtube-dl tries to download the best quality, but sometimes you may want to download other format. The simplest case is requesting a specific format, for example `-f 22`. You can get the list of available formats using `--list-formats`, you can also use a file extension (currently it supports aac, m4a, mp3, mp4, ogg, wav, webm) or the special names `best`, `bestvideo`, `bestaudio` and `worst`. -If you want to download multiple videos and they don't have the same formats available, you can specify the order of preference using slashes, as in `-f 22/17/18`. You can also filter the video results by putting a condition in brackets, as in `-f "best[height=720]"` (or `-f "[filesize>10M]"`). This works for filesize, height, width, tbr, abr, vbr, asr, and fps and the comparisons <, <=, >, >=, =, != and for ext, acodec, vcodec, container, and protocol and the comparisons =, != . Formats for which the value is not known are excluded unless you put a question mark (?) after the operator. You can combine format filters, so `-f "[height <=? 720][tbr>500]"` selects up to 720p videos (or videos where the height is not known) with a bitrate of at least 500 KBit/s. Use commas to download multiple formats, such as `-f 136/137/mp4/bestvideo,140/m4a/bestaudio`. You can merge the video and audio of two formats into a single file using `-f +` (requires ffmpeg or avconv), for example `-f bestvideo+bestaudio`. +If you want to download multiple videos and they don't have the same formats available, you can specify the order of preference using slashes, as in `-f 22/17/18`. You can also filter the video results by putting a condition in brackets, as in `-f "best[height=720]"` (or `-f "[filesize>10M]"`). This works for filesize, height, width, tbr, abr, vbr, asr, and fps and the comparisons <, <=, >, >=, =, != and for ext, acodec, vcodec, container, and protocol and the comparisons =, != . Formats for which the value is not known are excluded unless you put a question mark (?) after the operator. You can combine format filters, so `-f "[height <=? 720][tbr>500]"` selects up to 720p videos (or videos where the height is not known) with a bitrate of at least 500 KBit/s. Use commas to download multiple formats, such as `-f 136/137/mp4/bestvideo,140/m4a/bestaudio`. You can merge the video and audio of two formats into a single file using `-f +` (requires ffmpeg or avconv), for example `-f bestvideo+bestaudio`. Format selectors can also be grouped using parentheses, for example if you want to download the best mp4 and webm formats with a height lower than 480 you can use `-f '(mp4,webm)[height<480]'`. Since the end of April 2015 and version 2015.04.26 youtube-dl uses `-f bestvideo+bestaudio/best` as default format selection (see #5447, #5456). If ffmpeg or avconv are installed this results in downloading `bestvideo` and `bestaudio` separately and muxing them together into a single file giving the best overall quality available. Otherwise it falls back to `best` and results in downloading best available quality served as a single file. `best` is also needed for videos that don't come from YouTube because they don't provide the audio and video in two different files. If you want to only download some dash formats (for example if you are not interested in getting videos with a resolution higher than 1080p), you can add `-f bestvideo[height<=?1080]+bestaudio/best` to your configuration file. Note that if you use youtube-dl to stream to `stdout` (and most likely to pipe it to your media player then), i.e. you explicitly specify output template as `-o -`, youtube-dl still uses `-f best` format selection in order to start content delivery immediately to your player and not to wait until `bestvideo` and `bestaudio` are downloaded and muxed. @@ -439,6 +440,12 @@ Either prepend `http://www.youtube.com/watch?v=` or separate the ID from the opt youtube-dl -- -wNyEUrxzFU youtube-dl "http://www.youtube.com/watch?v=-wNyEUrxzFU" +### How do I pass cookies to youtube-dl? + +Use the `--cookies` option, for example `--cookies /path/to/cookies/file.txt`. Note that cookies file must be in Mozilla/Netscape format and the first line of cookies file must be either `# HTTP Cookie File` or `# Netscape HTTP Cookie File`. Make sure you have correct [newline format](https://en.wikipedia.org/wiki/Newline) in cookies file and convert newlines if necessary to correspond your OS, namely `CRLF` (`\r\n`) for Windows, `LF` (`\n`) for Linux and `CR` (`\r`) for Mac OS. `HTTP Error 400: Bad Request` when using `--cookies` is a good sign of invalid newline format. + +Passing cookies to youtube-dl is a good way to workaround login when particular extractor does not implement it explicitly. + ### Can you add support for this anime video site, or site which shows current movies for free? As a matter of policy (as well as legality), youtube-dl does not include support for services that specialize in infringing copyright. As a rule of thumb, if you cannot easily find a video that the service is quite obviously allowed to distribute (i.e. that has been uploaded by the creator, the creator's distributor, or is published under a free license), the service is probably unfit for inclusion to youtube-dl. @@ -538,7 +545,7 @@ If you want to add support for a new site, you can follow this quick list (assum ``` 5. Add an import in [`youtube_dl/extractor/__init__.py`](https://github.com/rg3/youtube-dl/blob/master/youtube_dl/extractor/__init__.py). 6. Run `python test/test_download.py TestDownload.test_YourExtractor`. This *should fail* at first, but you can continually re-run it until you're done. If you decide to add more than one test, then rename ``_TEST`` to ``_TESTS`` and make it into a list of dictionaries. The tests will be then be named `TestDownload.test_YourExtractor`, `TestDownload.test_YourExtractor_1`, `TestDownload.test_YourExtractor_2`, etc. -7. Have a look at [`youtube_dl/common/extractor/common.py`](https://github.com/rg3/youtube-dl/blob/master/youtube_dl/extractor/common.py) for possible helper methods and a [detailed description of what your extractor should return](https://github.com/rg3/youtube-dl/blob/master/youtube_dl/extractor/common.py#L38). Add tests and code for as many as you want. +7. Have a look at [`youtube_dl/common/extractor/common.py`](https://github.com/rg3/youtube-dl/blob/master/youtube_dl/extractor/common.py) for possible helper methods and a [detailed description of what your extractor should and may return](https://github.com/rg3/youtube-dl/blob/master/youtube_dl/extractor/common.py#L62-L200). Add tests and code for as many as you want. 8. If you can, check the code with [flake8](https://pypi.python.org/pypi/flake8). 9. When the tests pass, [add](http://git-scm.com/docs/git-add) the new files and [commit](http://git-scm.com/docs/git-commit) them and [push](http://git-scm.com/docs/git-push) the result, like this: @@ -566,7 +573,7 @@ with youtube_dl.YoutubeDL(ydl_opts) as ydl: ydl.download(['http://www.youtube.com/watch?v=BaW_jenozKc']) ``` -Most likely, you'll want to use various options. For a list of what can be done, have a look at [youtube_dl/YoutubeDL.py](https://github.com/rg3/youtube-dl/blob/master/youtube_dl/YoutubeDL.py#L69). For a start, if you want to intercept youtube-dl's output, set a `logger` object. +Most likely, you'll want to use various options. For a list of what can be done, have a look at [youtube_dl/YoutubeDL.py](https://github.com/rg3/youtube-dl/blob/master/youtube_dl/YoutubeDL.py#L117-L265). For a start, if you want to intercept youtube-dl's output, set a `logger` object. Here's a more complete example of a program that outputs only errors (and a short message after the download is finished), and downloads/converts the video to an mp3 file: diff --git a/docs/supportedsites.md b/docs/supportedsites.md index a84878026..e21471102 100644 --- a/docs/supportedsites.md +++ b/docs/supportedsites.md @@ -28,6 +28,7 @@ - **anitube.se** - **AnySex** - **Aparat** + - **AppleConnect** - **AppleDaily**: 臺灣蘋果日報 - **AppleTrailers** - **archive.org**: archive.org videos @@ -50,6 +51,7 @@ - **bambuser:channel** - **Bandcamp** - **Bandcamp:album** + - **bbc**: BBC - **bbc.co.uk**: BBC iPlayer - **BeatportPro** - **Beeg** @@ -84,7 +86,7 @@ - **chirbit:profile** - **Cinchcast** - **Cinemassacre** - - **clipfish** + - **Clipfish** - **cliphunter** - **Clipsyndicate** - **Cloudy** @@ -114,6 +116,7 @@ - **DailymotionCloud** - **daum.net** - **DBTV** + - **DCN** - **DctpTv** - **DeezerPlaylist** - **defense.gouv.fr** @@ -223,6 +226,7 @@ - **InternetVideoArchive** - **IPrima** - **iqiyi**: 爱奇艺 + - **Ir90Tv** - **ivi**: ivi.ru - **ivi:compilation**: ivi.ru compilations - **Izlesene** @@ -251,6 +255,7 @@ - **kuwo:song**: 酷我音乐 - **la7.tv** - **Laola1Tv** + - **Lecture2Go** - **Letv**: 乐视网 - **LetvPlaylist** - **LetvTv** @@ -347,7 +352,6 @@ - **NowTV** - **nowvideo**: NowVideo - **npo**: npo.nl and ntr.nl - - **npo**: npo.nl and ntr.nl - **npo.nl:live** - **npo.nl:radio** - **npo.nl:radio:fragment** @@ -373,6 +377,7 @@ - **parliamentlive.tv**: UK parliament videos - **Patreon** - **PBS** + - **Periscope**: Periscope - **PhilharmonieDeParis**: Philharmonie de Paris - **Phoenix** - **Photobucket** @@ -402,6 +407,7 @@ - **qqmusic:playlist**: QQ音乐 - 歌单 - **qqmusic:singer**: QQ音乐 - 歌手 - **qqmusic:toplist**: QQ音乐 - 排行榜 + - **Quickscope**: Quick Scope - **QuickVid** - **R7** - **radio.de** @@ -486,6 +492,7 @@ - **SportBox** - **SportBoxEmbed** - **SportDeutschland** + - **Sportschau** - **Srf** - **SRMediathek**: Saarländischer Rundfunk - **SSA** @@ -513,6 +520,7 @@ - **ted** - **TeleBruxelles** - **telecinco.es** + - **Telegraaf** - **TeleMB** - **TeleTask** - **TenPlay** @@ -611,11 +619,12 @@ - **Vimple**: Vimple - one-click video hosting - **Vine** - **vine:user** - - **vk.com** - - **vk.com:user-videos**: vk.com:All of a user's videos + - **vk**: VK + - **vk:uservideos**: VK - User's Videos - **Vodlocker** - **VoiceRepublic** - **Vporn** + - **vpro**: npo.nl and ntr.nl - **VRT** - **vube**: Vube.com - **VuClip** diff --git a/test/helper.py b/test/helper.py index e1129e58f..cb6eec8d9 100644 --- a/test/helper.py +++ b/test/helper.py @@ -133,8 +133,8 @@ def expect_info_dict(self, got_dict, expected_dict): elif isinstance(expected, compat_str) and expected.startswith('mincount:'): got = got_dict.get(info_field) self.assertTrue( - isinstance(got, list), - 'Expected field %s to be a list, but it is of type %s' % ( + isinstance(got, (list, dict)), + 'Expected field %s to be a list or a dict, but it is of type %s' % ( info_field, type(got).__name__)) expected_num = int(expected.partition(':')[2]) assertGreaterEqual( @@ -160,7 +160,7 @@ def expect_info_dict(self, got_dict, expected_dict): # Are checkable fields missing from the test case definition? test_info_dict = dict((key, value if not isinstance(value, compat_str) or len(value) < 250 else 'md5:' + md5(value)) for key, value in got_dict.items() - if value and key in ('id', 'title', 'description', 'uploader', 'upload_date', 'timestamp', 'uploader_id', 'location')) + if value and key in ('id', 'title', 'description', 'uploader', 'upload_date', 'timestamp', 'uploader_id', 'location', 'age_limit')) missing_keys = set(test_info_dict.keys()) - set(expected_dict.keys()) if missing_keys: def _repr(v): diff --git a/test/test_YoutubeDL.py b/test/test_YoutubeDL.py index a13c09ef4..0388c0bf3 100644 --- a/test/test_YoutubeDL.py +++ b/test/test_YoutubeDL.py @@ -15,7 +15,7 @@ from youtube_dl import YoutubeDL from youtube_dl.compat import compat_str from youtube_dl.extractor import YoutubeIE from youtube_dl.postprocessor.common import PostProcessor -from youtube_dl.utils import match_filter_func +from youtube_dl.utils import ExtractorError, match_filter_func TEST_URL = 'http://localhost/sample.mp4' @@ -105,6 +105,7 @@ class TestFormatSelection(unittest.TestCase): def test_format_selection(self): formats = [ {'format_id': '35', 'ext': 'mp4', 'preference': 1, 'url': TEST_URL}, + {'format_id': 'example-with-dashes', 'ext': 'webm', 'preference': 1, 'url': TEST_URL}, {'format_id': '45', 'ext': 'webm', 'preference': 2, 'url': TEST_URL}, {'format_id': '47', 'ext': 'webm', 'preference': 3, 'url': TEST_URL}, {'format_id': '2', 'ext': 'flv', 'preference': 4, 'url': TEST_URL}, @@ -136,6 +137,11 @@ class TestFormatSelection(unittest.TestCase): downloaded = ydl.downloaded_info_dicts[0] self.assertEqual(downloaded['format_id'], '35') + ydl = YDL({'format': 'example-with-dashes'}) + ydl.process_ie_result(info_dict.copy()) + downloaded = ydl.downloaded_info_dicts[0] + self.assertEqual(downloaded['format_id'], 'example-with-dashes') + def test_format_selection_audio(self): formats = [ {'format_id': 'audio-low', 'ext': 'webm', 'preference': 1, 'vcodec': 'none', 'url': TEST_URL}, @@ -229,21 +235,70 @@ class TestFormatSelection(unittest.TestCase): '141', '172', '140', '171', '139', ] - for f1id, f2id in zip(order, order[1:]): - f1 = YoutubeIE._formats[f1id].copy() - f1['format_id'] = f1id - f1['url'] = 'url:' + f1id - f2 = YoutubeIE._formats[f2id].copy() - f2['format_id'] = f2id - f2['url'] = 'url:' + f2id + def format_info(f_id): + info = YoutubeIE._formats[f_id].copy() + info['format_id'] = f_id + info['url'] = 'url:' + f_id + return info + formats_order = [format_info(f_id) for f_id in order] + + info_dict = _make_result(list(formats_order), extractor='youtube') + ydl = YDL({'format': 'bestvideo+bestaudio'}) + yie = YoutubeIE(ydl) + yie._sort_formats(info_dict['formats']) + ydl.process_ie_result(info_dict) + downloaded = ydl.downloaded_info_dicts[0] + self.assertEqual(downloaded['format_id'], '137+141') + self.assertEqual(downloaded['ext'], 'mp4') + + info_dict = _make_result(list(formats_order), extractor='youtube') + ydl = YDL({'format': 'bestvideo[height>=999999]+bestaudio/best'}) + yie = YoutubeIE(ydl) + yie._sort_formats(info_dict['formats']) + ydl.process_ie_result(info_dict) + downloaded = ydl.downloaded_info_dicts[0] + self.assertEqual(downloaded['format_id'], '38') + + info_dict = _make_result(list(formats_order), extractor='youtube') + ydl = YDL({'format': 'bestvideo/best,bestaudio'}) + yie = YoutubeIE(ydl) + yie._sort_formats(info_dict['formats']) + ydl.process_ie_result(info_dict) + downloaded_ids = [info['format_id'] for info in ydl.downloaded_info_dicts] + self.assertEqual(downloaded_ids, ['137', '141']) + + info_dict = _make_result(list(formats_order), extractor='youtube') + ydl = YDL({'format': '(bestvideo[ext=mp4],bestvideo[ext=webm])+bestaudio'}) + yie = YoutubeIE(ydl) + yie._sort_formats(info_dict['formats']) + ydl.process_ie_result(info_dict) + downloaded_ids = [info['format_id'] for info in ydl.downloaded_info_dicts] + self.assertEqual(downloaded_ids, ['137+141', '248+141']) + + info_dict = _make_result(list(formats_order), extractor='youtube') + ydl = YDL({'format': '(bestvideo[ext=mp4],bestvideo[ext=webm])[height<=720]+bestaudio'}) + yie = YoutubeIE(ydl) + yie._sort_formats(info_dict['formats']) + ydl.process_ie_result(info_dict) + downloaded_ids = [info['format_id'] for info in ydl.downloaded_info_dicts] + self.assertEqual(downloaded_ids, ['136+141', '247+141']) + + info_dict = _make_result(list(formats_order), extractor='youtube') + ydl = YDL({'format': '(bestvideo[ext=none]/bestvideo[ext=webm])+bestaudio'}) + yie = YoutubeIE(ydl) + yie._sort_formats(info_dict['formats']) + ydl.process_ie_result(info_dict) + downloaded_ids = [info['format_id'] for info in ydl.downloaded_info_dicts] + self.assertEqual(downloaded_ids, ['248+141']) + for f1, f2 in zip(formats_order, formats_order[1:]): info_dict = _make_result([f1, f2], extractor='youtube') ydl = YDL({'format': 'best/bestvideo'}) yie = YoutubeIE(ydl) yie._sort_formats(info_dict['formats']) ydl.process_ie_result(info_dict) downloaded = ydl.downloaded_info_dicts[0] - self.assertEqual(downloaded['format_id'], f1id) + self.assertEqual(downloaded['format_id'], f1['format_id']) info_dict = _make_result([f2, f1], extractor='youtube') ydl = YDL({'format': 'best/bestvideo'}) @@ -251,7 +306,18 @@ class TestFormatSelection(unittest.TestCase): yie._sort_formats(info_dict['formats']) ydl.process_ie_result(info_dict) downloaded = ydl.downloaded_info_dicts[0] - self.assertEqual(downloaded['format_id'], f1id) + self.assertEqual(downloaded['format_id'], f1['format_id']) + + def test_invalid_format_specs(self): + def assert_syntax_error(format_spec): + ydl = YDL({'format': format_spec}) + info_dict = _make_result([{'format_id': 'foo', 'url': TEST_URL}]) + self.assertRaises(SyntaxError, ydl.process_ie_result, info_dict) + + assert_syntax_error('bestvideo,,best') + assert_syntax_error('+bestaudio') + assert_syntax_error('bestvideo+') + assert_syntax_error('/') def test_format_filtering(self): formats = [ @@ -308,6 +374,18 @@ class TestFormatSelection(unittest.TestCase): downloaded = ydl.downloaded_info_dicts[0] self.assertEqual(downloaded['format_id'], 'G') + ydl = YDL({'format': 'all[width>=400][width<=600]'}) + ydl.process_ie_result(info_dict) + downloaded_ids = [info['format_id'] for info in ydl.downloaded_info_dicts] + self.assertEqual(downloaded_ids, ['B', 'C', 'D']) + + ydl = YDL({'format': 'best[height<40]'}) + try: + ydl.process_ie_result(info_dict) + except ExtractorError: + pass + self.assertEqual(ydl.downloaded_info_dicts, []) + class TestYoutubeDL(unittest.TestCase): def test_subtitles(self): diff --git a/test/test_download.py b/test/test_download.py index 1110357a7..284418834 100644 --- a/test/test_download.py +++ b/test/test_download.py @@ -136,7 +136,9 @@ def generator(test_case): # We're not using .download here sine that is just a shim # for outside error handling, and returns the exit code # instead of the result dict. - res_dict = ydl.extract_info(test_case['url']) + res_dict = ydl.extract_info( + test_case['url'], + force_generic_extractor=params.get('force_generic_extractor', False)) except (DownloadError, ExtractorError) as err: # Check if the exception is not a network related one if not err.exc_info[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError, compat_http_client.BadStatusLine) or (err.exc_info[0] == compat_HTTPError and err.exc_info[1].code == 503): diff --git a/test/test_utils.py b/test/test_utils.py index e13e11b59..a759b2da9 100644 --- a/test/test_utils.py +++ b/test/test_utils.py @@ -235,12 +235,21 @@ class TestUtil(unittest.TestCase): + ''' doc = xml.etree.ElementTree.fromstring(testxml) + self.assertEqual(find_xpath_attr(doc, './/fourohfour', 'n'), None) self.assertEqual(find_xpath_attr(doc, './/fourohfour', 'n', 'v'), None) + self.assertEqual(find_xpath_attr(doc, './/node', 'n'), None) + self.assertEqual(find_xpath_attr(doc, './/node', 'n', 'v'), None) + self.assertEqual(find_xpath_attr(doc, './/node', 'x'), doc[1]) self.assertEqual(find_xpath_attr(doc, './/node', 'x', 'a'), doc[1]) + self.assertEqual(find_xpath_attr(doc, './/node', 'x', 'b'), doc[3]) + self.assertEqual(find_xpath_attr(doc, './/node', 'y'), doc[2]) self.assertEqual(find_xpath_attr(doc, './/node', 'y', 'c'), doc[2]) + self.assertEqual(find_xpath_attr(doc, './/node', 'y', 'd'), doc[3]) + self.assertEqual(find_xpath_attr(doc, './/node', 'x', ''), doc[4]) def test_xpath_with_ns(self): testxml = ''' @@ -324,6 +333,7 @@ class TestUtil(unittest.TestCase): self.assertEqual(parse_duration('02:03:04'), 7384) self.assertEqual(parse_duration('01:02:03:04'), 93784) self.assertEqual(parse_duration('1 hour 3 minutes'), 3780) + self.assertEqual(parse_duration('87 Min.'), 5220) def test_fix_xml_ampersands(self): self.assertEqual( diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py index 00af78e06..cad6b026e 100755 --- a/youtube_dl/YoutubeDL.py +++ b/youtube_dl/YoutubeDL.py @@ -21,24 +21,24 @@ import subprocess import socket import sys import time +import tokenize import traceback if os.name == 'nt': import ctypes from .compat import ( - compat_basestring, compat_cookiejar, compat_expanduser, compat_get_terminal_size, compat_http_client, compat_kwargs, compat_str, + compat_tokenize_tokenize, compat_urllib_error, compat_urllib_request, ) from .utils import ( - escape_url, ContentTooShortError, date_from_str, DateRange, @@ -49,7 +49,6 @@ from .utils import ( ExtractorError, format_bytes, formatSeconds, - HEADRequest, locked_file, make_HTTPS_handler, MaxDownloadsReached, @@ -853,8 +852,8 @@ class YoutubeDL(object): else: raise Exception('Invalid result type: %s' % result_type) - def _apply_format_filter(self, format_spec, available_formats): - " Returns a tuple of the remaining format_spec and filtered formats " + def _build_format_filter(self, filter_spec): + " Returns a function to filter the formats according to the filter_spec " OPERATORS = { '<': operator.lt, @@ -864,13 +863,13 @@ class YoutubeDL(object): '=': operator.eq, '!=': operator.ne, } - operator_rex = re.compile(r'''(?x)\s*\[ + operator_rex = re.compile(r'''(?x)\s* (?Pwidth|height|tbr|abr|vbr|asr|filesize|fps) \s*(?P%s)(?P\s*\?)?\s* (?P[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?) - \]$ + $ ''' % '|'.join(map(re.escape, OPERATORS.keys()))) - m = operator_rex.search(format_spec) + m = operator_rex.search(filter_spec) if m: try: comparison_value = int(m.group('value')) @@ -881,7 +880,7 @@ class YoutubeDL(object): if comparison_value is None: raise ValueError( 'Invalid value %r in format specification %r' % ( - m.group('value'), format_spec)) + m.group('value'), filter_spec)) op = OPERATORS[m.group('op')] if not m: @@ -889,85 +888,283 @@ class YoutubeDL(object): '=': operator.eq, '!=': operator.ne, } - str_operator_rex = re.compile(r'''(?x)\s*\[ + str_operator_rex = re.compile(r'''(?x) \s*(?Pext|acodec|vcodec|container|protocol) \s*(?P%s)(?P\s*\?)? \s*(?P[a-zA-Z0-9_-]+) - \s*\]$ + \s*$ ''' % '|'.join(map(re.escape, STR_OPERATORS.keys()))) - m = str_operator_rex.search(format_spec) + m = str_operator_rex.search(filter_spec) if m: comparison_value = m.group('value') op = STR_OPERATORS[m.group('op')] if not m: - raise ValueError('Invalid format specification %r' % format_spec) + raise ValueError('Invalid filter specification %r' % filter_spec) def _filter(f): actual_value = f.get(m.group('key')) if actual_value is None: return m.group('none_inclusive') return op(actual_value, comparison_value) - new_formats = [f for f in available_formats if _filter(f)] + return _filter + + def build_format_selector(self, format_spec): + def syntax_error(note, start): + message = ( + 'Invalid format specification: ' + '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1])) + return SyntaxError(message) + + PICKFIRST = 'PICKFIRST' + MERGE = 'MERGE' + SINGLE = 'SINGLE' + GROUP = 'GROUP' + FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters']) + + def _parse_filter(tokens): + filter_parts = [] + for type, string, start, _, _ in tokens: + if type == tokenize.OP and string == ']': + return ''.join(filter_parts) + else: + filter_parts.append(string) + + def _remove_unused_ops(tokens): + # Remove operators that we don't use and join them with the sourrounding strings + # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9' + ALLOWED_OPS = ('/', '+', ',', '(', ')') + last_string, last_start, last_end, last_line = None, None, None, None + for type, string, start, end, line in tokens: + if type == tokenize.OP and string == '[': + if last_string: + yield tokenize.NAME, last_string, last_start, last_end, last_line + last_string = None + yield type, string, start, end, line + # everything inside brackets will be handled by _parse_filter + for type, string, start, end, line in tokens: + yield type, string, start, end, line + if type == tokenize.OP and string == ']': + break + elif type == tokenize.OP and string in ALLOWED_OPS: + if last_string: + yield tokenize.NAME, last_string, last_start, last_end, last_line + last_string = None + yield type, string, start, end, line + elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]: + if not last_string: + last_string = string + last_start = start + last_end = end + else: + last_string += string + if last_string: + yield tokenize.NAME, last_string, last_start, last_end, last_line + + def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False): + selectors = [] + current_selector = None + for type, string, start, _, _ in tokens: + # ENCODING is only defined in python 3.x + if type == getattr(tokenize, 'ENCODING', None): + continue + elif type in [tokenize.NAME, tokenize.NUMBER]: + current_selector = FormatSelector(SINGLE, string, []) + elif type == tokenize.OP: + if string == ')': + if not inside_group: + # ')' will be handled by the parentheses group + tokens.restore_last_token() + break + elif inside_merge and string in ['/', ',']: + tokens.restore_last_token() + break + elif inside_choice and string == ',': + tokens.restore_last_token() + break + elif string == ',': + if not current_selector: + raise syntax_error('"," must follow a format selector', start) + selectors.append(current_selector) + current_selector = None + elif string == '/': + if not current_selector: + raise syntax_error('"/" must follow a format selector', start) + first_choice = current_selector + second_choice = _parse_format_selection(tokens, inside_choice=True) + current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), []) + elif string == '[': + if not current_selector: + current_selector = FormatSelector(SINGLE, 'best', []) + format_filter = _parse_filter(tokens) + current_selector.filters.append(format_filter) + elif string == '(': + if current_selector: + raise syntax_error('Unexpected "("', start) + group = _parse_format_selection(tokens, inside_group=True) + current_selector = FormatSelector(GROUP, group, []) + elif string == '+': + video_selector = current_selector + audio_selector = _parse_format_selection(tokens, inside_merge=True) + if not video_selector or not audio_selector: + raise syntax_error('"+" must be between two format selectors', start) + current_selector = FormatSelector(MERGE, (video_selector, audio_selector), []) + else: + raise syntax_error('Operator not recognized: "{0}"'.format(string), start) + elif type == tokenize.ENDMARKER: + break + if current_selector: + selectors.append(current_selector) + return selectors + + def _build_selector_function(selector): + if isinstance(selector, list): + fs = [_build_selector_function(s) for s in selector] + + def selector_function(formats): + for f in fs: + for format in f(formats): + yield format + return selector_function + elif selector.type == GROUP: + selector_function = _build_selector_function(selector.selector) + elif selector.type == PICKFIRST: + fs = [_build_selector_function(s) for s in selector.selector] + + def selector_function(formats): + for f in fs: + picked_formats = list(f(formats)) + if picked_formats: + return picked_formats + return [] + elif selector.type == SINGLE: + format_spec = selector.selector + + def selector_function(formats): + formats = list(formats) + if not formats: + return + if format_spec == 'all': + for f in formats: + yield f + elif format_spec in ['best', 'worst', None]: + format_idx = 0 if format_spec == 'worst' else -1 + audiovideo_formats = [ + f for f in formats + if f.get('vcodec') != 'none' and f.get('acodec') != 'none'] + if audiovideo_formats: + yield audiovideo_formats[format_idx] + # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format + elif (all(f.get('acodec') != 'none' for f in formats) or + all(f.get('vcodec') != 'none' for f in formats)): + yield formats[format_idx] + elif format_spec == 'bestaudio': + audio_formats = [ + f for f in formats + if f.get('vcodec') == 'none'] + if audio_formats: + yield audio_formats[-1] + elif format_spec == 'worstaudio': + audio_formats = [ + f for f in formats + if f.get('vcodec') == 'none'] + if audio_formats: + yield audio_formats[0] + elif format_spec == 'bestvideo': + video_formats = [ + f for f in formats + if f.get('acodec') == 'none'] + if video_formats: + yield video_formats[-1] + elif format_spec == 'worstvideo': + video_formats = [ + f for f in formats + if f.get('acodec') == 'none'] + if video_formats: + yield video_formats[0] + else: + extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] + if format_spec in extensions: + filter_f = lambda f: f['ext'] == format_spec + else: + filter_f = lambda f: f['format_id'] == format_spec + matches = list(filter(filter_f, formats)) + if matches: + yield matches[-1] + elif selector.type == MERGE: + def _merge(formats_info): + format_1, format_2 = [f['format_id'] for f in formats_info] + # The first format must contain the video and the + # second the audio + if formats_info[0].get('vcodec') == 'none': + self.report_error('The first format must ' + 'contain the video, try using ' + '"-f %s+%s"' % (format_2, format_1)) + return + output_ext = ( + formats_info[0]['ext'] + if self.params.get('merge_output_format') is None + else self.params['merge_output_format']) + return { + 'requested_formats': formats_info, + 'format': '%s+%s' % (formats_info[0].get('format'), + formats_info[1].get('format')), + 'format_id': '%s+%s' % (formats_info[0].get('format_id'), + formats_info[1].get('format_id')), + 'width': formats_info[0].get('width'), + 'height': formats_info[0].get('height'), + 'resolution': formats_info[0].get('resolution'), + 'fps': formats_info[0].get('fps'), + 'vcodec': formats_info[0].get('vcodec'), + 'vbr': formats_info[0].get('vbr'), + 'stretched_ratio': formats_info[0].get('stretched_ratio'), + 'acodec': formats_info[1].get('acodec'), + 'abr': formats_info[1].get('abr'), + 'ext': output_ext, + } + video_selector, audio_selector = map(_build_selector_function, selector.selector) - new_format_spec = format_spec[:-len(m.group(0))] - if not new_format_spec: - new_format_spec = 'best' + def selector_function(formats): + formats = list(formats) + for pair in itertools.product(video_selector(formats), audio_selector(formats)): + yield _merge(pair) - return (new_format_spec, new_formats) + filters = [self._build_format_filter(f) for f in selector.filters] - def select_format(self, format_spec, available_formats): - while format_spec.endswith(']'): - format_spec, available_formats = self._apply_format_filter( - format_spec, available_formats) - if not available_formats: - return None + def final_selector(formats): + for _filter in filters: + formats = list(filter(_filter, formats)) + return selector_function(formats) + return final_selector - if format_spec in ['best', 'worst', None]: - format_idx = 0 if format_spec == 'worst' else -1 - audiovideo_formats = [ - f for f in available_formats - if f.get('vcodec') != 'none' and f.get('acodec') != 'none'] - if audiovideo_formats: - return audiovideo_formats[format_idx] - # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format - elif (all(f.get('acodec') != 'none' for f in available_formats) or - all(f.get('vcodec') != 'none' for f in available_formats)): - return available_formats[format_idx] - elif format_spec == 'bestaudio': - audio_formats = [ - f for f in available_formats - if f.get('vcodec') == 'none'] - if audio_formats: - return audio_formats[-1] - elif format_spec == 'worstaudio': - audio_formats = [ - f for f in available_formats - if f.get('vcodec') == 'none'] - if audio_formats: - return audio_formats[0] - elif format_spec == 'bestvideo': - video_formats = [ - f for f in available_formats - if f.get('acodec') == 'none'] - if video_formats: - return video_formats[-1] - elif format_spec == 'worstvideo': - video_formats = [ - f for f in available_formats - if f.get('acodec') == 'none'] - if video_formats: - return video_formats[0] - else: - extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] - if format_spec in extensions: - filter_f = lambda f: f['ext'] == format_spec - else: - filter_f = lambda f: f['format_id'] == format_spec - matches = list(filter(filter_f, available_formats)) - if matches: - return matches[-1] - return None + stream = io.BytesIO(format_spec.encode('utf-8')) + try: + tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline))) + except tokenize.TokenError: + raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec))) + + class TokenIterator(object): + def __init__(self, tokens): + self.tokens = tokens + self.counter = 0 + + def __iter__(self): + return self + + def __next__(self): + if self.counter >= len(self.tokens): + raise StopIteration() + value = self.tokens[self.counter] + self.counter += 1 + return value + + next = __next__ + + def restore_last_token(self): + self.counter -= 1 + + parsed_selector = _parse_format_selection(iter(TokenIterator(tokens))) + return _build_selector_function(parsed_selector) def _calc_headers(self, info_dict): res = std_headers.copy() @@ -1104,62 +1301,15 @@ class YoutubeDL(object): if req_format is None: req_format_list = [] if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and - info_dict['extractor'] in ['youtube', 'ted']): + info_dict['extractor'] in ['youtube', 'ted'] and + not info_dict.get('is_live')): merger = FFmpegMergerPP(self) if merger.available and merger.can_merge(): req_format_list.append('bestvideo+bestaudio') req_format_list.append('best') req_format = '/'.join(req_format_list) - formats_to_download = [] - if req_format == 'all': - formats_to_download = formats - else: - for rfstr in req_format.split(','): - # We can accept formats requested in the format: 34/5/best, we pick - # the first that is available, starting from left - req_formats = rfstr.split('/') - for rf in req_formats: - if re.match(r'.+?\+.+?', rf) is not None: - # Two formats have been requested like '137+139' - format_1, format_2 = rf.split('+') - formats_info = (self.select_format(format_1, formats), - self.select_format(format_2, formats)) - if all(formats_info): - # The first format must contain the video and the - # second the audio - if formats_info[0].get('vcodec') == 'none': - self.report_error('The first format must ' - 'contain the video, try using ' - '"-f %s+%s"' % (format_2, format_1)) - return - output_ext = ( - formats_info[0]['ext'] - if self.params.get('merge_output_format') is None - else self.params['merge_output_format']) - selected_format = { - 'requested_formats': formats_info, - 'format': '%s+%s' % (formats_info[0].get('format'), - formats_info[1].get('format')), - 'format_id': '%s+%s' % (formats_info[0].get('format_id'), - formats_info[1].get('format_id')), - 'width': formats_info[0].get('width'), - 'height': formats_info[0].get('height'), - 'resolution': formats_info[0].get('resolution'), - 'fps': formats_info[0].get('fps'), - 'vcodec': formats_info[0].get('vcodec'), - 'vbr': formats_info[0].get('vbr'), - 'stretched_ratio': formats_info[0].get('stretched_ratio'), - 'acodec': formats_info[1].get('acodec'), - 'abr': formats_info[1].get('abr'), - 'ext': output_ext, - } - else: - selected_format = None - else: - selected_format = self.select_format(rf, formats) - if selected_format is not None: - formats_to_download.append(selected_format) - break + format_selector = self.build_format_selector(req_format) + formats_to_download = list(format_selector(formats)) if not formats_to_download: raise ExtractorError('requested format not available', expected=True) @@ -1707,27 +1857,6 @@ class YoutubeDL(object): def urlopen(self, req): """ Start an HTTP download """ - - # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not - # always respected by websites, some tend to give out URLs with non percent-encoded - # non-ASCII characters (see telemb.py, ard.py [#3412]) - # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991) - # To work around aforementioned issue we will replace request's original URL with - # percent-encoded one - req_is_string = isinstance(req, compat_basestring) - url = req if req_is_string else req.get_full_url() - url_escaped = escape_url(url) - - # Substitute URL if any change after escaping - if url != url_escaped: - if req_is_string: - req = url_escaped - else: - req_type = HEADRequest if req.get_method() == 'HEAD' else compat_urllib_request.Request - req = req_type( - url_escaped, data=req.data, headers=req.headers, - origin_req_host=req.origin_req_host, unverifiable=req.unverifiable) - return self._opener.open(req, timeout=self._socket_timeout) def print_debug_header(self): diff --git a/youtube_dl/compat.py b/youtube_dl/compat.py index 0c57c7aeb..ace5bd716 100644 --- a/youtube_dl/compat.py +++ b/youtube_dl/compat.py @@ -42,6 +42,11 @@ try: except ImportError: # Python 2 import cookielib as compat_cookiejar +try: + import http.cookies as compat_cookies +except ImportError: # Python 2 + import Cookie as compat_cookies + try: import html.entities as compat_html_entities except ImportError: # Python 2 @@ -431,11 +436,17 @@ except TypeError: # Python 2.6 yield n n += step +if sys.version_info >= (3, 0): + from tokenize import tokenize as compat_tokenize_tokenize +else: + from tokenize import generate_tokens as compat_tokenize_tokenize + __all__ = [ 'compat_HTTPError', 'compat_basestring', 'compat_chr', 'compat_cookiejar', + 'compat_cookies', 'compat_expanduser', 'compat_get_terminal_size', 'compat_getenv', @@ -451,6 +462,7 @@ __all__ = [ 'compat_socket_create_connection', 'compat_str', 'compat_subprocess_get_DEVNULL', + 'compat_tokenize_tokenize', 'compat_urllib_error', 'compat_urllib_parse', 'compat_urllib_parse_unquote', diff --git a/youtube_dl/downloader/__init__.py b/youtube_dl/downloader/__init__.py index f110830c4..dccc59212 100644 --- a/youtube_dl/downloader/__init__.py +++ b/youtube_dl/downloader/__init__.py @@ -8,6 +8,7 @@ from .hls import NativeHlsFD from .http import HttpFD from .rtsp import RtspFD from .rtmp import RtmpFD +from .dash import DashSegmentsFD from ..utils import ( determine_protocol, @@ -20,6 +21,7 @@ PROTOCOL_MAP = { 'mms': RtspFD, 'rtsp': RtspFD, 'f4m': F4mFD, + 'http_dash_segments': DashSegmentsFD, } diff --git a/youtube_dl/downloader/dash.py b/youtube_dl/downloader/dash.py new file mode 100644 index 000000000..8b6fa2753 --- /dev/null +++ b/youtube_dl/downloader/dash.py @@ -0,0 +1,66 @@ +from __future__ import unicode_literals + +import re + +from .common import FileDownloader +from ..compat import compat_urllib_request + + +class DashSegmentsFD(FileDownloader): + """ + Download segments in a DASH manifest + """ + def real_download(self, filename, info_dict): + self.report_destination(filename) + tmpfilename = self.temp_name(filename) + base_url = info_dict['url'] + segment_urls = info_dict['segment_urls'] + + is_test = self.params.get('test', False) + remaining_bytes = self._TEST_FILE_SIZE if is_test else None + byte_counter = 0 + + def append_url_to_file(outf, target_url, target_name, remaining_bytes=None): + self.to_screen('[DashSegments] %s: Downloading %s' % (info_dict['id'], target_name)) + req = compat_urllib_request.Request(target_url) + if remaining_bytes is not None: + req.add_header('Range', 'bytes=0-%d' % (remaining_bytes - 1)) + + data = self.ydl.urlopen(req).read() + + if remaining_bytes is not None: + data = data[:remaining_bytes] + + outf.write(data) + return len(data) + + def combine_url(base_url, target_url): + if re.match(r'^https?://', target_url): + return target_url + return '%s%s%s' % (base_url, '' if base_url.endswith('/') else '/', target_url) + + with open(tmpfilename, 'wb') as outf: + append_url_to_file( + outf, combine_url(base_url, info_dict['initialization_url']), + 'initialization segment') + for i, segment_url in enumerate(segment_urls): + segment_len = append_url_to_file( + outf, combine_url(base_url, segment_url), + 'segment %d / %d' % (i + 1, len(segment_urls)), + remaining_bytes) + byte_counter += segment_len + if remaining_bytes is not None: + remaining_bytes -= segment_len + if remaining_bytes <= 0: + break + + self.try_rename(tmpfilename, filename) + + self._hook_progress({ + 'downloaded_bytes': byte_counter, + 'total_bytes': byte_counter, + 'filename': filename, + 'status': 'finished', + }) + + return True diff --git a/youtube_dl/downloader/external.py b/youtube_dl/downloader/external.py index 1d5cc9904..6c310346c 100644 --- a/youtube_dl/downloader/external.py +++ b/youtube_dl/downloader/external.py @@ -45,11 +45,13 @@ class ExternalFD(FileDownloader): def supports(cls, info_dict): return info_dict['protocol'] in ('http', 'https', 'ftp', 'ftps') - def _source_address(self, command_option): - source_address = self.params.get('source_address') - if source_address is None: + def _option(self, command_option, param): + param = self.params.get(param) + if param is None: return [] - return [command_option, source_address] + if isinstance(param, bool): + return [command_option] + return [command_option, param] def _configuration_args(self, default=[]): ex_args = self.params.get('external_downloader_args') @@ -77,7 +79,17 @@ class CurlFD(ExternalFD): cmd = [self.exe, '--location', '-o', tmpfilename] for key, val in info_dict['http_headers'].items(): cmd += ['--header', '%s: %s' % (key, val)] - cmd += self._source_address('--interface') + cmd += self._option('--interface', 'source_address') + cmd += self._configuration_args() + cmd += ['--', info_dict['url']] + return cmd + + +class AxelFD(ExternalFD): + def _make_cmd(self, tmpfilename, info_dict): + cmd = [self.exe, '-o', tmpfilename] + for key, val in info_dict['http_headers'].items(): + cmd += ['-H', '%s: %s' % (key, val)] cmd += self._configuration_args() cmd += ['--', info_dict['url']] return cmd @@ -88,7 +100,9 @@ class WgetFD(ExternalFD): cmd = [self.exe, '-O', tmpfilename, '-nv', '--no-cookies'] for key, val in info_dict['http_headers'].items(): cmd += ['--header', '%s: %s' % (key, val)] - cmd += self._source_address('--bind-address') + cmd += self._option('--bind-address', 'source_address') + cmd += self._option('--proxy', 'proxy') + cmd += self._option('--no-check-certificate', 'nocheckcertificate') cmd += self._configuration_args() cmd += ['--', info_dict['url']] return cmd @@ -105,7 +119,8 @@ class Aria2cFD(ExternalFD): cmd += ['--out', os.path.basename(tmpfilename)] for key, val in info_dict['http_headers'].items(): cmd += ['--header', '%s: %s' % (key, val)] - cmd += self._source_address('--interface') + cmd += self._option('--interface', 'source_address') + cmd += self._option('--all-proxy', 'proxy') cmd += ['--', info_dict['url']] return cmd diff --git a/youtube_dl/downloader/f4m.py b/youtube_dl/downloader/f4m.py index b1a858c45..275564b59 100644 --- a/youtube_dl/downloader/f4m.py +++ b/youtube_dl/downloader/f4m.py @@ -7,8 +7,7 @@ import os import time import xml.etree.ElementTree as etree -from .common import FileDownloader -from .http import HttpFD +from .fragment import FragmentFD from ..compat import ( compat_urlparse, compat_urllib_error, @@ -16,8 +15,6 @@ from ..compat import ( from ..utils import ( struct_pack, struct_unpack, - encodeFilename, - sanitize_open, xpath_text, ) @@ -226,16 +223,13 @@ def _add_ns(prop): return '{http://ns.adobe.com/f4m/1.0}%s' % prop -class HttpQuietDownloader(HttpFD): - def to_screen(self, *args, **kargs): - pass - - -class F4mFD(FileDownloader): +class F4mFD(FragmentFD): """ A downloader for f4m manifests or AdobeHDS. """ + FD_NAME = 'f4m' + def _get_unencrypted_media(self, doc): media = doc.findall(_add_ns('media')) if not media: @@ -288,7 +282,7 @@ class F4mFD(FileDownloader): def real_download(self, filename, info_dict): man_url = info_dict['url'] requested_bitrate = info_dict.get('tbr') - self.to_screen('[download] Downloading f4m manifest') + self.to_screen('[%s] Downloading f4m manifest' % self.FD_NAME) manifest = self.ydl.urlopen(man_url).read() doc = etree.fromstring(manifest) @@ -320,67 +314,20 @@ class F4mFD(FileDownloader): # For some akamai manifests we'll need to add a query to the fragment url akamai_pv = xpath_text(doc, _add_ns('pv-2.0')) - self.report_destination(filename) - http_dl = HttpQuietDownloader( - self.ydl, - { - 'continuedl': True, - 'quiet': True, - 'noprogress': True, - 'ratelimit': self.params.get('ratelimit', None), - 'test': self.params.get('test', False), - } - ) - tmpfilename = self.temp_name(filename) - (dest_stream, tmpfilename) = sanitize_open(tmpfilename, 'wb') + ctx = { + 'filename': filename, + 'total_frags': total_frags, + } + + self._prepare_frag_download(ctx) + + dest_stream = ctx['dest_stream'] write_flv_header(dest_stream) if not live: write_metadata_tag(dest_stream, metadata) - # This dict stores the download progress, it's updated by the progress - # hook - state = { - 'status': 'downloading', - 'downloaded_bytes': 0, - 'frag_index': 0, - 'frag_count': total_frags, - 'filename': filename, - 'tmpfilename': tmpfilename, - } - start = time.time() - - def frag_progress_hook(s): - if s['status'] not in ('downloading', 'finished'): - return - - frag_total_bytes = s.get('total_bytes', 0) - if s['status'] == 'finished': - state['downloaded_bytes'] += frag_total_bytes - state['frag_index'] += 1 - - estimated_size = ( - (state['downloaded_bytes'] + frag_total_bytes) / - (state['frag_index'] + 1) * total_frags) - time_now = time.time() - state['total_bytes_estimate'] = estimated_size - state['elapsed'] = time_now - start - - if s['status'] == 'finished': - progress = self.calc_percent(state['frag_index'], total_frags) - else: - frag_downloaded_bytes = s['downloaded_bytes'] - frag_progress = self.calc_percent(frag_downloaded_bytes, - frag_total_bytes) - progress = self.calc_percent(state['frag_index'], total_frags) - progress += frag_progress / float(total_frags) - - state['eta'] = self.calc_eta( - start, time_now, estimated_size, state['downloaded_bytes'] + frag_downloaded_bytes) - state['speed'] = s.get('speed') - self._hook_progress(state) - - http_dl.add_progress_hook(frag_progress_hook) + self._start_frag_download(ctx) frags_filenames = [] while fragments_list: @@ -391,9 +338,9 @@ class F4mFD(FileDownloader): url += '?' + akamai_pv.strip(';') if info_dict.get('extra_param_to_segment_url'): url += info_dict.get('extra_param_to_segment_url') - frag_filename = '%s-%s' % (tmpfilename, name) + frag_filename = '%s-%s' % (ctx['tmpfilename'], name) try: - success = http_dl.download(frag_filename, {'url': url}) + success = ctx['dl'].download(frag_filename, {'url': url}) if not success: return False with open(frag_filename, 'rb') as down: @@ -425,20 +372,9 @@ class F4mFD(FileDownloader): msg = 'Missed %d fragments' % (fragments_list[0][1] - (frag_i + 1)) self.report_warning(msg) - dest_stream.close() + self._finish_frag_download(ctx) - elapsed = time.time() - start - self.try_rename(tmpfilename, filename) for frag_file in frags_filenames: os.remove(frag_file) - fsize = os.path.getsize(encodeFilename(filename)) - self._hook_progress({ - 'downloaded_bytes': fsize, - 'total_bytes': fsize, - 'filename': filename, - 'status': 'finished', - 'elapsed': elapsed, - }) - return True diff --git a/youtube_dl/downloader/fragment.py b/youtube_dl/downloader/fragment.py new file mode 100644 index 000000000..5a64b29ee --- /dev/null +++ b/youtube_dl/downloader/fragment.py @@ -0,0 +1,111 @@ +from __future__ import division, unicode_literals + +import os +import time + +from .common import FileDownloader +from .http import HttpFD +from ..utils import ( + encodeFilename, + sanitize_open, +) + + +class HttpQuietDownloader(HttpFD): + def to_screen(self, *args, **kargs): + pass + + +class FragmentFD(FileDownloader): + """ + A base file downloader class for fragmented media (e.g. f4m/m3u8 manifests). + """ + + def _prepare_and_start_frag_download(self, ctx): + self._prepare_frag_download(ctx) + self._start_frag_download(ctx) + + def _prepare_frag_download(self, ctx): + self.to_screen('[%s] Total fragments: %d' % (self.FD_NAME, ctx['total_frags'])) + self.report_destination(ctx['filename']) + dl = HttpQuietDownloader( + self.ydl, + { + 'continuedl': True, + 'quiet': True, + 'noprogress': True, + 'ratelimit': self.params.get('ratelimit', None), + 'retries': self.params.get('retries', 0), + 'test': self.params.get('test', False), + } + ) + tmpfilename = self.temp_name(ctx['filename']) + dest_stream, tmpfilename = sanitize_open(tmpfilename, 'wb') + ctx.update({ + 'dl': dl, + 'dest_stream': dest_stream, + 'tmpfilename': tmpfilename, + }) + + def _start_frag_download(self, ctx): + total_frags = ctx['total_frags'] + # This dict stores the download progress, it's updated by the progress + # hook + state = { + 'status': 'downloading', + 'downloaded_bytes': 0, + 'frag_index': 0, + 'frag_count': total_frags, + 'filename': ctx['filename'], + 'tmpfilename': ctx['tmpfilename'], + } + start = time.time() + ctx['started'] = start + + def frag_progress_hook(s): + if s['status'] not in ('downloading', 'finished'): + return + + frag_total_bytes = s.get('total_bytes', 0) + if s['status'] == 'finished': + state['downloaded_bytes'] += frag_total_bytes + state['frag_index'] += 1 + + estimated_size = ( + (state['downloaded_bytes'] + frag_total_bytes) / + (state['frag_index'] + 1) * total_frags) + time_now = time.time() + state['total_bytes_estimate'] = estimated_size + state['elapsed'] = time_now - start + + if s['status'] == 'finished': + progress = self.calc_percent(state['frag_index'], total_frags) + else: + frag_downloaded_bytes = s['downloaded_bytes'] + frag_progress = self.calc_percent(frag_downloaded_bytes, + frag_total_bytes) + progress = self.calc_percent(state['frag_index'], total_frags) + progress += frag_progress / float(total_frags) + + state['eta'] = self.calc_eta( + start, time_now, estimated_size, state['downloaded_bytes'] + frag_downloaded_bytes) + state['speed'] = s.get('speed') + self._hook_progress(state) + + ctx['dl'].add_progress_hook(frag_progress_hook) + + return start + + def _finish_frag_download(self, ctx): + ctx['dest_stream'].close() + elapsed = time.time() - ctx['started'] + self.try_rename(ctx['tmpfilename'], ctx['filename']) + fsize = os.path.getsize(encodeFilename(ctx['filename'])) + + self._hook_progress({ + 'downloaded_bytes': fsize, + 'total_bytes': fsize, + 'filename': ctx['filename'], + 'status': 'finished', + 'elapsed': elapsed, + }) diff --git a/youtube_dl/downloader/hls.py b/youtube_dl/downloader/hls.py index 8be4f4249..2b6c3370f 100644 --- a/youtube_dl/downloader/hls.py +++ b/youtube_dl/downloader/hls.py @@ -4,12 +4,11 @@ import os import re import subprocess -from ..postprocessor.ffmpeg import FFmpegPostProcessor from .common import FileDownloader -from ..compat import ( - compat_urlparse, - compat_urllib_request, -) +from .fragment import FragmentFD + +from ..compat import compat_urlparse +from ..postprocessor.ffmpeg import FFmpegPostProcessor from ..utils import ( encodeArgument, encodeFilename, @@ -33,6 +32,8 @@ class HlsFD(FileDownloader): for opt in (ffpp.executable, '-y', '-i', url, '-f', 'mp4', '-c', 'copy', '-bsf:a', 'aac_adtstoasc')] args.append(encodeFilename(tmpfilename, True)) + self._debug_cmd(args) + retval = subprocess.call(args) if retval == 0: fsize = os.path.getsize(encodeFilename(tmpfilename)) @@ -51,54 +52,50 @@ class HlsFD(FileDownloader): return False -class NativeHlsFD(FileDownloader): +class NativeHlsFD(FragmentFD): """ A more limited implementation that does not require ffmpeg """ + FD_NAME = 'hlsnative' + def real_download(self, filename, info_dict): - url = info_dict['url'] - self.report_destination(filename) - tmpfilename = self.temp_name(filename) + man_url = info_dict['url'] + self.to_screen('[%s] Downloading m3u8 manifest' % self.FD_NAME) + manifest = self.ydl.urlopen(man_url).read() - self.to_screen( - '[hlsnative] %s: Downloading m3u8 manifest' % info_dict['id']) - data = self.ydl.urlopen(url).read() - s = data.decode('utf-8', 'ignore') - segment_urls = [] + s = manifest.decode('utf-8', 'ignore') + fragment_urls = [] for line in s.splitlines(): line = line.strip() if line and not line.startswith('#'): segment_url = ( line if re.match(r'^https?://', line) - else compat_urlparse.urljoin(url, line)) - segment_urls.append(segment_url) - - is_test = self.params.get('test', False) - remaining_bytes = self._TEST_FILE_SIZE if is_test else None - byte_counter = 0 - with open(tmpfilename, 'wb') as outf: - for i, segurl in enumerate(segment_urls): - self.to_screen( - '[hlsnative] %s: Downloading segment %d / %d' % - (info_dict['id'], i + 1, len(segment_urls))) - seg_req = compat_urllib_request.Request(segurl) - if remaining_bytes is not None: - seg_req.add_header('Range', 'bytes=0-%d' % (remaining_bytes - 1)) - - segment = self.ydl.urlopen(seg_req).read() - if remaining_bytes is not None: - segment = segment[:remaining_bytes] - remaining_bytes -= len(segment) - outf.write(segment) - byte_counter += len(segment) - if remaining_bytes is not None and remaining_bytes <= 0: + else compat_urlparse.urljoin(man_url, line)) + fragment_urls.append(segment_url) + # We only download the first fragment during the test + if self.params.get('test', False): break - self._hook_progress({ - 'downloaded_bytes': byte_counter, - 'total_bytes': byte_counter, + ctx = { 'filename': filename, - 'status': 'finished', - }) - self.try_rename(tmpfilename, filename) + 'total_frags': len(fragment_urls), + } + + self._prepare_and_start_frag_download(ctx) + + frags_filenames = [] + for i, frag_url in enumerate(fragment_urls): + frag_filename = '%s-Frag%d' % (ctx['tmpfilename'], i) + success = ctx['dl'].download(frag_filename, {'url': frag_url}) + if not success: + return False + with open(frag_filename, 'rb') as down: + ctx['dest_stream'].write(down.read()) + frags_filenames.append(frag_filename) + + self._finish_frag_download(ctx) + + for frag_file in frags_filenames: + os.remove(frag_file) + return True diff --git a/youtube_dl/downloader/http.py b/youtube_dl/downloader/http.py index b7f144af9..a29f5cf31 100644 --- a/youtube_dl/downloader/http.py +++ b/youtube_dl/downloader/http.py @@ -4,6 +4,7 @@ import errno import os import socket import time +import re from .common import FileDownloader from ..compat import ( @@ -57,6 +58,24 @@ 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: diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py index 6c1d4f08a..7459a1944 100644 --- a/youtube_dl/extractor/__init__.py +++ b/youtube_dl/extractor/__init__.py @@ -19,9 +19,14 @@ from .anysex import AnySexIE from .aol import AolIE from .allocine import AllocineIE from .aparat import AparatIE +from .appleconnect import AppleConnectIE from .appletrailers import AppleTrailersIE from .archiveorg import ArchiveOrgIE -from .ard import ARDIE, ARDMediathekIE +from .ard import ( + ARDIE, + ARDMediathekIE, + SportschauIE, +) from .arte import ( ArteTvIE, ArteTVPlus7IE, @@ -38,7 +43,10 @@ from .azubu import AzubuIE from .baidu import BaiduVideoIE from .bambuser import BambuserIE, BambuserChannelIE from .bandcamp import BandcampIE, BandcampAlbumIE -from .bbccouk import BBCCoUkIE +from .bbc import ( + BBCCoUkIE, + BBCIE, +) from .beeg import BeegIE from .behindkink import BehindKinkIE from .beatportpro import BeatportProIE @@ -110,6 +118,7 @@ from .dailymotion import ( ) from .daum import DaumIE from .dbtv import DBTVIE +from .dcn import DCNIE from .dctp import DctpTvIE from .deezer import DeezerPlaylistIE from .dfb import DFBIE @@ -149,6 +158,7 @@ from .eporner import EpornerIE from .eroprofile import EroProfileIE from .escapist import EscapistIE from .espn import ESPNIE +from .esri import EsriVideoIE from .everyonesmixtape import EveryonesMixtapeIE from .exfm import ExfmIE from .expotv import ExpoTVIE @@ -233,11 +243,16 @@ from .imdb import ( ) from .imgur import ImgurIE from .ina import InaIE +from .indavideo import ( + IndavideoIE, + IndavideoEmbedIE, +) from .infoq import InfoQIE from .instagram import InstagramIE, InstagramUserIE from .internetvideoarchive import InternetVideoArchiveIE from .iprima import IPrimaIE from .iqiyi import IqiyiIE +from .ir90tv import Ir90TvIE from .ivi import ( IviIE, IviCompilationIE @@ -270,6 +285,7 @@ from .kuwo import ( ) from .la7 import LA7IE from .laola1tv import Laola1TvIE +from .lecture2go import Lecture2GoIE from .letv import ( LetvIE, LetvTvIE, @@ -421,6 +437,10 @@ from .orf import ( from .parliamentliveuk import ParliamentLiveUKIE from .patreon import PatreonIE from .pbs import PBSIE +from .periscope import ( + PeriscopeIE, + QuickscopeIE, +) from .philharmoniedeparis import PhilharmonieDeParisIE from .phoenix import PhoenixIE from .photobucket import PhotobucketIE @@ -476,6 +496,7 @@ from .rtl2 import RTL2IE from .rtp import RTPIE from .rts import RTSIE from .rtve import RTVEALaCartaIE, RTVELiveIE, RTVEInfantilIE +from .rtvnh import RTVNHIE from .ruhd import RUHDIE from .rutube import ( RutubeIE, @@ -582,6 +603,7 @@ from .techtalks import TechTalksIE from .ted import TEDIE from .telebruxelles import TeleBruxellesIE from .telecinco import TelecincoIE +from .telegraaf import TelegraafIE from .telemb import TeleMBIE from .teletask import TeleTaskIE from .tenplay import TenPlayIE diff --git a/youtube_dl/extractor/appleconnect.py b/youtube_dl/extractor/appleconnect.py new file mode 100644 index 000000000..ea7a70393 --- /dev/null +++ b/youtube_dl/extractor/appleconnect.py @@ -0,0 +1,50 @@ +# coding: utf-8 +from __future__ import unicode_literals + +from .common import InfoExtractor +from ..utils import ( + str_to_int, + ExtractorError +) + + +class AppleConnectIE(InfoExtractor): + _VALID_URL = r'https?://itunes\.apple\.com/\w{0,2}/?post/idsa\.(?P[\w-]+)' + _TEST = { + 'url': 'https://itunes.apple.com/us/post/idsa.4ab17a39-2720-11e5-96c5-a5b38f6c42d3', + 'md5': '10d0f2799111df4cb1c924520ca78f98', + 'info_dict': { + 'id': '4ab17a39-2720-11e5-96c5-a5b38f6c42d3', + 'ext': 'm4v', + 'title': 'Energy', + 'uploader': 'Drake', + 'thumbnail': 'http://is5.mzstatic.com/image/thumb/Video5/v4/78/61/c5/7861c5fa-ad6d-294b-1464-cf7605b911d6/source/1920x1080sr.jpg', + 'upload_date': '20150710', + 'timestamp': 1436545535, + }, + } + + def _real_extract(self, url): + video_id = self._match_id(url) + webpage = self._download_webpage(url, video_id) + + try: + video_json = self._html_search_regex( + r'class="auc-video-data">(\{.*?\})', webpage, 'json') + except ExtractorError: + raise ExtractorError('This post doesn\'t contain a video', expected=True) + + video_data = self._parse_json(video_json, video_id) + timestamp = str_to_int(self._html_search_regex(r'data-timestamp="(\d+)"', webpage, 'timestamp')) + like_count = str_to_int(self._html_search_regex(r'(\d+) Loves', webpage, 'like count')) + + return { + 'id': video_id, + 'url': video_data['sslSrc'], + 'title': video_data['title'], + 'description': video_data['description'], + 'uploader': video_data['artistName'], + 'thumbnail': video_data['artworkUrl'], + 'timestamp': timestamp, + 'like_count': like_count, + } diff --git a/youtube_dl/extractor/ard.py b/youtube_dl/extractor/ard.py index 6a35ea463..6f465789b 100644 --- a/youtube_dl/extractor/ard.py +++ b/youtube_dl/extractor/ard.py @@ -8,6 +8,7 @@ from .generic import GenericIE from ..utils import ( determine_ext, ExtractorError, + get_element_by_attribute, qualities, int_or_none, parse_duration, @@ -22,19 +23,125 @@ class ARDMediathekIE(InfoExtractor): _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.daserste\.de)/(?:.*/)(?P[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?' _TESTS = [{ - 'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht', - 'only_matching': True, + 'url': 'http://www.ardmediathek.de/tv/Dokumentation-und-Reportage/Ich-liebe-das-Leben-trotzdem/rbb-Fernsehen/Video?documentId=29582122&bcastId=3822114', + 'info_dict': { + 'id': '29582122', + 'ext': 'mp4', + 'title': 'Ich liebe das Leben trotzdem', + 'description': 'md5:45e4c225c72b27993314b31a84a5261c', + 'duration': 4557, + }, + 'params': { + # m3u8 download + 'skip_download': True, + }, }, { - 'url': 'http://www.ardmediathek.de/tv/Tatort/Das-Wunder-von-Wolbeck-Video-tgl-ab-20/Das-Erste/Video?documentId=22490580&bcastId=602916', + 'url': 'http://www.ardmediathek.de/tv/Tatort/Tatort-Scheinwelten-H%C3%B6rfassung-Video/Das-Erste/Video?documentId=29522730&bcastId=602916', + 'md5': 'f4d98b10759ac06c0072bbcd1f0b9e3e', 'info_dict': { - 'id': '22490580', + 'id': '29522730', 'ext': 'mp4', - 'title': 'Das Wunder von Wolbeck (Video tgl. ab 20 Uhr)', - 'description': 'Auf einem restaurierten Hof bei Wolbeck wird der Heilpraktiker Raffael Lembeck eines morgens von seiner Frau Stella tot aufgefunden. Das Opfer war offensichtlich in seiner Praxis zu Fall gekommen und ist dann verblutet, erklärt Prof. Boerne am Tatort.', + 'title': 'Tatort: Scheinwelten - Hörfassung (Video tgl. ab 20 Uhr)', + 'description': 'md5:196392e79876d0ac94c94e8cdb2875f1', + 'duration': 5252, }, - 'skip': 'Blocked outside of Germany', + }, { + # audio + 'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086', + 'md5': '219d94d8980b4f538c7fcb0865eb7f2c', + 'info_dict': { + 'id': '28488308', + 'ext': 'mp3', + 'title': 'Tod eines Fußballers', + 'description': 'md5:f6e39f3461f0e1f54bfa48c8875c86ef', + 'duration': 3240, + }, + }, { + 'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht', + 'only_matching': True, }] + def _extract_media_info(self, media_info_url, webpage, video_id): + media_info = self._download_json( + media_info_url, video_id, 'Downloading media JSON') + + formats = self._extract_formats(media_info, video_id) + + if not formats: + if '"fsk"' in webpage: + raise ExtractorError( + 'This video is only available after 20:00', expected=True) + elif media_info.get('_geoblocked'): + raise ExtractorError('This video is not available due to geo restriction', expected=True) + + self._sort_formats(formats) + + duration = int_or_none(media_info.get('_duration')) + thumbnail = media_info.get('_previewImage') + + subtitles = {} + subtitle_url = media_info.get('_subtitleUrl') + if subtitle_url: + subtitles['de'] = [{ + 'ext': 'srt', + 'url': subtitle_url, + }] + + return { + 'id': video_id, + 'duration': duration, + 'thumbnail': thumbnail, + 'formats': formats, + 'subtitles': subtitles, + } + + def _extract_formats(self, media_info, video_id): + type_ = media_info.get('_type') + media_array = media_info.get('_mediaArray', []) + formats = [] + for num, media in enumerate(media_array): + for stream in media.get('_mediaStreamArray', []): + stream_urls = stream.get('_stream') + if not stream_urls: + continue + if not isinstance(stream_urls, list): + stream_urls = [stream_urls] + quality = stream.get('_quality') + server = stream.get('_server') + for stream_url in stream_urls: + ext = determine_ext(stream_url) + if ext == 'f4m': + formats.extend(self._extract_f4m_formats( + stream_url + '?hdcore=3.1.1&plugin=aasp-3.1.1.69.124', + video_id, preference=-1, f4m_id='hds')) + elif ext == 'm3u8': + formats.extend(self._extract_m3u8_formats( + stream_url, video_id, 'mp4', preference=1, m3u8_id='hls')) + else: + if server and server.startswith('rtmp'): + f = { + 'url': server, + 'play_path': stream_url, + 'format_id': 'a%s-rtmp-%s' % (num, quality), + } + elif stream_url.startswith('http'): + f = { + 'url': stream_url, + 'format_id': 'a%s-%s-%s' % (num, ext, quality) + } + else: + continue + m = re.search(r'_(?P\d+)x(?P\d+)\.mp4$', stream_url) + if m: + f.update({ + 'width': int(m.group('width')), + 'height': int(m.group('height')), + }) + if type_ == 'audio': + f['vcodec'] = 'none' + formats.append(f) + return formats + def _real_extract(self, url): # determine video id from url m = re.match(self._VALID_URL, url) @@ -92,46 +199,22 @@ class ARDMediathekIE(InfoExtractor): 'format_id': fid, 'url': furl, }) + self._sort_formats(formats) + info = { + 'formats': formats, + } else: # request JSON file - media_info = self._download_json( - 'http://www.ardmediathek.de/play/media/%s' % video_id, video_id) - # The second element of the _mediaArray contains the standard http urls - streams = media_info['_mediaArray'][1]['_mediaStreamArray'] - if not streams: - if '"fsk"' in webpage: - raise ExtractorError('This video is only available after 20:00') - - formats = [] - for s in streams: - if type(s['_stream']) == list: - for index, url in enumerate(s['_stream'][::-1]): - quality = s['_quality'] + index - formats.append({ - 'quality': quality, - 'url': url, - 'format_id': '%s-%s' % (determine_ext(url), quality) - }) - continue - - format = { - 'quality': s['_quality'], - 'url': s['_stream'], - } - - format['format_id'] = '%s-%s' % ( - determine_ext(format['url']), format['quality']) + info = self._extract_media_info( + 'http://www.ardmediathek.de/play/media/%s' % video_id, webpage, video_id) - formats.append(format) - - self._sort_formats(formats) - - return { + info.update({ 'id': video_id, 'title': title, 'description': description, - 'formats': formats, 'thumbnail': thumbnail, - } + }) + + return info class ARDIE(InfoExtractor): @@ -189,3 +272,41 @@ class ARDIE(InfoExtractor): 'upload_date': upload_date, 'thumbnail': thumbnail, } + + +class SportschauIE(ARDMediathekIE): + IE_NAME = 'Sportschau' + _VALID_URL = r'(?Phttps?://(?:www\.)?sportschau\.de/(?:[^/]+/)+video(?P[^/#?]+))\.html' + _TESTS = [{ + 'url': 'http://www.sportschau.de/tourdefrance/videoseppeltkokainhatnichtsmitklassischemdopingzutun100.html', + 'info_dict': { + 'id': 'seppeltkokainhatnichtsmitklassischemdopingzutun100', + 'ext': 'mp4', + 'title': 'Seppelt: "Kokain hat nichts mit klassischem Doping zu tun"', + 'thumbnail': 're:^https?://.*\.jpg$', + 'description': 'Der ARD-Doping Experte Hajo Seppelt gibt seine Einschätzung zum ersten Dopingfall der diesjährigen Tour de France um den Italiener Luca Paolini ab.', + }, + 'params': { + # m3u8 download + 'skip_download': True, + }, + }] + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + video_id = mobj.group('id') + base_url = mobj.group('baseurl') + + webpage = self._download_webpage(url, video_id) + title = get_element_by_attribute('class', 'headline', webpage) + description = self._html_search_meta('description', webpage, 'description') + + info = self._extract_media_info( + base_url + '-mc_defaultQuality-h.json', webpage, video_id) + + info.update({ + 'title': title, + 'description': description, + }) + + return info diff --git a/youtube_dl/extractor/bbc.py b/youtube_dl/extractor/bbc.py new file mode 100644 index 000000000..abc5a44a1 --- /dev/null +++ b/youtube_dl/extractor/bbc.py @@ -0,0 +1,780 @@ +# coding: utf-8 +from __future__ import unicode_literals + +import re +import xml.etree.ElementTree + +from .common import InfoExtractor +from ..utils import ( + ExtractorError, + float_or_none, + int_or_none, + parse_duration, + parse_iso8601, +) +from ..compat import compat_HTTPError + + +class BBCCoUkIE(InfoExtractor): + IE_NAME = 'bbc.co.uk' + IE_DESC = 'BBC iPlayer' + _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:(?:(?:programmes|iplayer(?:/[^/]+)?/(?:episode|playlist))/)|music/clips[/#])(?P[\da-z]{8})' + + _MEDIASELECTOR_URLS = [ + 'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/pc/vpid/%s', + ] + + _TESTS = [ + { + 'url': 'http://www.bbc.co.uk/programmes/b039g8p7', + 'info_dict': { + 'id': 'b039d07m', + 'ext': 'flv', + 'title': 'Kaleidoscope, Leonard Cohen', + 'description': 'The Canadian poet and songwriter reflects on his musical career.', + 'duration': 1740, + }, + 'params': { + # rtmp download + 'skip_download': True, + } + }, + { + 'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/', + 'info_dict': { + 'id': 'b00yng1d', + 'ext': 'flv', + 'title': 'The Man in Black: Series 3: The Printed Name', + 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.", + 'duration': 1800, + }, + 'params': { + # rtmp download + 'skip_download': True, + }, + 'skip': 'Episode is no longer available on BBC iPlayer Radio', + }, + { + 'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/', + 'info_dict': { + 'id': 'b00yng1d', + 'ext': 'flv', + 'title': 'The Voice UK: Series 3: Blind Auditions 5', + 'description': "Emma Willis and Marvin Humes present the fifth set of blind auditions in the singing competition, as the coaches continue to build their teams based on voice alone.", + 'duration': 5100, + }, + 'params': { + # rtmp download + 'skip_download': True, + }, + 'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only', + }, + { + 'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion', + 'info_dict': { + 'id': 'b03k3pb7', + 'ext': 'flv', + 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction", + 'description': '2. Invasion', + 'duration': 3600, + }, + 'params': { + # rtmp download + 'skip_download': True, + }, + 'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only', + }, { + 'url': 'http://www.bbc.co.uk/programmes/b04v20dw', + 'info_dict': { + 'id': 'b04v209v', + 'ext': 'flv', + 'title': 'Pete Tong, The Essential New Tune Special', + 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!", + 'duration': 10800, + }, + 'params': { + # rtmp download + 'skip_download': True, + } + }, { + 'url': 'http://www.bbc.co.uk/music/clips/p02frcc3', + 'note': 'Audio', + 'info_dict': { + 'id': 'p02frcch', + 'ext': 'flv', + 'title': 'Pete Tong, Past, Present and Future Special, Madeon - After Hours mix', + 'description': 'French house superstar Madeon takes us out of the club and onto the after party.', + 'duration': 3507, + }, + 'params': { + # rtmp download + 'skip_download': True, + } + }, { + 'url': 'http://www.bbc.co.uk/music/clips/p025c0zz', + 'note': 'Video', + 'info_dict': { + 'id': 'p025c103', + 'ext': 'flv', + 'title': 'Reading and Leeds Festival, 2014, Rae Morris - Closer (Live on BBC Three)', + 'description': 'Rae Morris performs Closer for BBC Three at Reading 2014', + 'duration': 226, + }, + 'params': { + # rtmp download + 'skip_download': True, + } + }, { + 'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls', + 'info_dict': { + 'id': 'p02n76xf', + 'ext': 'flv', + 'title': 'Natural World, 2015-2016: 2. Super Powered Owls', + 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d', + 'duration': 3540, + }, + 'params': { + # rtmp download + 'skip_download': True, + }, + 'skip': 'geolocation', + }, { + 'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition', + 'info_dict': { + 'id': 'b05zmgw1', + 'ext': 'flv', + 'description': 'Kirsty Wark and Morgan Quaintance visit the Royal Academy as it prepares for its annual artistic extravaganza, meeting people who have come together to make the show unique.', + 'title': 'Royal Academy Summer Exhibition', + 'duration': 3540, + }, + 'params': { + # rtmp download + 'skip_download': True, + }, + 'skip': 'geolocation', + }, { + 'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4', + 'only_matching': True, + }, { + 'url': 'http://www.bbc.co.uk/music/clips#p02frcc3', + 'only_matching': True, + }, { + 'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo', + 'only_matching': True, + } + ] + + class MediaSelectionError(Exception): + def __init__(self, id): + self.id = id + + def _extract_asx_playlist(self, connection, programme_id): + asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist') + return [ref.get('href') for ref in asx.findall('./Entry/ref')] + + def _extract_connection(self, connection, programme_id): + formats = [] + protocol = connection.get('protocol') + supplier = connection.get('supplier') + if protocol == 'http': + href = connection.get('href') + transfer_format = connection.get('transferFormat') + # ASX playlist + if supplier == 'asx': + for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)): + formats.append({ + 'url': ref, + 'format_id': 'ref%s_%s' % (i, supplier), + }) + # Skip DASH until supported + elif transfer_format == 'dash': + pass + # Direct link + else: + formats.append({ + 'url': href, + 'format_id': supplier, + }) + elif protocol == 'rtmp': + application = connection.get('application', 'ondemand') + auth_string = connection.get('authString') + identifier = connection.get('identifier') + server = connection.get('server') + formats.append({ + 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string), + 'play_path': identifier, + 'app': '%s?%s' % (application, auth_string), + 'page_url': 'http://www.bbc.co.uk', + 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf', + 'rtmp_live': False, + 'ext': 'flv', + 'format_id': supplier, + }) + return formats + + def _extract_items(self, playlist): + return playlist.findall('./{http://bbc.co.uk/2008/emp/playlist}item') + + def _extract_medias(self, media_selection): + error = media_selection.find('./{http://bbc.co.uk/2008/mp/mediaselection}error') + if error is not None: + raise BBCCoUkIE.MediaSelectionError(error.get('id')) + return media_selection.findall('./{http://bbc.co.uk/2008/mp/mediaselection}media') + + def _extract_connections(self, media): + return media.findall('./{http://bbc.co.uk/2008/mp/mediaselection}connection') + + def _extract_video(self, media, programme_id): + formats = [] + vbr = int_or_none(media.get('bitrate')) + vcodec = media.get('encoding') + service = media.get('service') + width = int_or_none(media.get('width')) + height = int_or_none(media.get('height')) + file_size = int_or_none(media.get('media_file_size')) + for connection in self._extract_connections(media): + conn_formats = self._extract_connection(connection, programme_id) + for format in conn_formats: + format.update({ + 'format_id': '%s_%s' % (service, format['format_id']), + 'width': width, + 'height': height, + 'vbr': vbr, + 'vcodec': vcodec, + 'filesize': file_size, + }) + formats.extend(conn_formats) + return formats + + def _extract_audio(self, media, programme_id): + formats = [] + abr = int_or_none(media.get('bitrate')) + acodec = media.get('encoding') + service = media.get('service') + for connection in self._extract_connections(media): + conn_formats = self._extract_connection(connection, programme_id) + for format in conn_formats: + format.update({ + 'format_id': '%s_%s' % (service, format['format_id']), + 'abr': abr, + 'acodec': acodec, + }) + formats.extend(conn_formats) + return formats + + def _get_subtitles(self, media, programme_id): + subtitles = {} + for connection in self._extract_connections(media): + captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions') + lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en') + subtitles[lang] = [ + { + 'url': connection.get('href'), + 'ext': 'ttml', + }, + ] + return subtitles + + def _raise_extractor_error(self, media_selection_error): + raise ExtractorError( + '%s returned error: %s' % (self.IE_NAME, media_selection_error.id), + expected=True) + + def _download_media_selector(self, programme_id): + last_exception = None + for mediaselector_url in self._MEDIASELECTOR_URLS: + try: + return self._download_media_selector_url( + mediaselector_url % programme_id, programme_id) + except BBCCoUkIE.MediaSelectionError as e: + if e.id == 'notukerror': + last_exception = e + continue + self._raise_extractor_error(e) + self._raise_extractor_error(last_exception) + + def _download_media_selector_url(self, url, programme_id=None): + try: + media_selection = self._download_xml( + url, programme_id, 'Downloading media selection XML') + except ExtractorError as ee: + if isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 403: + media_selection = xml.etree.ElementTree.fromstring(ee.cause.read().decode('utf-8')) + else: + raise + return self._process_media_selector(media_selection, programme_id) + + def _process_media_selector(self, media_selection, programme_id): + formats = [] + subtitles = None + + for media in self._extract_medias(media_selection): + kind = media.get('kind') + if kind == 'audio': + formats.extend(self._extract_audio(media, programme_id)) + elif kind == 'video': + formats.extend(self._extract_video(media, programme_id)) + elif kind == 'captions': + subtitles = self.extract_subtitles(media, programme_id) + return formats, subtitles + + def _download_playlist(self, playlist_id): + try: + playlist = self._download_json( + 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id, + playlist_id, 'Downloading playlist JSON') + + version = playlist.get('defaultAvailableVersion') + if version: + smp_config = version['smpConfig'] + title = smp_config['title'] + description = smp_config['summary'] + for item in smp_config['items']: + kind = item['kind'] + if kind != 'programme' and kind != 'radioProgramme': + continue + programme_id = item.get('vpid') + duration = int_or_none(item.get('duration')) + formats, subtitles = self._download_media_selector(programme_id) + return programme_id, title, description, duration, formats, subtitles + except ExtractorError as ee: + if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404): + raise + + # fallback to legacy playlist + return self._process_legacy_playlist(playlist_id) + + def _process_legacy_playlist_url(self, url, display_id): + playlist = self._download_legacy_playlist_url(url, display_id) + return self._extract_from_legacy_playlist(playlist, display_id) + + def _process_legacy_playlist(self, playlist_id): + return self._process_legacy_playlist_url( + 'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id) + + def _download_legacy_playlist_url(self, url, playlist_id=None): + return self._download_xml( + url, playlist_id, 'Downloading legacy playlist XML') + + def _extract_from_legacy_playlist(self, playlist, playlist_id): + no_items = playlist.find('./{http://bbc.co.uk/2008/emp/playlist}noItems') + if no_items is not None: + reason = no_items.get('reason') + if reason == 'preAvailability': + msg = 'Episode %s is not yet available' % playlist_id + elif reason == 'postAvailability': + msg = 'Episode %s is no longer available' % playlist_id + elif reason == 'noMedia': + msg = 'Episode %s is not currently available' % playlist_id + else: + msg = 'Episode %s is not available: %s' % (playlist_id, reason) + raise ExtractorError(msg, expected=True) + + for item in self._extract_items(playlist): + kind = item.get('kind') + if kind != 'programme' and kind != 'radioProgramme': + continue + title = playlist.find('./{http://bbc.co.uk/2008/emp/playlist}title').text + description = playlist.find('./{http://bbc.co.uk/2008/emp/playlist}summary').text + + def get_programme_id(item): + def get_from_attributes(item): + for p in('identifier', 'group'): + value = item.get(p) + if value and re.match(r'^[pb][\da-z]{7}$', value): + return value + get_from_attributes(item) + mediator = item.find('./{http://bbc.co.uk/2008/emp/playlist}mediator') + if mediator is not None: + return get_from_attributes(mediator) + + programme_id = get_programme_id(item) + duration = int_or_none(item.get('duration')) + # TODO: programme_id can be None and media items can be incorporated right inside + # playlist's item (e.g. http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu) + # as f4m and m3u8 + formats, subtitles = self._download_media_selector(programme_id) + + return programme_id, title, description, duration, formats, subtitles + + def _real_extract(self, url): + group_id = self._match_id(url) + + webpage = self._download_webpage(url, group_id, 'Downloading video page') + + programme_id = None + + tviplayer = self._search_regex( + r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById', + webpage, 'player', default=None) + + if tviplayer: + player = self._parse_json(tviplayer, group_id).get('player', {}) + duration = int_or_none(player.get('duration')) + programme_id = player.get('vpid') + + if not programme_id: + programme_id = self._search_regex( + r'"vpid"\s*:\s*"([\da-z]{8})"', webpage, 'vpid', fatal=False, default=None) + + if programme_id: + formats, subtitles = self._download_media_selector(programme_id) + title = self._og_search_title(webpage) + description = self._search_regex( + r'

([^<]+)

', + webpage, 'description', fatal=False) + else: + programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id) + + self._sort_formats(formats) + + return { + 'id': programme_id, + 'title': title, + 'description': description, + 'thumbnail': self._og_search_thumbnail(webpage, default=None), + 'duration': duration, + 'formats': formats, + 'subtitles': subtitles, + } + + +class BBCIE(BBCCoUkIE): + IE_NAME = 'bbc' + IE_DESC = 'BBC' + _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P[^/#?]+)' + + _MEDIASELECTOR_URLS = [ + # Provides more formats, namely direct mp4 links, but fails on some videos with + # notukerror for non UK (?) users (e.g. + # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret) + 'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s', + # Provides fewer formats, but works everywhere for everybody (hopefully) + 'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s', + ] + + _TESTS = [{ + # article with multiple videos embedded with data-media-meta containing + # playlist.sxml, externalId and no direct video links + 'url': 'http://www.bbc.com/news/world-europe-32668511', + 'info_dict': { + 'id': 'world-europe-32668511', + 'title': 'Russia stages massive WW2 parade despite Western boycott', + 'description': 'md5:00ff61976f6081841f759a08bf78cc9c', + }, + 'playlist_count': 2, + }, { + # article with multiple videos embedded with data-media-meta (more videos) + 'url': 'http://www.bbc.com/news/business-28299555', + 'info_dict': { + 'id': 'business-28299555', + 'title': 'Farnborough Airshow: Video highlights', + 'description': 'BBC reports and video highlights at the Farnborough Airshow.', + }, + 'playlist_count': 9, + 'skip': 'Save time', + }, { + # article with multiple videos embedded with `new SMP()` + 'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460', + 'info_dict': { + 'id': '3662a707-0af9-3149-963f-47bea720b460', + 'title': 'BBC Blogs - Adam Curtis - BUGGER', + }, + 'playlist_count': 18, + }, { + # single video embedded with mediaAssetPage.init() + 'url': 'http://www.bbc.com/news/world-europe-32041533', + 'info_dict': { + 'id': 'p02mprgb', + 'ext': 'mp4', + 'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV', + 'duration': 47, + 'timestamp': 1427219242, + 'upload_date': '20150324', + }, + 'params': { + # rtmp download + 'skip_download': True, + } + }, { + # article with single video embedded with data-media-meta containing + # direct video links (for now these are extracted) and playlist.xml (with + # media items as f4m and m3u8 - currently unsupported) + 'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu', + 'info_dict': { + 'id': '150615_telabyad_kentin_cogu', + 'ext': 'mp4', + 'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde", + 'duration': 47, + 'timestamp': 1434397334, + 'upload_date': '20150615', + }, + 'params': { + 'skip_download': True, + } + }, { + # single video embedded with mediaAssetPage.init() (regional section) + 'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw', + 'info_dict': { + 'id': '150619_video_honduras_militares_hospitales_corrupcion_aw', + 'ext': 'mp4', + 'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción', + 'duration': 87, + 'timestamp': 1434713142, + 'upload_date': '20150619', + }, + 'params': { + 'skip_download': True, + } + }, { + # single video from video playlist embedded with vxp-playlist-data JSON + 'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376', + 'info_dict': { + 'id': 'p02w6qjc', + 'ext': 'mp4', + 'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''', + 'duration': 56, + }, + 'params': { + 'skip_download': True, + } + }, { + # single video story with digitalData + 'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret', + 'info_dict': { + 'id': 'p02q6gc4', + 'ext': 'flv', + 'title': 'Sri Lanka’s spicy secret', + 'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.', + 'timestamp': 1437674293, + 'upload_date': '20150723', + }, + 'params': { + # rtmp download + 'skip_download': True, + } + }, { + # single video story without digitalData + 'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star', + 'info_dict': { + 'id': 'p018zqqg', + 'ext': 'mp4', + 'title': 'Hyundai Santa Fe Sport: Rock star', + 'description': 'md5:b042a26142c4154a6e472933cf20793d', + 'timestamp': 1368473503, + 'upload_date': '20130513', + }, + 'params': { + # rtmp download + 'skip_download': True, + } + }, { + # single video with playlist.sxml URL + 'url': 'http://www.bbc.com/sport/0/football/33653409', + 'info_dict': { + 'id': 'p02xycnp', + 'ext': 'mp4', + 'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?', + 'description': 'md5:398fca0e2e701c609d726e034fa1fc89', + 'duration': 140, + }, + 'params': { + # rtmp download + 'skip_download': True, + } + }, { + # single video with playlist URL from weather section + 'url': 'http://www.bbc.com/weather/features/33601775', + 'only_matching': True, + }, { + # custom redirection to www.bbc.com + 'url': 'http://www.bbc.co.uk/news/science-environment-33661876', + 'only_matching': True, + }] + + @classmethod + def suitable(cls, url): + return False if BBCCoUkIE.suitable(url) else super(BBCIE, cls).suitable(url) + + def _extract_from_media_meta(self, media_meta, video_id): + # Direct links to media in media metadata (e.g. + # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu) + # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml + source_files = media_meta.get('sourceFiles') + if source_files: + return [{ + 'url': f['url'], + 'format_id': format_id, + 'ext': f.get('encoding'), + 'tbr': float_or_none(f.get('bitrate'), 1000), + 'filesize': int_or_none(f.get('filesize')), + } for format_id, f in source_files.items() if f.get('url')], [] + + programme_id = media_meta.get('externalId') + if programme_id: + return self._download_media_selector(programme_id) + + # Process playlist.sxml as legacy playlist + href = media_meta.get('href') + if href: + playlist = self._download_legacy_playlist_url(href) + _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id) + return formats, subtitles + + return [], [] + + def _real_extract(self, url): + playlist_id = self._match_id(url) + + webpage = self._download_webpage(url, playlist_id) + + timestamp = parse_iso8601(self._search_regex( + [r'"datePublished":\s*"([^"]+)', + r']+property="article:published_time"[^>]+content="([^"]+)"', + r'itemprop="datePublished"[^>]+datetime="([^"]+)"'], + webpage, 'date', default=None)) + + # single video with playlist.sxml URL (e.g. http://www.bbc.com/sport/0/football/3365340ng) + playlist = self._search_regex( + r']+name="playlist"[^>]+value="([^"]+)"', + webpage, 'playlist', default=None) + if playlist: + programme_id, title, description, duration, formats, subtitles = \ + self._process_legacy_playlist_url(playlist, playlist_id) + self._sort_formats(formats) + return { + 'id': programme_id, + 'title': title, + 'description': description, + 'duration': duration, + 'timestamp': timestamp, + 'formats': formats, + 'subtitles': subtitles, + } + + # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret) + programme_id = self._search_regex( + [r'data-video-player-vpid="([\da-z]{8})"', + r']+name="externalIdentifier"[^>]+value="([\da-z]{8})"'], + webpage, 'vpid', default=None) + if programme_id: + formats, subtitles = self._download_media_selector(programme_id) + self._sort_formats(formats) + # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star) + digital_data = self._parse_json( + self._search_regex( + r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'), + programme_id, fatal=False) + page_info = digital_data.get('page', {}).get('pageInfo', {}) + title = page_info.get('pageName') or self._og_search_title(webpage) + description = page_info.get('description') or self._og_search_description(webpage) + timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp + return { + 'id': programme_id, + 'title': title, + 'description': description, + 'timestamp': timestamp, + 'formats': formats, + 'subtitles': subtitles, + } + + playlist_title = self._html_search_regex( + r'(.*?)(?:\s*-\s*BBC [^ ]+)?', webpage, 'playlist title') + playlist_description = self._og_search_description(webpage, default=None) + + def extract_all(pattern): + return list(filter(None, map( + lambda s: self._parse_json(s, playlist_id, fatal=False), + re.findall(pattern, webpage)))) + + # Multiple video article (e.g. + # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460) + EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+[\da-z]{8}(?:\b[^"]+)?' + entries = [] + for match in extract_all(r'new\s+SMP\(({.+?})\)'): + embed_url = match.get('playerSettings', {}).get('externalEmbedUrl') + if embed_url and re.match(EMBED_URL, embed_url): + entries.append(embed_url) + entries.extend(re.findall( + r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage)) + if entries: + return self.playlist_result( + [self.url_result(entry, 'BBCCoUk') for entry in entries], + playlist_id, playlist_title, playlist_description) + + # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511) + medias = extract_all(r"data-media-meta='({[^']+})'") + + if not medias: + # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international) + media_asset = self._search_regex( + r'mediaAssetPage\.init\(\s*({.+?}), "/', + webpage, 'media asset', default=None) + if media_asset: + media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False) + medias = [] + for video in media_asset_page.get('videos', {}).values(): + medias.extend(video.values()) + + if not medias: + # Multiple video playlist with single `now playing` entry (e.g. + # http://www.bbc.com/news/video_and_audio/must_see/33767813) + vxp_playlist = self._parse_json( + self._search_regex( + r']+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)', + webpage, 'playlist data'), + playlist_id) + playlist_medias = [] + for item in vxp_playlist: + media = item.get('media') + if not media: + continue + playlist_medias.append(media) + # Download single video if found media with asset id matching the video id from URL + if item.get('advert', {}).get('assetId') == playlist_id: + medias = [media] + break + # Fallback to the whole playlist + if not medias: + medias = playlist_medias + + entries = [] + for num, media_meta in enumerate(medias, start=1): + formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id) + if not formats: + continue + self._sort_formats(formats) + + video_id = media_meta.get('externalId') + if not video_id: + video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num) + + title = media_meta.get('caption') + if not title: + title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num) + + duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration')) + + images = [] + for image in media_meta.get('images', {}).values(): + images.extend(image.values()) + if 'image' in media_meta: + images.append(media_meta['image']) + + thumbnails = [{ + 'url': image.get('href'), + 'width': int_or_none(image.get('width')), + 'height': int_or_none(image.get('height')), + } for image in images] + + entries.append({ + 'id': video_id, + 'title': title, + 'thumbnails': thumbnails, + 'duration': duration, + 'timestamp': timestamp, + 'formats': formats, + 'subtitles': subtitles, + }) + + return self.playlist_result(entries, playlist_id, playlist_title, playlist_description) diff --git a/youtube_dl/extractor/bbccouk.py b/youtube_dl/extractor/bbccouk.py deleted file mode 100644 index 5825d2867..000000000 --- a/youtube_dl/extractor/bbccouk.py +++ /dev/null @@ -1,379 +0,0 @@ -from __future__ import unicode_literals - -import xml.etree.ElementTree - -from .common import InfoExtractor -from ..utils import ( - ExtractorError, - int_or_none, -) -from ..compat import compat_HTTPError - - -class BBCCoUkIE(InfoExtractor): - IE_NAME = 'bbc.co.uk' - IE_DESC = 'BBC iPlayer' - _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:(?:(?:programmes|iplayer(?:/[^/]+)?/(?:episode|playlist))/)|music/clips[/#])(?P[\da-z]{8})' - - _TESTS = [ - { - 'url': 'http://www.bbc.co.uk/programmes/b039g8p7', - 'info_dict': { - 'id': 'b039d07m', - 'ext': 'flv', - 'title': 'Kaleidoscope, Leonard Cohen', - 'description': 'The Canadian poet and songwriter reflects on his musical career.', - 'duration': 1740, - }, - 'params': { - # rtmp download - 'skip_download': True, - } - }, - { - 'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/', - 'info_dict': { - 'id': 'b00yng1d', - 'ext': 'flv', - 'title': 'The Man in Black: Series 3: The Printed Name', - 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.", - 'duration': 1800, - }, - 'params': { - # rtmp download - 'skip_download': True, - }, - 'skip': 'Episode is no longer available on BBC iPlayer Radio', - }, - { - 'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/', - 'info_dict': { - 'id': 'b00yng1d', - 'ext': 'flv', - 'title': 'The Voice UK: Series 3: Blind Auditions 5', - 'description': "Emma Willis and Marvin Humes present the fifth set of blind auditions in the singing competition, as the coaches continue to build their teams based on voice alone.", - 'duration': 5100, - }, - 'params': { - # rtmp download - 'skip_download': True, - }, - 'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only', - }, - { - 'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion', - 'info_dict': { - 'id': 'b03k3pb7', - 'ext': 'flv', - 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction", - 'description': '2. Invasion', - 'duration': 3600, - }, - 'params': { - # rtmp download - 'skip_download': True, - }, - 'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only', - }, { - 'url': 'http://www.bbc.co.uk/programmes/b04v20dw', - 'info_dict': { - 'id': 'b04v209v', - 'ext': 'flv', - 'title': 'Pete Tong, The Essential New Tune Special', - 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!", - 'duration': 10800, - }, - 'params': { - # rtmp download - 'skip_download': True, - } - }, { - 'url': 'http://www.bbc.co.uk/music/clips/p02frcc3', - 'note': 'Audio', - 'info_dict': { - 'id': 'p02frcch', - 'ext': 'flv', - 'title': 'Pete Tong, Past, Present and Future Special, Madeon - After Hours mix', - 'description': 'French house superstar Madeon takes us out of the club and onto the after party.', - 'duration': 3507, - }, - 'params': { - # rtmp download - 'skip_download': True, - } - }, { - 'url': 'http://www.bbc.co.uk/music/clips/p025c0zz', - 'note': 'Video', - 'info_dict': { - 'id': 'p025c103', - 'ext': 'flv', - 'title': 'Reading and Leeds Festival, 2014, Rae Morris - Closer (Live on BBC Three)', - 'description': 'Rae Morris performs Closer for BBC Three at Reading 2014', - 'duration': 226, - }, - 'params': { - # rtmp download - 'skip_download': True, - } - }, { - 'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls', - 'info_dict': { - 'id': 'p02n76xf', - 'ext': 'flv', - 'title': 'Natural World, 2015-2016: 2. Super Powered Owls', - 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d', - 'duration': 3540, - }, - 'params': { - # rtmp download - 'skip_download': True, - }, - 'skip': 'geolocation', - }, { - 'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition', - 'info_dict': { - 'id': 'b05zmgw1', - 'ext': 'flv', - 'description': 'Kirsty Wark and Morgan Quaintance visit the Royal Academy as it prepares for its annual artistic extravaganza, meeting people who have come together to make the show unique.', - 'title': 'Royal Academy Summer Exhibition', - 'duration': 3540, - }, - 'params': { - # rtmp download - 'skip_download': True, - }, - 'skip': 'geolocation', - }, { - 'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4', - 'only_matching': True, - }, { - 'url': 'http://www.bbc.co.uk/music/clips#p02frcc3', - 'only_matching': True, - }, { - 'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo', - 'only_matching': True, - } - ] - - def _extract_asx_playlist(self, connection, programme_id): - asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist') - return [ref.get('href') for ref in asx.findall('./Entry/ref')] - - def _extract_connection(self, connection, programme_id): - formats = [] - protocol = connection.get('protocol') - supplier = connection.get('supplier') - if protocol == 'http': - href = connection.get('href') - # ASX playlist - if supplier == 'asx': - for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)): - formats.append({ - 'url': ref, - 'format_id': 'ref%s_%s' % (i, supplier), - }) - # Direct link - else: - formats.append({ - 'url': href, - 'format_id': supplier, - }) - elif protocol == 'rtmp': - application = connection.get('application', 'ondemand') - auth_string = connection.get('authString') - identifier = connection.get('identifier') - server = connection.get('server') - formats.append({ - 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string), - 'play_path': identifier, - 'app': '%s?%s' % (application, auth_string), - 'page_url': 'http://www.bbc.co.uk', - 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf', - 'rtmp_live': False, - 'ext': 'flv', - 'format_id': supplier, - }) - return formats - - def _extract_items(self, playlist): - return playlist.findall('./{http://bbc.co.uk/2008/emp/playlist}item') - - def _extract_medias(self, media_selection): - error = media_selection.find('./{http://bbc.co.uk/2008/mp/mediaselection}error') - if error is not None: - raise ExtractorError( - '%s returned error: %s' % (self.IE_NAME, error.get('id')), expected=True) - return media_selection.findall('./{http://bbc.co.uk/2008/mp/mediaselection}media') - - def _extract_connections(self, media): - return media.findall('./{http://bbc.co.uk/2008/mp/mediaselection}connection') - - def _extract_video(self, media, programme_id): - formats = [] - vbr = int(media.get('bitrate')) - vcodec = media.get('encoding') - service = media.get('service') - width = int(media.get('width')) - height = int(media.get('height')) - file_size = int(media.get('media_file_size')) - for connection in self._extract_connections(media): - conn_formats = self._extract_connection(connection, programme_id) - for format in conn_formats: - format.update({ - 'format_id': '%s_%s' % (service, format['format_id']), - 'width': width, - 'height': height, - 'vbr': vbr, - 'vcodec': vcodec, - 'filesize': file_size, - }) - formats.extend(conn_formats) - return formats - - def _extract_audio(self, media, programme_id): - formats = [] - abr = int(media.get('bitrate')) - acodec = media.get('encoding') - service = media.get('service') - for connection in self._extract_connections(media): - conn_formats = self._extract_connection(connection, programme_id) - for format in conn_formats: - format.update({ - 'format_id': '%s_%s' % (service, format['format_id']), - 'abr': abr, - 'acodec': acodec, - }) - formats.extend(conn_formats) - return formats - - def _get_subtitles(self, media, programme_id): - subtitles = {} - for connection in self._extract_connections(media): - captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions') - lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en') - subtitles[lang] = [ - { - 'url': connection.get('href'), - 'ext': 'ttml', - }, - ] - return subtitles - - def _download_media_selector(self, programme_id): - try: - media_selection = self._download_xml( - 'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/pc/vpid/%s' % programme_id, - programme_id, 'Downloading media selection XML') - except ExtractorError as ee: - if isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 403: - media_selection = xml.etree.ElementTree.fromstring(ee.cause.read().decode('utf-8')) - else: - raise - - formats = [] - subtitles = None - - for media in self._extract_medias(media_selection): - kind = media.get('kind') - if kind == 'audio': - formats.extend(self._extract_audio(media, programme_id)) - elif kind == 'video': - formats.extend(self._extract_video(media, programme_id)) - elif kind == 'captions': - subtitles = self.extract_subtitles(media, programme_id) - - return formats, subtitles - - def _download_playlist(self, playlist_id): - try: - playlist = self._download_json( - 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id, - playlist_id, 'Downloading playlist JSON') - - version = playlist.get('defaultAvailableVersion') - if version: - smp_config = version['smpConfig'] - title = smp_config['title'] - description = smp_config['summary'] - for item in smp_config['items']: - kind = item['kind'] - if kind != 'programme' and kind != 'radioProgramme': - continue - programme_id = item.get('vpid') - duration = int(item.get('duration')) - formats, subtitles = self._download_media_selector(programme_id) - return programme_id, title, description, duration, formats, subtitles - except ExtractorError as ee: - if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404): - raise - - # fallback to legacy playlist - playlist = self._download_xml( - 'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, - playlist_id, 'Downloading legacy playlist XML') - - no_items = playlist.find('./{http://bbc.co.uk/2008/emp/playlist}noItems') - if no_items is not None: - reason = no_items.get('reason') - if reason == 'preAvailability': - msg = 'Episode %s is not yet available' % playlist_id - elif reason == 'postAvailability': - msg = 'Episode %s is no longer available' % playlist_id - elif reason == 'noMedia': - msg = 'Episode %s is not currently available' % playlist_id - else: - msg = 'Episode %s is not available: %s' % (playlist_id, reason) - raise ExtractorError(msg, expected=True) - - for item in self._extract_items(playlist): - kind = item.get('kind') - if kind != 'programme' and kind != 'radioProgramme': - continue - title = playlist.find('./{http://bbc.co.uk/2008/emp/playlist}title').text - description = playlist.find('./{http://bbc.co.uk/2008/emp/playlist}summary').text - programme_id = item.get('identifier') - duration = int(item.get('duration')) - formats, subtitles = self._download_media_selector(programme_id) - - return programme_id, title, description, duration, formats, subtitles - - def _real_extract(self, url): - group_id = self._match_id(url) - - webpage = self._download_webpage(url, group_id, 'Downloading video page') - - programme_id = None - - tviplayer = self._search_regex( - r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById', - webpage, 'player', default=None) - - if tviplayer: - player = self._parse_json(tviplayer, group_id).get('player', {}) - duration = int_or_none(player.get('duration')) - programme_id = player.get('vpid') - - if not programme_id: - programme_id = self._search_regex( - r'"vpid"\s*:\s*"([\da-z]{8})"', webpage, 'vpid', fatal=False, default=None) - - if programme_id: - formats, subtitles = self._download_media_selector(programme_id) - title = self._og_search_title(webpage) - description = self._search_regex( - r'

([^<]+)

', - webpage, 'description', fatal=False) - else: - programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id) - - self._sort_formats(formats) - - return { - 'id': programme_id, - 'title': title, - 'description': description, - 'thumbnail': self._og_search_thumbnail(webpage, default=None), - 'duration': duration, - 'formats': formats, - 'subtitles': subtitles, - } diff --git a/youtube_dl/extractor/breakcom.py b/youtube_dl/extractor/breakcom.py index 809287d14..aa08051b1 100644 --- a/youtube_dl/extractor/breakcom.py +++ b/youtube_dl/extractor/breakcom.py @@ -18,6 +18,7 @@ class BreakIE(InfoExtractor): 'id': '2468056', 'ext': 'mp4', 'title': 'When Girls Act Like D-Bags', + 'age_limit': 13, } }, { 'url': 'http://www.break.com/video/ugc/baby-flex-2773063', diff --git a/youtube_dl/extractor/canalplus.py b/youtube_dl/extractor/canalplus.py index 699b4f7d0..57e0cda2c 100644 --- a/youtube_dl/extractor/canalplus.py +++ b/youtube_dl/extractor/canalplus.py @@ -106,15 +106,11 @@ class CanalplusIE(InfoExtractor): continue format_id = fmt.tag if format_id == 'HLS': - hls_formats = self._extract_m3u8_formats(format_url, video_id, 'flv') - for fmt in hls_formats: - fmt['preference'] = preference(format_id) - formats.extend(hls_formats) + formats.extend(self._extract_m3u8_formats( + format_url, video_id, 'mp4', preference=preference(format_id))) elif format_id == 'HDS': - hds_formats = self._extract_f4m_formats(format_url + '?hdcore=2.11.3', video_id) - for fmt in hds_formats: - fmt['preference'] = preference(format_id) - formats.extend(hds_formats) + formats.extend(self._extract_f4m_formats( + format_url + '?hdcore=2.11.3', video_id, preference=preference(format_id))) else: formats.append({ 'url': format_url, diff --git a/youtube_dl/extractor/clipfish.py b/youtube_dl/extractor/clipfish.py index a5c3cb7c6..7af903571 100644 --- a/youtube_dl/extractor/clipfish.py +++ b/youtube_dl/extractor/clipfish.py @@ -1,53 +1,68 @@ from __future__ import unicode_literals import re -import time -import xml.etree.ElementTree from .common import InfoExtractor from ..utils import ( - ExtractorError, - parse_duration, + determine_ext, + int_or_none, + js_to_json, + parse_iso8601, + remove_end, ) class ClipfishIE(InfoExtractor): - IE_NAME = 'clipfish' - - _VALID_URL = r'^https?://(?:www\.)?clipfish\.de/.*?/video/(?P[0-9]+)/' + _VALID_URL = r'https?://(?:www\.)?clipfish\.de/(?:[^/]+/)+video/(?P[0-9]+)' _TEST = { 'url': 'http://www.clipfish.de/special/game-trailer/video/3966754/fifa-14-e3-2013-trailer/', - 'md5': '2521cd644e862936cf2e698206e47385', + 'md5': '79bc922f3e8a9097b3d68a93780fd475', 'info_dict': { 'id': '3966754', 'ext': 'mp4', 'title': 'FIFA 14 - E3 2013 Trailer', + 'timestamp': 1370938118, + 'upload_date': '20130611', 'duration': 82, - }, - 'skip': 'Blocked in the US' + } } def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - video_id = mobj.group(1) - - info_url = ('http://www.clipfish.de/devxml/videoinfo/%s?ts=%d' % - (video_id, int(time.time()))) - doc = self._download_xml( - info_url, video_id, note='Downloading info page') - title = doc.find('title').text - video_url = doc.find('filename').text - if video_url is None: - xml_bytes = xml.etree.ElementTree.tostring(doc) - raise ExtractorError('Cannot find video URL in document %r' % - xml_bytes) - thumbnail = doc.find('imageurl').text - duration = parse_duration(doc.find('duration').text) + video_id = self._match_id(url) + + webpage = self._download_webpage(url, video_id) + + video_info = self._parse_json( + js_to_json(self._html_search_regex( + '(?s)videoObject\s*=\s*({.+?});', webpage, 'video object')), + video_id) + + formats = [] + for video_url in re.findall(r'var\s+videourl\s*=\s*"([^"]+)"', webpage): + ext = determine_ext(video_url) + if ext == 'm3u8': + formats.append({ + 'url': video_url.replace('de.hls.fra.clipfish.de', 'hls.fra.clipfish.de'), + 'ext': 'mp4', + 'format_id': 'hls', + }) + else: + formats.append({ + 'url': video_url, + 'format_id': ext, + }) + self._sort_formats(formats) + + title = remove_end(self._og_search_title(webpage), ' - Video') + thumbnail = self._og_search_thumbnail(webpage) + duration = int_or_none(video_info.get('length')) + timestamp = parse_iso8601(self._html_search_meta('uploadDate', webpage, 'upload date')) return { 'id': video_id, 'title': title, - 'url': video_url, + 'formats': formats, 'thumbnail': thumbnail, 'duration': duration, + 'timestamp': timestamp, } diff --git a/youtube_dl/extractor/comcarcoff.py b/youtube_dl/extractor/comcarcoff.py index 9c25b2223..81f3d7697 100644 --- a/youtube_dl/extractor/comcarcoff.py +++ b/youtube_dl/extractor/comcarcoff.py @@ -36,7 +36,7 @@ class ComCarCoffIE(InfoExtractor): webpage, 'full data json')) video_id = full_data['activeVideo']['video'] - video_data = full_data['videos'][video_id] + video_data = full_data.get('videos', {}).get(video_id) or full_data['singleshots'][video_id] thumbnails = [{ 'url': video_data['images']['thumb'], }, { diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py index b9014fc23..16ae4b98f 100644 --- a/youtube_dl/extractor/common.py +++ b/youtube_dl/extractor/common.py @@ -14,10 +14,13 @@ import xml.etree.ElementTree from ..compat import ( compat_cookiejar, + compat_cookies, compat_HTTPError, compat_http_client, compat_urllib_error, + compat_urllib_parse, compat_urllib_parse_urlparse, + compat_urllib_request, compat_urlparse, compat_str, ) @@ -35,6 +38,9 @@ from ..utils import ( RegexNotFoundError, sanitize_filename, unescapeHTML, + url_basename, + xpath_text, + xpath_with_ns, ) @@ -65,7 +71,7 @@ class InfoExtractor(object): Potential fields: * url Mandatory. The URL of the video file - * ext Will be calculated from url if missing + * ext Will be calculated from URL if missing * format A human-readable description of the format ("mp4 container with h264/opus"). Calculated from the format_id, width, height. @@ -155,7 +161,7 @@ class InfoExtractor(object): lower to higher preference, each element is a dictionary with the "ext" entry and one of: * "data": The subtitles file contents - * "url": A url pointing to the subtitles file + * "url": A URL pointing to the subtitles file automatic_captions: Like 'subtitles', used by the YoutubeIE for automatically generated captions duration: Length of the video in seconds, as an integer. @@ -176,13 +182,18 @@ class InfoExtractor(object): Set to "root" to indicate that this is a comment to the original video. age_limit: Age restriction for the video, as an integer (years) - webpage_url: The url to the video webpage, if given to youtube-dl it + webpage_url: The URL to the video webpage, if given to youtube-dl it should allow to get the same result again. (It will be set by YoutubeDL if it's missing) categories: A list of categories that the video falls in, for example ["Sports", "Berlin"] + tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"] is_live: True, False, or None (=unknown). Whether this video is a live stream that goes on instead of a fixed-length video. + start_time: Time in seconds where the reproduction should start, as + specified in the URL. + end_time: Time in seconds where the reproduction should end, as + specified in the URL. Unless mentioned otherwise, the fields should be Unicode strings. @@ -193,8 +204,8 @@ class InfoExtractor(object): There must be a key "entries", which is a list, an iterable, or a PagedList object, each element of which is a valid dictionary by this specification. - Additionally, playlists can have "title" and "id" attributes with the same - semantics as videos (see above). + Additionally, playlists can have "title", "description" and "id" attributes + with the same semantics as videos (see above). _type "multi_video" indicates that there are multiple videos that @@ -501,7 +512,7 @@ class InfoExtractor(object): # Methods for following #608 @staticmethod def url_result(url, ie=None, video_id=None, video_title=None): - """Returns a url that points to a page that should be processed""" + """Returns a URL that points to a page that should be processed""" # TODO: ie should be the class used for getting the info video_info = {'_type': 'url', 'url': url, @@ -626,6 +637,12 @@ class InfoExtractor(object): template % (content_re, property_re), ] + @staticmethod + def _meta_regex(prop): + return r'''(?isx)]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1) + [^>]+?content=(["\'])(?P.*?)\2''' % re.escape(prop) + def _og_search_property(self, prop, html, name=None, **kargs): if name is None: name = 'OpenGraph %s' % prop @@ -635,7 +652,7 @@ class InfoExtractor(object): return unescapeHTML(escaped) def _og_search_thumbnail(self, html, **kargs): - return self._og_search_property('image', html, 'thumbnail url', fatal=False, **kargs) + return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs) def _og_search_description(self, html, **kargs): return self._og_search_property('description', html, fatal=False, **kargs) @@ -656,9 +673,7 @@ class InfoExtractor(object): if display_name is None: display_name = name return self._html_search_regex( - r'''(?isx)]+(?:itemprop|name|property)=(["\']?)%s\1) - [^>]+?content=(["\'])(?P.*?)\2''' % re.escape(name), + self._meta_regex(name), html, display_name, fatal=fatal, group='content', **kwargs) def _dc_search_uploader(self, html): @@ -967,69 +982,210 @@ class InfoExtractor(object): self._sort_formats(formats) return formats - # TODO: improve extraction - def _extract_smil_formats(self, smil_url, video_id, fatal=True): - smil = self._download_xml( - smil_url, video_id, 'Downloading SMIL file', - 'Unable to download SMIL file', fatal=fatal) + @staticmethod + def _xpath_ns(path, namespace=None): + if not namespace: + return path + out = [] + for c in path.split('/'): + if not c or c == '.': + out.append(c) + else: + out.append('{%s}%s' % (namespace, c)) + return '/'.join(out) + + def _extract_smil_formats(self, smil_url, video_id, fatal=True, f4m_params=None): + smil = self._download_smil(smil_url, video_id, fatal=fatal) + if smil is False: assert not fatal return [] - base = smil.find('./head/meta').get('base') + namespace = self._parse_smil_namespace(smil) + + return self._parse_smil_formats( + smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) + + def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None): + smil = self._download_smil(smil_url, video_id, fatal=fatal) + if smil is False: + return {} + return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params) + + def _download_smil(self, smil_url, video_id, fatal=True): + return self._download_xml( + smil_url, video_id, 'Downloading SMIL file', + 'Unable to download SMIL file', fatal=fatal) + + def _parse_smil(self, smil, smil_url, video_id, f4m_params=None): + namespace = self._parse_smil_namespace(smil) + + formats = self._parse_smil_formats( + smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) + subtitles = self._parse_smil_subtitles(smil, namespace=namespace) + + video_id = os.path.splitext(url_basename(smil_url))[0] + title = None + description = None + for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): + name = meta.attrib.get('name') + content = meta.attrib.get('content') + if not name or not content: + continue + if not title and name == 'title': + title = content + elif not description and name in ('description', 'abstract'): + description = content + + return { + 'id': video_id, + 'title': title or video_id, + 'description': description, + 'formats': formats, + 'subtitles': subtitles, + } + + def _parse_smil_namespace(self, smil): + return self._search_regex( + r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None) + + def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None): + base = smil_url + for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): + b = meta.get('base') or meta.get('httpBase') + if b: + base = b + break formats = [] rtmp_count = 0 - if smil.findall('./body/seq/video'): - video = smil.findall('./body/seq/video')[0] - fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count) - formats.extend(fmts) - else: - for video in smil.findall('./body/switch/video'): - fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count) - formats.extend(fmts) + http_count = 0 + + videos = smil.findall(self._xpath_ns('.//video', namespace)) + for video in videos: + src = video.get('src') + if not src: + continue + + bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000) + filesize = int_or_none(video.get('size') or video.get('fileSize')) + width = int_or_none(video.get('width')) + height = int_or_none(video.get('height')) + proto = video.get('proto') + ext = video.get('ext') + src_ext = determine_ext(src) + streamer = video.get('streamer') or base + + if proto == 'rtmp' or streamer.startswith('rtmp'): + rtmp_count += 1 + formats.append({ + 'url': streamer, + 'play_path': src, + 'ext': 'flv', + 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate), + 'tbr': bitrate, + 'filesize': filesize, + 'width': width, + 'height': height, + }) + continue + + src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src) + + if proto == 'm3u8' or src_ext == 'm3u8': + formats.extend(self._extract_m3u8_formats( + src_url, video_id, ext or 'mp4', m3u8_id='hls')) + continue + + if src_ext == 'f4m': + f4m_url = src_url + if not f4m_params: + f4m_params = { + 'hdcore': '3.2.0', + 'plugin': 'flowplayer-3.2.0.1', + } + f4m_url += '&' if '?' in f4m_url else '?' + f4m_url += compat_urllib_parse.urlencode(f4m_params) + formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds')) + continue + + if src_url.startswith('http'): + http_count += 1 + formats.append({ + 'url': src_url, + 'ext': ext or src_ext or 'flv', + 'format_id': 'http-%d' % (bitrate or http_count), + 'tbr': bitrate, + 'filesize': filesize, + 'width': width, + 'height': height, + }) + continue self._sort_formats(formats) return formats - def _parse_smil_video(self, video, video_id, base, rtmp_count): - src = video.get('src') - if not src: - return [], rtmp_count - bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000) - width = int_or_none(video.get('width')) - height = int_or_none(video.get('height')) - proto = video.get('proto') - if not proto: - if base: - if base.startswith('rtmp'): - proto = 'rtmp' - elif base.startswith('http'): - proto = 'http' - ext = video.get('ext') - if proto == 'm3u8': - return self._extract_m3u8_formats(src, video_id, ext), rtmp_count - elif proto == 'rtmp': - rtmp_count += 1 - streamer = video.get('streamer') or base - return ([{ - 'url': streamer, - 'play_path': src, - 'ext': 'flv', - 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate), - 'tbr': bitrate, - 'width': width, - 'height': height, - }], rtmp_count) - elif proto.startswith('http'): - return ([{ - 'url': base + src, - 'ext': ext or 'flv', - 'tbr': bitrate, - 'width': width, - 'height': height, - }], rtmp_count) + def _parse_smil_subtitles(self, smil, namespace=None): + subtitles = {} + for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))): + src = textstream.get('src') + if not src: + continue + ext = textstream.get('ext') or determine_ext(src) + if not ext: + type_ = textstream.get('type') + if type_ == 'text/srt': + ext = 'srt' + lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') + subtitles.setdefault(lang, []).append({ + 'url': src, + 'ext': ext, + }) + return subtitles + + def _extract_xspf_playlist(self, playlist_url, playlist_id, fatal=True): + xspf = self._download_xml( + playlist_url, playlist_id, 'Downloading xpsf playlist', + 'Unable to download xspf manifest', fatal=fatal) + if xspf is False: + return [] + return self._parse_xspf(xspf, playlist_id) + + def _parse_xspf(self, playlist, playlist_id): + NS_MAP = { + 'xspf': 'http://xspf.org/ns/0/', + 's1': 'http://static.streamone.nl/player/ns/0', + } + + entries = [] + for track in playlist.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)): + title = xpath_text( + track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id) + description = xpath_text( + track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description') + thumbnail = xpath_text( + track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail') + duration = float_or_none( + xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000) + + formats = [{ + 'url': location.text, + 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)), + 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))), + 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))), + } for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP))] + self._sort_formats(formats) + + entries.append({ + 'id': playlist_id, + 'title': title, + 'description': description, + 'thumbnail': thumbnail, + 'duration': duration, + 'formats': formats, + }) + return entries def _live_title(self, name): """ Generate the title for a live video """ @@ -1065,6 +1221,12 @@ class InfoExtractor(object): None, '/', True, False, expire_time, '', None, None, None) self._downloader.cookiejar.set_cookie(cookie) + def _get_cookies(self, url): + """ Return a compat_cookies.SimpleCookie with the cookies for the url """ + req = compat_urllib_request.Request(url) + self._downloader.cookiejar.add_cookie_header(req) + return compat_cookies.SimpleCookie(req.get_header('Cookie')) + def get_testcases(self, include_onlymatching=False): t = getattr(self, '_TEST', None) if t: @@ -1116,7 +1278,7 @@ class InfoExtractor(object): class SearchInfoExtractor(InfoExtractor): """ Base class for paged search queries extractors. - They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query} + They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query} Instances should define _SEARCH_KEY and _MAX_RESULTS. """ diff --git a/youtube_dl/extractor/dailymotion.py b/youtube_dl/extractor/dailymotion.py index 1a41c0db1..2d90b2224 100644 --- a/youtube_dl/extractor/dailymotion.py +++ b/youtube_dl/extractor/dailymotion.py @@ -13,8 +13,9 @@ from ..compat import ( ) from ..utils import ( ExtractorError, + determine_ext, int_or_none, - orderedSet, + parse_iso8601, str_to_int, unescapeHTML, ) @@ -28,10 +29,16 @@ class DailymotionBaseInfoExtractor(InfoExtractor): request.add_header('Cookie', 'family_filter=off; ff=off') return request + def _download_webpage_handle_no_ff(self, url, *args, **kwargs): + request = self._build_request(url) + return self._download_webpage_handle(request, *args, **kwargs) + + def _download_webpage_no_ff(self, url, *args, **kwargs): + request = self._build_request(url) + return self._download_webpage(request, *args, **kwargs) -class DailymotionIE(DailymotionBaseInfoExtractor): - """Information Extractor for Dailymotion""" +class DailymotionIE(DailymotionBaseInfoExtractor): _VALID_URL = r'(?i)(?:https?://)?(?:(www|touch)\.)?dailymotion\.[a-z]{2,3}/(?:(embed|#)/)?video/(?P[^/?_]+)' IE_NAME = 'dailymotion' @@ -50,10 +57,17 @@ class DailymotionIE(DailymotionBaseInfoExtractor): 'info_dict': { 'id': 'x2iuewm', 'ext': 'mp4', - 'uploader': 'IGN', 'title': 'Steam Machine Models, Pricing Listed on Steam Store - IGN News', - 'upload_date': '20150306', + 'description': 'Several come bundled with the Steam Controller.', + 'thumbnail': 're:^https?:.*\.(?:jpg|png)$', 'duration': 74, + 'timestamp': 1425657362, + 'upload_date': '20150306', + 'uploader': 'IGN', + 'uploader_id': 'xijv66', + 'age_limit': 0, + 'view_count': int, + 'comment_count': int, } }, # Vevo video @@ -87,38 +101,106 @@ class DailymotionIE(DailymotionBaseInfoExtractor): def _real_extract(self, url): video_id = self._match_id(url) - url = 'https://www.dailymotion.com/video/%s' % video_id - # Retrieve video webpage to extract further information - request = self._build_request(url) - webpage = self._download_webpage(request, video_id) + webpage = self._download_webpage_no_ff( + 'https://www.dailymotion.com/video/%s' % video_id, video_id) + + age_limit = self._rta_search(webpage) - # Extract URL, uploader and title from webpage - self.report_extraction(video_id) + description = self._og_search_description(webpage) or self._html_search_meta( + 'description', webpage, 'description') - # It may just embed a vevo video: - m_vevo = re.search( + view_count = str_to_int(self._search_regex( + [r']+itemprop="interactionCount"[^>]+content="UserPlays:(\d+)"', + r'video_views_count[^>]+>\s+([\d\.,]+)'], + webpage, 'view count', fatal=False)) + comment_count = int_or_none(self._search_regex( + r']+itemprop="interactionCount"[^>]+content="UserComments:(\d+)"', + webpage, 'comment count', fatal=False)) + + player_v5 = self._search_regex( + r'playerV5\s*=\s*dmp\.create\([^,]+?,\s*({.+?})\);', + webpage, 'player v5', default=None) + if player_v5: + player = self._parse_json(player_v5, video_id) + metadata = player['metadata'] + formats = [] + for quality, media_list in metadata['qualities'].items(): + for media in media_list: + media_url = media.get('url') + if not media_url: + continue + type_ = media.get('type') + if type_ == 'application/vnd.lumberjack.manifest': + continue + if type_ == 'application/x-mpegURL' or determine_ext(media_url) == 'm3u8': + formats.extend(self._extract_m3u8_formats( + media_url, video_id, 'mp4', m3u8_id='hls')) + else: + f = { + 'url': media_url, + 'format_id': quality, + } + m = re.search(r'H264-(?P\d+)x(?P\d+)', media_url) + if m: + f.update({ + 'width': int(m.group('width')), + 'height': int(m.group('height')), + }) + formats.append(f) + self._sort_formats(formats) + + title = metadata['title'] + duration = int_or_none(metadata.get('duration')) + timestamp = int_or_none(metadata.get('created_time')) + thumbnail = metadata.get('poster_url') + uploader = metadata.get('owner', {}).get('screenname') + uploader_id = metadata.get('owner', {}).get('id') + + subtitles = {} + for subtitle_lang, subtitle in metadata.get('subtitles', {}).get('data', {}).items(): + subtitles[subtitle_lang] = [{ + 'ext': determine_ext(subtitle_url), + 'url': subtitle_url, + } for subtitle_url in subtitle.get('urls', [])] + + return { + 'id': video_id, + 'title': title, + 'description': description, + 'thumbnail': thumbnail, + 'duration': duration, + 'timestamp': timestamp, + 'uploader': uploader, + 'uploader_id': uploader_id, + 'age_limit': age_limit, + 'view_count': view_count, + 'comment_count': comment_count, + 'formats': formats, + 'subtitles': subtitles, + } + + # vevo embed + vevo_id = self._search_regex( r'[\w]*)', - webpage) - if m_vevo is not None: - vevo_id = m_vevo.group('id') - self.to_screen('Vevo video detected: %s' % vevo_id) - return self.url_result('vevo:%s' % vevo_id, ie='Vevo') + webpage, 'vevo embed', default=None) + if vevo_id: + return self.url_result('vevo:%s' % vevo_id, 'Vevo') - age_limit = self._rta_search(webpage) + # fallback old player + embed_page = self._download_webpage_no_ff( + 'https://www.dailymotion.com/embed/video/%s' % video_id, + video_id, 'Downloading embed page') + + timestamp = parse_iso8601(self._html_search_meta( + 'video:release_date', webpage, 'upload date')) + + info = self._parse_json( + self._search_regex( + r'var info = ({.*?}),$', embed_page, + 'video info', flags=re.MULTILINE), + video_id) - video_upload_date = None - mobj = re.search(r'', webpage) - if mobj is not None: - video_upload_date = mobj.group(1) + mobj.group(2) + mobj.group(3) - - embed_url = 'https://www.dailymotion.com/embed/video/%s' % video_id - embed_request = self._build_request(embed_url) - embed_page = self._download_webpage( - embed_request, video_id, 'Downloading embed page') - info = self._search_regex(r'var info = ({.*?}),$', embed_page, - 'video info', flags=re.MULTILINE) - info = json.loads(info) if info.get('error') is not None: msg = 'Couldn\'t get video, Dailymotion says: %s' % info['error']['title'] raise ExtractorError(msg, expected=True) @@ -139,16 +221,11 @@ class DailymotionIE(DailymotionBaseInfoExtractor): 'width': width, 'height': height, }) - if not formats: - raise ExtractorError('Unable to extract video URL') + self._sort_formats(formats) # subtitles video_subtitles = self.extract_subtitles(video_id, webpage) - view_count = str_to_int(self._search_regex( - r'video_views_count[^>]+>\s+([\d\.,]+)', - webpage, 'view count', fatal=False)) - title = self._og_search_title(webpage, default=None) if title is None: title = self._html_search_regex( @@ -159,8 +236,9 @@ class DailymotionIE(DailymotionBaseInfoExtractor): 'id': video_id, 'formats': formats, 'uploader': info['owner.screenname'], - 'upload_date': video_upload_date, + 'timestamp': timestamp, 'title': title, + 'description': description, 'subtitles': video_subtitles, 'thumbnail': info['thumbnail_url'], 'age_limit': age_limit, @@ -199,18 +277,26 @@ class DailymotionPlaylistIE(DailymotionBaseInfoExtractor): }] def _extract_entries(self, id): - video_ids = [] + video_ids = set() + processed_urls = set() for pagenum in itertools.count(1): - request = self._build_request(self._PAGE_TEMPLATE % (id, pagenum)) - webpage = self._download_webpage(request, - id, 'Downloading page %s' % pagenum) + page_url = self._PAGE_TEMPLATE % (id, pagenum) + webpage, urlh = self._download_webpage_handle_no_ff( + page_url, id, 'Downloading page %s' % pagenum) + if urlh.geturl() in processed_urls: + self.report_warning('Stopped at duplicated page %s, which is the same as %s' % ( + page_url, urlh.geturl()), id) + break - video_ids.extend(re.findall(r'data-xid="(.+?)"', webpage)) + processed_urls.add(urlh.geturl()) + + for video_id in re.findall(r'data-xid="(.+?)"', webpage): + if video_id not in video_ids: + yield self.url_result('http://www.dailymotion.com/video/%s' % video_id, 'Dailymotion') + video_ids.add(video_id) if re.search(self._MORE_PAGES_INDICATOR, webpage) is None: break - return [self.url_result('http://www.dailymotion.com/video/%s' % video_id, 'Dailymotion') - for video_id in orderedSet(video_ids)] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) @@ -227,7 +313,7 @@ class DailymotionPlaylistIE(DailymotionBaseInfoExtractor): class DailymotionUserIE(DailymotionPlaylistIE): IE_NAME = 'dailymotion:user' - _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?:(?:old/)?user/)?(?P[^/]+)$' + _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?!(?:embed|#|video|playlist)/)(?:(?:old/)?user/)?(?P[^/]+)' _PAGE_TEMPLATE = 'http://www.dailymotion.com/user/%s/%s' _TESTS = [{ 'url': 'https://www.dailymotion.com/user/nqtv', @@ -236,6 +322,17 @@ class DailymotionUserIE(DailymotionPlaylistIE): 'title': 'Rémi Gaillard', }, 'playlist_mincount': 100, + }, { + 'url': 'http://www.dailymotion.com/user/UnderProject', + 'info_dict': { + 'id': 'UnderProject', + 'title': 'UnderProject', + }, + 'playlist_mincount': 1800, + 'expected_warnings': [ + 'Stopped at duplicated page', + ], + 'skip': 'Takes too long time', }] def _real_extract(self, url): @@ -286,8 +383,7 @@ class DailymotionCloudIE(DailymotionBaseInfoExtractor): def _real_extract(self, url): video_id = self._match_id(url) - request = self._build_request(url) - webpage = self._download_webpage(request, video_id) + webpage = self._download_webpage_no_ff(url, video_id) title = self._html_search_regex(r'([^>]+)', webpage, 'title') diff --git a/youtube_dl/extractor/dcn.py b/youtube_dl/extractor/dcn.py new file mode 100644 index 000000000..82261e25c --- /dev/null +++ b/youtube_dl/extractor/dcn.py @@ -0,0 +1,84 @@ +# coding: utf-8 +from __future__ import unicode_literals + +from .common import InfoExtractor +from ..compat import ( + compat_urllib_parse, + compat_urllib_request, +) +from ..utils import ( + int_or_none, + parse_iso8601, +) + + +class DCNIE(InfoExtractor): + _VALID_URL = r'https?://(?:www\.)?dcndigital\.ae/(?:#/)?(?:video/.+|show/\d+/.+?)/(?P\d+)' + _TEST = { + 'url': 'http://www.dcndigital.ae/#/show/199074/%D8%B1%D8%AD%D9%84%D8%A9-%D8%A7%D9%84%D8%B9%D9%85%D8%B1-%D8%A7%D9%84%D8%AD%D9%84%D9%82%D8%A9-1/17375/6887', + 'info_dict': + { + 'id': '17375', + 'ext': 'mp4', + 'title': 'رحلة العمر : الحلقة 1', + 'description': 'md5:0156e935d870acb8ef0a66d24070c6d6', + 'thumbnail': 're:^https?://.*\.jpg$', + 'duration': 2041, + 'timestamp': 1227504126, + 'upload_date': '20081124', + }, + 'params': { + # m3u8 download + 'skip_download': True, + }, + } + + def _real_extract(self, url): + video_id = self._match_id(url) + + request = compat_urllib_request.Request( + 'http://admin.mangomolo.com/analytics/index.php/plus/video?id=%s' % video_id, + headers={'Origin': 'http://www.dcndigital.ae'}) + + video = self._download_json(request, video_id) + title = video.get('title_en') or video['title_ar'] + + webpage = self._download_webpage( + 'http://admin.mangomolo.com/analytics/index.php/customers/embed/video?' + + compat_urllib_parse.urlencode({ + 'id': video['id'], + 'user_id': video['user_id'], + 'signature': video['signature'], + 'countries': 'Q0M=', + 'filter': 'DENY', + }), video_id) + + m3u8_url = self._html_search_regex(r'file:\s*"([^"]+)', webpage, 'm3u8 url') + formats = self._extract_m3u8_formats( + m3u8_url, video_id, 'mp4', entry_protocol='m3u8_native', m3u8_id='hls') + + rtsp_url = self._search_regex( + r']+href="(rtsp://[^"]+)"', webpage, 'rtsp url', fatal=False) + if rtsp_url: + formats.append({ + 'url': rtsp_url, + 'format_id': 'rtsp', + }) + + self._sort_formats(formats) + + img = video.get('img') + thumbnail = 'http://admin.mangomolo.com/analytics/%s' % img if img else None + duration = int_or_none(video.get('duration')) + description = video.get('description_en') or video.get('description_ar') + timestamp = parse_iso8601(video.get('create_time') or video.get('update_time'), ' ') + + return { + 'id': video_id, + 'title': title, + 'description': description, + 'thumbnail': thumbnail, + 'duration': duration, + 'timestamp': timestamp, + 'formats': formats, + } diff --git a/youtube_dl/extractor/dhm.py b/youtube_dl/extractor/dhm.py index 3ed1f1663..44e0c5d4d 100644 --- a/youtube_dl/extractor/dhm.py +++ b/youtube_dl/extractor/dhm.py @@ -1,10 +1,7 @@ from __future__ import unicode_literals from .common import InfoExtractor -from ..utils import ( - xpath_text, - parse_duration, -) +from ..utils import parse_duration class DHMIE(InfoExtractor): @@ -34,24 +31,14 @@ class DHMIE(InfoExtractor): }] def _real_extract(self, url): - video_id = self._match_id(url) + playlist_id = self._match_id(url) - webpage = self._download_webpage(url, video_id) + webpage = self._download_webpage(url, playlist_id) playlist_url = self._search_regex( r"file\s*:\s*'([^']+)'", webpage, 'playlist url') - playlist = self._download_xml(playlist_url, video_id) - - track = playlist.find( - './{http://xspf.org/ns/0/}trackList/{http://xspf.org/ns/0/}track') - - video_url = xpath_text( - track, './{http://xspf.org/ns/0/}location', - 'video url', fatal=True) - thumbnail = xpath_text( - track, './{http://xspf.org/ns/0/}image', - 'thumbnail') + entries = self._extract_xspf_playlist(playlist_url, playlist_id) title = self._search_regex( [r'dc:title="([^"]+)"', r' »([^<]+)'], @@ -63,11 +50,10 @@ class DHMIE(InfoExtractor): r'Length\s*\s*:\s*([^<]+)', webpage, 'duration', default=None)) - return { - 'id': video_id, - 'url': video_url, + entries[0].update({ 'title': title, 'description': description, 'duration': duration, - 'thumbnail': thumbnail, - } + }) + + return self.playlist_result(entries, playlist_id) diff --git a/youtube_dl/extractor/esri.py b/youtube_dl/extractor/esri.py new file mode 100644 index 000000000..bf5d2019f --- /dev/null +++ b/youtube_dl/extractor/esri.py @@ -0,0 +1,74 @@ +# coding: utf-8 +from __future__ import unicode_literals + +import re + +from .common import InfoExtractor +from ..compat import compat_urlparse +from ..utils import ( + int_or_none, + parse_filesize, + unified_strdate, +) + + +class EsriVideoIE(InfoExtractor): + _VALID_URL = r'https?://video\.esri\.com/watch/(?P[0-9]+)' + _TEST = { + 'url': 'https://video.esri.com/watch/1124/arcgis-online-_dash_-developing-applications', + 'md5': 'd4aaf1408b221f1b38227a9bbaeb95bc', + 'info_dict': { + 'id': '1124', + 'ext': 'mp4', + 'title': 'ArcGIS Online - Developing Applications', + 'description': 'Jeremy Bartley demonstrates how to develop applications with ArcGIS Online.', + 'thumbnail': 're:^https?://.*\.jpg$', + 'duration': 185, + 'upload_date': '20120419', + } + } + + def _real_extract(self, url): + video_id = self._match_id(url) + + webpage = self._download_webpage(url, video_id) + + formats = [] + for width, height, content in re.findall( + r'(?s)
  • (\d+)x(\d+):(.+?)
    • ', webpage): + for video_url, ext, filesize in re.findall( + r']+href="([^"]+)">([^<]+) \(([^<]+)\)', content): + formats.append({ + 'url': compat_urlparse.urljoin(url, video_url), + 'ext': ext.lower(), + 'format_id': '%s-%s' % (ext.lower(), height), + 'width': int(width), + 'height': int(height), + 'filesize_approx': parse_filesize(filesize), + }) + self._sort_formats(formats) + + title = self._html_search_meta('title', webpage, 'title') + description = self._html_search_meta( + 'description', webpage, 'description', fatal=False) + + thumbnail = self._html_search_meta('thumbnail', webpage, 'thumbnail', fatal=False) + if thumbnail: + thumbnail = re.sub(r'_[st]\.jpg$', '_x.jpg', thumbnail) + + duration = int_or_none(self._search_regex( + [r'var\s+videoSeconds\s*=\s*(\d+)', r"'duration'\s*:\s*(\d+)"], + webpage, 'duration', fatal=False)) + + upload_date = unified_strdate(self._html_search_meta( + 'last-modified', webpage, 'upload date', fatal=None)) + + return { + 'id': video_id, + 'title': title, + 'description': description, + 'thumbnail': thumbnail, + 'duration': duration, + 'upload_date': upload_date, + 'formats': formats + } diff --git a/youtube_dl/extractor/facebook.py b/youtube_dl/extractor/facebook.py index e17bb9aea..178a7ca4c 100644 --- a/youtube_dl/extractor/facebook.py +++ b/youtube_dl/extractor/facebook.py @@ -17,6 +17,8 @@ from ..utils import ( int_or_none, limit_length, urlencode_postdata, + get_element_by_id, + clean_html, ) @@ -42,6 +44,7 @@ class FacebookIE(InfoExtractor): 'id': '637842556329505', 'ext': 'mp4', 'title': 're:Did you know Kei Nishikori is the first Asian man to ever reach a Grand Slam', + 'uploader': 'Tennis on Facebook', } }, { 'note': 'Video without discernible title', @@ -50,6 +53,7 @@ class FacebookIE(InfoExtractor): 'id': '274175099429670', 'ext': 'mp4', 'title': 'Facebook video #274175099429670', + 'uploader': 'Asif Nawab Butt', }, 'expected_warnings': [ 'title' @@ -161,6 +165,7 @@ class FacebookIE(InfoExtractor): video_title = limit_length(video_title, 80) if not video_title: video_title = 'Facebook video #%s' % video_id + uploader = clean_html(get_element_by_id('fbPhotoPageAuthorName', webpage)) return { 'id': video_id, @@ -168,4 +173,5 @@ class FacebookIE(InfoExtractor): 'formats': formats, 'duration': int_or_none(video_data.get('video_duration')), 'thumbnail': video_data.get('thumbnail_src'), + 'uploader': uploader, } diff --git a/youtube_dl/extractor/fc2.py b/youtube_dl/extractor/fc2.py index 1ccc1a964..e4f7195a8 100644 --- a/youtube_dl/extractor/fc2.py +++ b/youtube_dl/extractor/fc2.py @@ -86,7 +86,7 @@ class FC2IE(InfoExtractor): info_url = ( "http://video.fc2.com/ginfo.php?mimi={1:s}&href={2:s}&v={0:s}&fversion=WIN%2011%2C6%2C602%2C180&from=2&otag=0&upid={0:s}&tk=null&". - format(video_id, mimi, compat_urllib_request.quote(refer, safe='').replace('.', '%2E'))) + format(video_id, mimi, compat_urllib_request.quote(refer, safe=b'').replace('.', '%2E'))) info_webpage = self._download_webpage( info_url, video_id, note='Downloading info page') diff --git a/youtube_dl/extractor/fourtube.py b/youtube_dl/extractor/fourtube.py index b2284ab01..3bb4f6239 100644 --- a/youtube_dl/extractor/fourtube.py +++ b/youtube_dl/extractor/fourtube.py @@ -32,6 +32,7 @@ class FourTubeIE(InfoExtractor): 'view_count': int, 'like_count': int, 'categories': list, + 'age_limit': 18, } } diff --git a/youtube_dl/extractor/francetv.py b/youtube_dl/extractor/francetv.py index b2c984bf2..75723c00d 100644 --- a/youtube_dl/extractor/francetv.py +++ b/youtube_dl/extractor/francetv.py @@ -6,15 +6,11 @@ import re import json from .common import InfoExtractor -from ..compat import ( - compat_urllib_parse_urlparse, - compat_urlparse, -) +from ..compat import compat_urlparse from ..utils import ( clean_html, ExtractorError, int_or_none, - float_or_none, parse_duration, determine_ext, ) @@ -59,12 +55,12 @@ class FranceTVBaseInfoExtractor(InfoExtractor): # See https://github.com/rg3/youtube-dl/issues/3963 # m3u8 urls work fine continue - video_url_parsed = compat_urllib_parse_urlparse(video_url) f4m_url = self._download_webpage( - 'http://hdfauth.francetv.fr/esi/TA?url=%s' % video_url_parsed.path, + 'http://hdfauth.francetv.fr/esi/TA?url=%s' % video_url, video_id, 'Downloading f4m manifest token', fatal=False) if f4m_url: - formats.extend(self._extract_f4m_formats(f4m_url, video_id, 1, format_id)) + formats.extend(self._extract_f4m_formats( + f4m_url + '&hdcore=3.7.0&plugin=aasp-3.7.0.39.44', video_id, 1, format_id)) elif ext == 'm3u8': formats.extend(self._extract_m3u8_formats(video_url, video_id, 'mp4', m3u8_id=format_id)) elif video_url.startswith('rtmp'): @@ -87,7 +83,7 @@ class FranceTVBaseInfoExtractor(InfoExtractor): 'title': info['titre'], 'description': clean_html(info['synopsis']), 'thumbnail': compat_urlparse.urljoin('http://pluzz.francetv.fr', info['image']), - 'duration': float_or_none(info.get('real_duration'), 1000) or parse_duration(info['duree']), + 'duration': int_or_none(info.get('real_duration')) or parse_duration(info['duree']), 'timestamp': int_or_none(info['diffusion']['timestamp']), 'formats': formats, } @@ -160,11 +156,21 @@ class FranceTvInfoIE(FranceTVBaseInfoExtractor): class FranceTVIE(FranceTVBaseInfoExtractor): IE_NAME = 'francetv' IE_DESC = 'France 2, 3, 4, 5 and Ô' - _VALID_URL = r'''(?x)https?://www\.france[2345o]\.fr/ - (?: - emissions/.*?/(videos|emissions)/(?P[^/?]+) - | (emissions?|jt)/(?P[^/?]+) - )''' + _VALID_URL = r'''(?x) + https?:// + (?: + (?:www\.)?france[2345o]\.fr/ + (?: + emissions/[^/]+/(?:videos|diffusions)| + emission/[^/]+| + videos| + jt + ) + /| + embed\.francetv\.fr/\?ue= + ) + (?P[^/?]+) + ''' _TESTS = [ # france2 @@ -221,24 +227,46 @@ class FranceTVIE(FranceTVBaseInfoExtractor): }, # franceo { - 'url': 'http://www.franceo.fr/jt/info-afrique/04-12-2013', - 'md5': '52f0bfe202848b15915a2f39aaa8981b', + 'url': 'http://www.franceo.fr/jt/info-soir/18-07-2015', + 'md5': '47d5816d3b24351cdce512ad7ab31da8', 'info_dict': { - 'id': '108634970', + 'id': '125377621', 'ext': 'flv', - 'title': 'Infô Afrique', - 'description': 'md5:ebf346da789428841bee0fd2a935ea55', - 'upload_date': '20140915', - 'timestamp': 1410822000, + 'title': 'Infô soir', + 'description': 'md5:01b8c6915a3d93d8bbbd692651714309', + 'upload_date': '20150718', + 'timestamp': 1437241200, + 'duration': 414, + }, + }, + { + # francetv embed + 'url': 'http://embed.francetv.fr/?ue=8d7d3da1e3047c42ade5a5d7dfd3fc87', + 'info_dict': { + 'id': 'EV_30231', + 'ext': 'flv', + 'title': 'Alcaline, le concert avec Calogero', + 'description': 'md5:61f08036dcc8f47e9cfc33aed08ffaff', + 'upload_date': '20150226', + 'timestamp': 1424989860, + 'duration': 5400, }, }, + { + 'url': 'http://www.france4.fr/emission/highlander/diffusion-du-17-07-2015-04h05', + 'only_matching': True, + }, + { + 'url': 'http://www.franceo.fr/videos/125377617', + 'only_matching': True, + } ] def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - webpage = self._download_webpage(url, mobj.group('key') or mobj.group('id')) + video_id = self._match_id(url) + webpage = self._download_webpage(url, video_id) video_id, catalogue = self._html_search_regex( - r'href="http://videos\.francetv\.fr/video/([^@]+@[^"]+)"', + r'href="http://videos?\.francetv\.fr/video/([^@]+@[^"]+)"', webpage, 'video ID').split('@') return self._extract_video(video_id, catalogue) diff --git a/youtube_dl/extractor/funnyordie.py b/youtube_dl/extractor/funnyordie.py index dd87257c4..f5f13689c 100644 --- a/youtube_dl/extractor/funnyordie.py +++ b/youtube_dl/extractor/funnyordie.py @@ -53,7 +53,7 @@ class FunnyOrDieIE(InfoExtractor): for bitrate in bitrates: for link in links: formats.append({ - 'url': '%s%d.%s' % (link[0], bitrate, link[1]), + 'url': self._proto_relative_url('%s%d.%s' % (link[0], bitrate, link[1])), 'format_id': '%s-%d' % (link[1], bitrate), 'vbr': bitrate, }) diff --git a/youtube_dl/extractor/gdcvault.py b/youtube_dl/extractor/gdcvault.py index 43f916412..a6834db43 100644 --- a/youtube_dl/extractor/gdcvault.py +++ b/youtube_dl/extractor/gdcvault.py @@ -7,7 +7,10 @@ from ..compat import ( compat_urllib_parse, compat_urllib_request, ) -from ..utils import remove_end +from ..utils import ( + remove_end, + HEADRequest, +) class GDCVaultIE(InfoExtractor): @@ -73,10 +76,20 @@ class GDCVaultIE(InfoExtractor): return video_formats def _parse_flv(self, xml_description): - video_formats = [] + formats = [] akamai_url = xml_description.find('./metadata/akamaiHost').text + audios = xml_description.find('./metadata/audios') + if audios is not None: + for audio in audios: + formats.append({ + 'url': 'rtmp://%s/ondemand?ovpfv=1.1' % akamai_url, + 'play_path': remove_end(audio.get('url'), '.flv'), + 'ext': 'flv', + 'vcodec': 'none', + 'format_id': audio.get('code'), + }) slide_video_path = xml_description.find('./metadata/slideVideo').text - video_formats.append({ + formats.append({ 'url': 'rtmp://%s/ondemand?ovpfv=1.1' % akamai_url, 'play_path': remove_end(slide_video_path, '.flv'), 'ext': 'flv', @@ -86,7 +99,7 @@ class GDCVaultIE(InfoExtractor): 'format_id': 'slides', }) speaker_video_path = xml_description.find('./metadata/speakerVideo').text - video_formats.append({ + formats.append({ 'url': 'rtmp://%s/ondemand?ovpfv=1.1' % akamai_url, 'play_path': remove_end(speaker_video_path, '.flv'), 'ext': 'flv', @@ -95,7 +108,7 @@ class GDCVaultIE(InfoExtractor): 'preference': -1, 'format_id': 'speaker', }) - return video_formats + return formats def _login(self, webpage_url, display_id): (username, password) = self._get_login_info() @@ -133,16 +146,18 @@ class GDCVaultIE(InfoExtractor): r's1\.addVariable\("file",\s*encodeURIComponent\("(/[^"]+)"\)\);', start_page, 'url', default=None) if direct_url: - video_url = 'http://www.gdcvault.com/' + direct_url title = self._html_search_regex( r'Session Name\s*(.*?)', start_page, 'title') + video_url = 'http://www.gdcvault.com' + direct_url + # resolve the url so that we can detect the correct extension + head = self._request_webpage(HEADRequest(video_url), video_id) + video_url = head.geturl() return { 'id': video_id, 'display_id': display_id, 'url': video_url, - 'ext': 'flv', 'title': title, } @@ -168,8 +183,8 @@ class GDCVaultIE(InfoExtractor): # Fallback to the older format xml_name = self._html_search_regex(r'