diff options
| -rw-r--r-- | yt_dlp/downloader/common.py | 6 | ||||
| -rw-r--r-- | yt_dlp/downloader/http.py | 3 | ||||
| -rw-r--r-- | yt_dlp/extractor/fptplay.py | 37 | ||||
| -rw-r--r-- | yt_dlp/utils.py | 95 |
4 files changed, 86 insertions, 55 deletions
diff --git a/yt_dlp/downloader/common.py b/yt_dlp/downloader/common.py index afd2f2e38..cbfea7a65 100644 --- a/yt_dlp/downloader/common.py +++ b/yt_dlp/downloader/common.py @@ -11,6 +11,7 @@ from ..utils import ( encodeFilename, error_to_compat_str, format_bytes, + LockingUnsupportedError, sanitize_open, shell_quote, timeconvert, @@ -234,7 +235,10 @@ class FileDownloader(object): @wrap_file_access('open', fatal=True) def sanitize_open(self, filename, open_mode): - return sanitize_open(filename, open_mode) + f, filename = sanitize_open(filename, open_mode) + if not getattr(f, 'locked', None): + self.write_debug(f'{LockingUnsupportedError.msg}. Proceeding without locking', only_once=True) + return f, filename @wrap_file_access('remove') def try_remove(self, filename): diff --git a/yt_dlp/downloader/http.py b/yt_dlp/downloader/http.py index 591a9b08d..a232168fa 100644 --- a/yt_dlp/downloader/http.py +++ b/yt_dlp/downloader/http.py @@ -145,7 +145,8 @@ class HttpFD(FileDownloader): or content_len < range_end) if accept_content_len: ctx.content_len = content_len - ctx.data_len = min(content_len, req_end or content_len) - (req_start or 0) + if content_len or req_end: + ctx.data_len = min(content_len or req_end, req_end or content_len) - (req_start or 0) return # 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 diff --git a/yt_dlp/extractor/fptplay.py b/yt_dlp/extractor/fptplay.py index a34e90bb1..c23fe6c53 100644 --- a/yt_dlp/extractor/fptplay.py +++ b/yt_dlp/extractor/fptplay.py @@ -7,12 +7,14 @@ import urllib.parse from .common import InfoExtractor from ..utils import ( + clean_html, join_nonempty, + strip_or_none, ) class FptplayIE(InfoExtractor): - _VALID_URL = r'https?://fptplay\.vn/(?P<type>xem-video)/[^/]+\-(?P<id>\w+)(?:/tap-(?P<episode>[^/]+)?/?(?:[?#]|$)|)' + _VALID_URL = r'https?://fptplay\.vn/xem-video/[^/]+\-(?P<id>\w+)(?:/tap-(?P<episode>\d+)?/?(?:[?#]|$)|)' _GEO_COUNTRIES = ['VN'] IE_NAME = 'fptplay' IE_DESC = 'fptplay.vn' @@ -22,7 +24,7 @@ class FptplayIE(InfoExtractor): 'info_dict': { 'id': '621a123016f369ebbde55945', 'ext': 'mp4', - 'title': 'Nhân Duyên Đại Nhân Xin Dừng Bước - Ms. Cupid In Love', + 'title': 'Nhân Duyên Đại Nhân Xin Dừng Bước - Tập 1A', 'description': 'md5:23cf7d1ce0ade8e21e76ae482e6a8c6c', }, }, { @@ -31,25 +33,42 @@ class FptplayIE(InfoExtractor): 'info_dict': { 'id': '61f3aa8a6b3b1d2e73c60eb5', 'ext': 'mp4', - 'title': 'Má Tôi Là Đại Gia - 3', + 'title': 'Má Tôi Là Đại Gia - Tập 3', 'description': 'md5:ff8ba62fb6e98ef8875c42edff641d1c', }, }, { + 'url': 'https://fptplay.vn/xem-video/lap-toi-do-giam-under-the-skin-6222d9684ec7230fa6e627a2/tap-4', + 'md5': 'bcb06c55ec14786d7d4eda07fa1ccbb9', + 'info_dict': { + 'id': '6222d9684ec7230fa6e627a2', + 'ext': 'mp4', + 'title': 'Lạp Tội Đồ Giám - Tập 2B', + 'description': 'md5:e5a47e9d35fbf7e9479ca8a77204908b', + }, + }, { 'url': 'https://fptplay.vn/xem-video/nha-co-chuyen-hi-alls-well-ends-well-1997-6218995f6af792ee370459f0', 'only_matching': True, }] def _real_extract(self, url): - type_url, video_id, episode = self._match_valid_url(url).group('type', 'id', 'episode') - webpage = self._download_webpage(url, video_id=video_id, fatal=False) - info = self._download_json(self.get_api_with_st_token(video_id, episode or 0), video_id) + video_id, slug_episode = self._match_valid_url(url).group('id', 'episode') + webpage = self._download_webpage(url, video_id=video_id, fatal=False) or '' + title = self._search_regex( + r'(?s)<h4\s+class="mb-1 text-2xl text-white"[^>]*>(.+)</h4>', webpage, 'title', fatal=False) + real_episode = slug_episode if not title else self._search_regex( + r'<p.+title="(?P<episode>[^">]+)"\s+class="epi-title active"', webpage, 'episode', fatal=False) + title = strip_or_none(title) or self._html_search_meta(('og:title', 'twitter:title'), webpage) + + info = self._download_json( + self.get_api_with_st_token(video_id, int(slug_episode) - 1 if slug_episode else 0), video_id) formats, subtitles = self._extract_m3u8_formats_and_subtitles(info['data']['url'], video_id, 'mp4') self._sort_formats(formats) return { 'id': video_id, - 'title': join_nonempty( - self._html_search_meta(('og:title', 'twitter:title'), webpage), episode, delim=' - '), - 'description': self._html_search_meta(['og:description', 'twitter:description'], webpage), + 'title': join_nonempty(title, real_episode, delim=' - '), + 'description': ( + clean_html(self._search_regex(r'<p\s+class="overflow-hidden"[^>]*>(.+)</p>', webpage, 'description')) + or self._html_search_meta(('og:description', 'twitter:description'), webpage)), 'formats': formats, 'subtitles': subtitles, } diff --git a/yt_dlp/utils.py b/yt_dlp/utils.py index 451aa18e1..d696bf3ab 100644 --- a/yt_dlp/utils.py +++ b/yt_dlp/utils.py @@ -674,26 +674,29 @@ def sanitize_open(filename, open_mode): It returns the tuple (stream, definitive_file_name). """ - try: - if filename == '-': - if sys.platform == 'win32': - import msvcrt - msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY) - return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename) - stream = locked_file(filename, open_mode, block=False).open() - return (stream, filename) - except (IOError, OSError) as err: - if err.errno in (errno.EACCES,): - raise + if filename == '-': + if sys.platform == 'win32': + import msvcrt + msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY) + return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename) - # In case of error, try to remove win32 forbidden chars - alt_filename = sanitize_path(filename) - if alt_filename == filename: - raise - else: - # An exception here should be caught in the caller - stream = locked_file(filename, open_mode, block=False).open() - return (stream, alt_filename) + for attempt in range(2): + try: + try: + if sys.platform == 'win32': + # FIXME: Windows only has mandatory locking which also locks the file from being read. + # So for now, don't lock the file on windows. Ref: https://github.com/yt-dlp/yt-dlp/issues/3124 + raise LockingUnsupportedError() + stream = locked_file(filename, open_mode, block=False).__enter__() + except LockingUnsupportedError: + stream = open(filename, open_mode) + return (stream, filename) + except (IOError, OSError) as err: + if attempt or err.errno in (errno.EACCES,): + raise + old_filename, filename = filename, sanitize_path(filename) + if old_filename == filename: + raise def timeconvert(timestr): @@ -2120,6 +2123,13 @@ def intlist_to_bytes(xs): return compat_struct_pack('%dB' % len(xs), *xs) +class LockingUnsupportedError(IOError): + msg = 'File locking is not supported on this platform' + + def __init__(self): + super().__init__(self.msg) + + # Cross-platform file locking if sys.platform == 'win32': import ctypes.wintypes @@ -2200,21 +2210,20 @@ else: fcntl.lockf(f, fcntl.LOCK_UN) except ImportError: - UNSUPPORTED_MSG = 'file locking is not supported on this platform' def _lock_file(f, exclusive, block): - raise IOError(UNSUPPORTED_MSG) + raise LockingUnsupportedError() def _unlock_file(f): - raise IOError(UNSUPPORTED_MSG) + raise LockingUnsupportedError() class locked_file(object): - _closed = False + locked = False def __init__(self, filename, mode, block=True, encoding=None): - assert mode in ['r', 'rb', 'a', 'ab', 'w', 'wb'] - self.f = io.open(filename, mode, encoding=encoding) + assert mode in {'r', 'rb', 'a', 'ab', 'w', 'wb'} + self.f = open(filename, mode, encoding=encoding) self.mode = mode self.block = block @@ -2222,36 +2231,34 @@ class locked_file(object): exclusive = 'r' not in self.mode try: _lock_file(self.f, exclusive, self.block) + self.locked = True except IOError: self.f.close() raise return self - def __exit__(self, etype, value, traceback): + def unlock(self): + if not self.locked: + return try: - if not self._closed: - _unlock_file(self.f) + _unlock_file(self.f) finally: - self.f.close() - self._closed = True - - def __iter__(self): - return iter(self.f) - - def write(self, *args): - return self.f.write(*args) + self.locked = False - def read(self, *args): - return self.f.read(*args) + def __exit__(self, *_): + try: + self.unlock() + finally: + self.f.close() - def flush(self): - self.f.flush() + open = __enter__ + close = __exit__ - def open(self): - return self.__enter__() + def __getattr__(self, attr): + return getattr(self.f, attr) - def close(self, *args): - self.__exit__(self, *args, value=False, traceback=False) + def __iter__(self): + return iter(self.f) def get_filesystem_encoding(): |