Merge pull request #8739 from remitamine/update_url_params

[youtube-dl] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 16b4324a4d1d9c8203b83b9b15e22b793e2b2697..d431aa6b726c59b40a9c48b3e3144f1d7a2c8db0 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -160,8 +160,6 @@ if sys.version_info >= (2, 7):
     def find_xpath_attr(node, xpath, key, val=None):
         """ Find the xpath xpath[@key=val] """
         assert re.match(r'^[a-zA-Z_-]+$', key)
-        if val:
-            assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val)
         expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
         return node.find(expr)
 else:
@@ -467,6 +465,10 @@ def encodeFilename(s, for_subprocess=False):
     if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
         return s
 
+    # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
+    if sys.platform.startswith('java'):
+        return s
+
     return s.encode(get_subprocess_encoding(), 'ignore')
 
 
@@ -905,9 +907,9 @@ def unified_strdate(date_str, day_first=True):
         '%d %b %Y',
         '%B %d %Y',
         '%b %d %Y',
-        '%b %dst %Y %I:%M%p',
-        '%b %dnd %Y %I:%M%p',
-        '%b %dth %Y %I:%M%p',
+        '%b %dst %Y %I:%M',
+        '%b %dnd %Y %I:%M',
+        '%b %dth %Y %I:%M',
         '%Y %m %d',
         '%Y-%m-%d',
         '%Y/%m/%d',
@@ -1586,9 +1588,12 @@ class PagedList(object):
 
 
 class OnDemandPagedList(PagedList):
-    def __init__(self, pagefunc, pagesize):
+    def __init__(self, pagefunc, pagesize, use_cache=False):
         self._pagefunc = pagefunc
         self._pagesize = pagesize
+        self._use_cache = use_cache
+        if use_cache:
+            self._cache = {}
 
     def getslice(self, start=0, end=None):
         res = []
@@ -1598,7 +1603,13 @@ class OnDemandPagedList(PagedList):
             if start >= nextfirstid:
                 continue
 
-            page_results = list(self._pagefunc(pagenum))
+            page_results = None
+            if self._use_cache:
+                page_results = self._cache.get(pagenum)
+            if page_results is None:
+                page_results = list(self._pagefunc(pagenum))
+            if self._use_cache:
+                self._cache[pagenum] = page_results
 
             startv = (
                 start % self._pagesize
@@ -1728,6 +1739,14 @@ def urlencode_postdata(*args, **kargs):
     return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
 
 
+def update_url_query(url, query):
+    parsed_url = compat_urlparse.urlparse(url)
+    qs = compat_parse_qs(parsed_url.query)
+    qs.update(query)
+    return compat_urlparse.urlunparse(parsed_url._replace(
+        query=compat_urllib_parse.urlencode(qs, True)))
+
+
 def encode_dict(d, encoding='utf-8'):
     def encode(v):
         return v.encode(encoding) if isinstance(v, compat_basestring) else v
@@ -1862,7 +1881,11 @@ def mimetype2ext(mt):
 
     return {
         '3gpp': '3gp',
+        'smptett+xml': 'tt',
+        'srt': 'srt',
+        'ttaf+xml': 'dfxp',
         'ttml+xml': 'ttml',
+        'vtt': 'vtt',
         'x-flv': 'flv',
         'x-mp4-fragmented': 'mp4',
         'x-ms-wmv': 'wmv',
@@ -2622,3 +2645,41 @@ def ohdave_rsa_encrypt(data, exponent, modulus):
     payload = int(binascii.hexlify(data[::-1]), 16)
     encrypted = pow(payload, exponent, modulus)
     return '%x' % encrypted
+
+
+def encode_base_n(num, n, table=None):
+    FULL_TABLE = '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
+    if not table:
+        table = FULL_TABLE[:n]
+
+    if n > len(table):
+        raise ValueError('base %d exceeds table length %d' % (n, len(table)))
+
+    if num == 0:
+        return table[0]
+
+    ret = ''
+    while num:
+        ret = table[num % n] + ret
+        num = num // n
+    return ret
+
+
+def decode_packed_codes(code):
+    mobj = re.search(
+        r"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)",
+        code)
+    obfucasted_code, base, count, symbols = mobj.groups()
+    base = int(base)
+    count = int(count)
+    symbols = symbols.split('|')
+    symbol_table = {}
+
+    while count:
+        count -= 1
+        base_n_count = encode_base_n(count, base)
+        symbol_table[base_n_count] = symbols[count] or base_n_count
+
+    return re.sub(
+        r'\b(\w+)\b', lambda mobj: symbol_table[mobj.group(0)],
+        obfucasted_code)