Fix: make XML compatible (#33)

MetaTube.bundle/Contents/Code/api_client.py

@@ -1,7 +1,7 @@
 # -*- coding: utf-8 -*-
 
+import utils
 from constants import *
-from utils import parse_date
 
 try:  # Python 2
     import httplib as http_status
@@ -48,8 +48,8 @@ def __init__(self, **data):
         self.blood_type = data['blood_type']  # type: str
         self.measurements = data['measurements']  # type: str
         self.nationality = data['nationality']  # type: str
-        self.birthday = parse_date(data['birthday'])  # type: datetime
-        self.debut_date = parse_date(data['debut_date'])  # type: datetime
+        self.birthday = utils.parse_date(data['birthday'])  # type: datetime
+        self.debut_date = utils.parse_date(data['debut_date'])  # type: datetime
 
 
 class MovieSearchResult(BaseInfoObject):
@@ -61,7 +61,7 @@ def __init__(self, **data):
         self.thumb_url = data['thumb_url']  # type: str
         self.score = float(data['score'])  # type: float
         self.actors = data.get('actors', [])  # type: list[str]
-        self.release_date = parse_date(data['release_date'])  # type: datetime
+        self.release_date = utils.parse_date(data['release_date'])  # type: datetime
 
 
 class MovieInfoObject(MovieSearchResult):
@@ -87,7 +87,7 @@ def __init__(self, **data):
         self.author = data['author']  # type: str
         self.comment = data['comment']  # type: str
         self.score = float(data['score'])  # type: float
-        self.date = parse_date(data['date'])  # type: datetime
+        self.date = utils.parse_date(data['date'])  # type: datetime
 
 
 class TranslationInfoObject(object):
@@ -137,7 +137,7 @@ def get_json(self, url, require_auth=False):
             headers['Authorization'] = 'Bearer {token}'.format(token=Prefs[KEY_API_TOKEN])
 
         with self.session.get(url=url, headers=headers) as r:
-            info = r.json()
+            info = utils.safe_unicode(r.json())
             data = info.get('data')
             error = info.get('error')
 

MetaTube.bundle/Contents/Code/utils.py

@@ -1,6 +1,7 @@
 # -*- coding: utf-8 -*-
 import os
 import re
+import sys
 from base64 import b64decode
 from datetime import datetime
 
@@ -11,6 +12,34 @@
 except ImportError:  # Python 3
     from urllib.parse import unquote
 
+# Python 3 compatible code
+if sys.version_info.major == 3:
+    unicode = str
+
+
+# Based on an answer by John Machin on Stack Overflow:
+# - http://stackoverflow.com/questions/8733233/filtering-out-certain-bytes-in-python
+def filter_invalid_xml_chars(s):
+    def is_valid_xml_char(i):
+        c = ord(i)
+        return (0x20 <= c <= 0xD7FF or
+                0xE000 <= c <= 0xFFFD or
+                0x10000 <= c <= 0x10FFFF or
+                c in (0x9, 0xA, 0xD))
+
+    return u''.join(c for c in s if is_valid_xml_char(c))
+
+
+def safe_unicode(o):
+    if isinstance(o, unicode):
+        return filter_invalid_xml_chars(o)
+    elif isinstance(o, list):
+        return [safe_unicode(v) for v in o]
+    elif isinstance(o, dict):
+        return dict((k, safe_unicode(v)) for k, v in o.items())
+    else:
+        return o
+
 
 def average(a):
     x = 0.0