[compat] compat_etree_fromstring: also decode the text attribute

[youtube-dl] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 1dc3153fd901e8fef6241e747f863a52a221e12d..c761ea22a4580216bb2484bd08ec7e8de1d724e4 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -3,6 +3,7 @@
 
 from __future__ import unicode_literals
 
+import base64
 import calendar
 import codecs
 import contextlib
@@ -35,6 +36,7 @@ import zlib
 from .compat import (
     compat_basestring,
     compat_chr,
+    compat_etree_fromstring,
     compat_html_entities,
     compat_http_client,
     compat_kwargs,
@@ -1371,7 +1373,12 @@ def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
             v = getattr(v, get_attr, None)
     if v == '':
         v = None
-    return default if v is None else (int(v) * invscale // scale)
+    if v is None:
+        return default
+    try:
+        return int(v) * invscale // scale
+    except ValueError:
+        return default
 
 
 def str_or_none(v, default=None):
@@ -1387,7 +1394,12 @@ def str_to_int(int_str):
 
 
 def float_or_none(v, scale=1, invscale=1, default=None):
-    return default if v is None else (float(v) * invscale / scale)
+    if v is None:
+        return default
+    try:
+        return float(v) * invscale / scale
+    except ValueError:
+        return default
 
 
 def parse_duration(s):
@@ -1640,29 +1652,6 @@ def encode_dict(d, encoding='utf-8'):
     return dict((k.encode(encoding), v.encode(encoding)) for k, v in d.items())
 
 
-try:
-    etree_iter = xml.etree.ElementTree.Element.iter
-except AttributeError:  # Python <=2.6
-    etree_iter = lambda n: n.findall('.//*')
-
-
-def parse_xml(s):
-    class TreeBuilder(xml.etree.ElementTree.TreeBuilder):
-        def doctype(self, name, pubid, system):
-            pass  # Ignore doctypes
-
-    parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder())
-    kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {}
-    tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
-    # Fix up XML parser in Python 2.x
-    if sys.version_info < (3, 0):
-        for n in etree_iter(tree):
-            if n.text is not None:
-                if not isinstance(n.text, compat_str):
-                    n.text = n.text.decode('utf-8')
-    return tree
-
-
 US_RATINGS = {
     'G': 0,
     'PG': 10,
@@ -1690,8 +1679,8 @@ def js_to_json(code):
         if v in ('true', 'false', 'null'):
             return v
         if v.startswith('"'):
-            return v
-        if v.startswith("'"):
+            v = re.sub(r"\\'", "'", v[1:-1])
+        elif v.startswith("'"):
             v = v[1:-1]
             v = re.sub(r"\\\\|\\'|\"", lambda m: {
                 '\\\\': '\\\\',
@@ -1785,6 +1774,10 @@ def urlhandle_detect_ext(url_handle):
     return mimetype2ext(getheader('Content-Type'))
 
 
+def encode_data_uri(data, mime_type):
+    return 'data:%s;base64,%s' % (mime_type, base64.b64encode(data).decode('ascii'))
+
+
 def age_restricted(content_limit, age_limit):
     """ Returns True iff the content should be blocked """
 
@@ -1959,7 +1952,7 @@ def dfxp2srt(dfxp_data):
 
         return out
 
-    dfxp = xml.etree.ElementTree.fromstring(dfxp_data.encode('utf-8'))
+    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')