diff options
author | Philipp Hagemeister <phihag@phihag.de> | 2013-09-22 00:35:03 +0200 |
---|---|---|
committer | Philipp Hagemeister <phihag@phihag.de> | 2013-09-22 00:35:03 +0200 |
commit | c4417ddb611e14b81fe56b6b32964c5802faf554 (patch) | |
tree | b48e51a91bd0c01e641bda652e6bcfa0ae39c238 | |
parent | 2f2ffea9cad7d30165a0171bf6e662bef2182ab4 (diff) |
[youtube] Add filesystem signature cache
-rw-r--r-- | youtube_dl/FileDownloader.py | 2 | ||||
-rw-r--r-- | youtube_dl/extractor/youtube.py | 35 |
2 files changed, 30 insertions, 7 deletions
diff --git a/youtube_dl/FileDownloader.py b/youtube_dl/FileDownloader.py index 0b5a5d77d..1eb71a80e 100644 --- a/youtube_dl/FileDownloader.py +++ b/youtube_dl/FileDownloader.py @@ -39,6 +39,8 @@ class FileDownloader(object): test: Download only first bytes to test the downloader. min_filesize: Skip files smaller than this size max_filesize: Skip files larger than this size + cachedir: Location of the cache files in the filesystem. + False to disable filesystem cache. """ params = None diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index 5c0ea2e43..63f59ae8f 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -4,8 +4,10 @@ import collections import itertools import io import json -import netrc +import operator +import os.path import re +import shutil import socket import string import struct @@ -422,13 +424,28 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): """Indicate the download will use the RTMP protocol.""" self.to_screen(u'RTMP download detected') - def _extract_signature_function(self, video_id, player_url): - id_m = re.match(r'.*-(?P<id>[a-zA-Z0-9]+)\.(?P<ext>[a-z]+)$', + def _extract_signature_function(self, video_id, player_url, slen): + id_m = re.match(r'.*-(?P<id>[a-zA-Z0-9_-]+)\.(?P<ext>[a-z]+)$', player_url) player_type = id_m.group('ext') player_id = id_m.group('id') - # TODO read from filesystem cache + # Read from filesystem cache + func_id = '%s_%s_%d' % (player_type, player_id, slen) + assert os.path.basename(func_id) == func_id + cache_dir = self.downloader.params.get('cachedir', + u'~/.youtube-dl/cache') + + if cache_dir is not False: + cache_fn = os.path.join(os.path.expanduser(cache_dir), + u'youtube-sigfuncs', + func_id + '.json') + try: + with io.open(cache_fn, '', encoding='utf-8') as cachef: + cache_spec = json.load(cachef) + return lambda s: u''.join(s[i] for i in cache_spec) + except OSError: + pass # No cache available if player_type == 'js': code = self._download_webpage( @@ -436,7 +453,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): note=u'Downloading %s player %s' % (player_type, player_id), errnote=u'Download of %s failed' % player_url) res = self._parse_sig_js(code) - elif player_tpye == 'swf': + elif player_type == 'swf': urlh = self._request_webpage( player_url, video_id, note=u'Downloading %s player %s' % (player_type, player_id), @@ -446,7 +463,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): else: assert False, 'Invalid player type %r' % player_type - # TODO write cache + if cache_dir is not False: + cache_res = res(map(compat_chr, range(slen))) + cache_spec = [ord(c) for c in cache_res] + shutil.makedirs(os.path.dirname(cache_fn)) + write_json_file(cache_spec, cache_fn) return res @@ -983,7 +1004,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): try: if player_url not in self._player_cache: func = self._extract_signature_function( - video_id, player_url + video_id, player_url, len(s) ) self._player_cache[player_url] = func return self._player_cache[player_url](s) |