Merge pull request #9358 from dstftw/hls-native-to-ffmpeg-delegation

[youtube-dl] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 5c4ab2748cc94c9e3c10d29e994546d67ada4aea..6e45737843fbc2594227ac4699a3fce549b81e32 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -14,8 +14,8 @@ import email.utils
 import errno
 import functools
 import gzip
-import itertools
 import io
+import itertools
 import json
 import locale
 import math
@@ -24,8 +24,8 @@ import os
 import pipes
 import platform
 import re
-import ssl
 import socket
+import ssl
 import struct
 import subprocess
 import sys
@@ -89,6 +89,11 @@ KNOWN_EXTENSIONS = (
     'wav',
     'f4f', 'f4m', 'm3u8', 'smil')
 
+# needed for sanitizing filenames in restricted mode
+ACCENT_CHARS = dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿ',
+                        itertools.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOUUUUYP', ['ss'],
+                                        'aaaaaa', ['ae'], 'ceeeeiiiionoooooouuuuypy')))
+
 
 def preferredencoding():
     """Get preferred encoding.
@@ -251,9 +256,9 @@ def get_element_by_attribute(attribute, value, html):
 
     m = re.search(r'''(?xs)
         <([a-zA-Z0-9:._-]+)
-         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
+         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'))*?
          \s+%s=['"]?%s['"]?
-         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
+         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'))*?
         \s*>
         (?P<content>.*?)
         </\1>
@@ -365,6 +370,8 @@ def sanitize_filename(s, restricted=False, is_id=False):
     Set is_id if this is not an arbitrary string, but an ID that should be kept if possible
     """
     def replace_insane(char):
+        if restricted and char in ACCENT_CHARS:
+            return ACCENT_CHARS[char]
         if char == '?' or ord(char) < 32 or ord(char) == 127:
             return ''
         elif char == '"':
@@ -1540,44 +1547,46 @@ def parse_duration(s):
 
     s = s.strip()
 
-    m = re.match(
-        r'''(?ix)(?:P?T)?
-        (?:
-            (?P<only_mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*|
-            (?P<only_hours>[0-9.]+)\s*(?:hours?)|
-
-            \s*(?P<hours_reversed>[0-9]+)\s*(?:[:h]|hours?)\s*(?P<mins_reversed>[0-9]+)\s*(?:[:m]|mins?\.?|minutes?)\s*|
-            (?:
+    days, hours, mins, secs, ms = [None] * 5
+    m = re.match(r'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?$', s)
+    if m:
+        days, hours, mins, secs, ms = m.groups()
+    else:
+        m = re.match(
+            r'''(?ix)(?:P?T)?
                 (?:
-                    (?:(?P<days>[0-9]+)\s*(?:[:d]|days?)\s*)?
-                    (?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*
+                    (?P<days>[0-9]+)\s*d(?:ays?)?\s*
                 )?
-                (?P<mins>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*
-            )?
-            (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)?
-        )$''', s)
-    if not m:
-        return None
-    res = 0
-    if m.group('only_mins'):
-        return float_or_none(m.group('only_mins'), invscale=60)
-    if m.group('only_hours'):
-        return float_or_none(m.group('only_hours'), invscale=60 * 60)
-    if m.group('secs'):
-        res += int(m.group('secs'))
-    if m.group('mins_reversed'):
-        res += int(m.group('mins_reversed')) * 60
-    if m.group('mins'):
-        res += int(m.group('mins')) * 60
-    if m.group('hours'):
-        res += int(m.group('hours')) * 60 * 60
-    if m.group('hours_reversed'):
-        res += int(m.group('hours_reversed')) * 60 * 60
-    if m.group('days'):
-        res += int(m.group('days')) * 24 * 60 * 60
-    if m.group('ms'):
-        res += float(m.group('ms'))
-    return res
+                (?:
+                    (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
+                )?
+                (?:
+                    (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
+                )?
+                (?:
+                    (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
+                )?$''', s)
+        if m:
+            days, hours, mins, secs, ms = m.groups()
+        else:
+            m = re.match(r'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)$', s)
+            if m:
+                hours, mins = m.groups()
+            else:
+                return None
+
+    duration = 0
+    if secs:
+        duration += float(secs)
+    if mins:
+        duration += float(mins) * 60
+    if hours:
+        duration += float(hours) * 60 * 60
+    if days:
+        duration += float(days) * 24 * 60 * 60
+    if ms:
+        duration += float(ms)
+    return duration
 
 
 def prepend_extension(filename, ext, expected_real_ext=None):
@@ -1792,6 +1801,8 @@ def urlencode_postdata(*args, **kargs):
 
 
 def update_url_query(url, query):
+    if not query:
+        return url
     parsed_url = compat_urlparse.urlparse(url)
     qs = compat_parse_qs(parsed_url.query)
     qs.update(query)
@@ -1931,6 +1942,9 @@ def error_to_compat_str(err):
 
 
 def mimetype2ext(mt):
+    if mt is None:
+        return None
+
     ext = {
         'audio/mp4': 'm4a',
     }.get(mt)
@@ -2131,6 +2145,7 @@ def dfxp2srt(dfxp_data):
     _x = functools.partial(xpath_with_ns, ns_map={
         'ttml': 'http://www.w3.org/ns/ttml',
         'ttaf1': 'http://www.w3.org/2006/10/ttaf1',
+        'ttaf1_0604': 'http://www.w3.org/2006/04/ttaf1',
     })
 
     class TTMLPElementParser(object):
@@ -2157,7 +2172,7 @@ def dfxp2srt(dfxp_data):
 
     dfxp = compat_etree_fromstring(dfxp_data.encode('utf-8'))
     out = []
-    paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall('.//p')
+    paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall(_x('.//ttaf1_0604:p')) or dfxp.findall('.//p')
 
     if not paras:
         raise ValueError('Invalid dfxp/TTML subtitle')