[compat] Add compat_urllib_parse_urlencode and eliminate encode_dict

[youtube-dl] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index ec186918cd8672ada2da2d5521e0ba8b22eb273d..eacd81bf9ad3c733c578ace81cd7bf03ebff4725 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -47,9 +47,11 @@ from .compat import (
     compat_str,
     compat_urllib_error,
     compat_urllib_parse,
+    compat_urllib_parse_urlencode,
     compat_urllib_parse_urlparse,
     compat_urllib_request,
     compat_urlparse,
+    compat_xpath,
     shlex_quote,
 )
 
@@ -165,12 +167,7 @@ if sys.version_info >= (2, 7):
         return node.find(expr)
 else:
     def find_xpath_attr(node, xpath, key, val=None):
-        # Here comes the crazy part: In 2.6, if the xpath is a unicode,
-        # .//node does not match if a node is a direct child of . !
-        if isinstance(xpath, compat_str):
-            xpath = xpath.encode('ascii')
-
-        for f in node.findall(xpath):
+        for f in node.findall(compat_xpath(xpath)):
             if key not in f.attrib:
                 continue
             if val is None or f.attrib.get(key) == val:
@@ -195,9 +192,7 @@ def xpath_with_ns(path, ns_map):
 
 def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
     def _find_xpath(xpath):
-        if sys.version_info < (2, 7):  # Crazy 2.6
-            xpath = xpath.encode('ascii')
-        return node.find(xpath)
+        return node.find(compat_xpath(xpath))
 
     if isinstance(xpath, (str, compat_str)):
         n = _find_xpath(xpath)
@@ -273,15 +268,17 @@ def get_element_by_attribute(attribute, value, html):
 
     return unescapeHTML(res)
 
+
 class HTMLAttributeParser(compat_HTMLParser):
     """Trivial HTML parser to gather the attributes for a single element"""
     def __init__(self):
-        self.attrs = { }
+        self.attrs = {}
         compat_HTMLParser.__init__(self)
 
     def handle_starttag(self, tag, attrs):
         self.attrs = dict(attrs)
 
+
 def extract_attributes(html_element):
     """Given a string for an HTML element such as
     <el
@@ -303,6 +300,7 @@ def extract_attributes(html_element):
     parser.close()
     return parser.attrs
 
+
 def clean_html(html):
     """Clean an HTML snippet into a readable string"""
 
@@ -1318,7 +1316,7 @@ def shell_quote(args):
 def smuggle_url(url, data):
     """ Pass additional data in a URL for internal use. """
 
-    sdata = compat_urllib_parse.urlencode(
+    sdata = compat_urllib_parse_urlencode(
         {'__youtubedl_smuggle': json.dumps(data)})
     return url + '#' + sdata
 
@@ -1349,7 +1347,7 @@ def format_bytes(bytes):
 def lookup_unit_table(unit_table, s):
     units_re = '|'.join(re.escape(u) for u in unit_table)
     m = re.match(
-        r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)' % units_re, s)
+        r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re, s)
     if not m:
         return None
     num_str = m.group('num').replace(',', '.')
@@ -1749,6 +1747,7 @@ def escape_url(url):
     """Escape URL as suggested by RFC 3986"""
     url_parsed = compat_urllib_parse_urlparse(url)
     return url_parsed._replace(
+        netloc=url_parsed.netloc.encode('idna').decode('ascii'),
         path=escape_rfc3986(url_parsed.path),
         params=escape_rfc3986(url_parsed.params),
         query=escape_rfc3986(url_parsed.query),
@@ -1758,7 +1757,8 @@ def escape_url(url):
 try:
     struct.pack('!I', 0)
 except TypeError:
-    # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument
+    # In Python 2.6 and 2.7.x < 2.7.7, struct requires a bytes argument
+    # See https://bugs.python.org/issue19099
     def struct_pack(spec, *args):
         if isinstance(spec, compat_str):
             spec = spec.encode('ascii')
@@ -1790,22 +1790,15 @@ def read_batch_urls(batch_fd):
 
 
 def urlencode_postdata(*args, **kargs):
-    return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
+    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)
-    qs = encode_dict(qs)
     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
-    return dict((encode(k), encode(v)) for k, v in d.items())
+        query=compat_urllib_parse_urlencode(qs, True)))
 
 
 def dict_get(d, key_or_keys, default=None, skip_false_values=True):