diff --git a/morss/crawler.py b/morss/crawler.py
index aa52fc8..fd471ab 100644
--- a/morss/crawler.py
+++ b/morss/crawler.py
@@ -2,14 +2,13 @@ import ssl
 import socket
 
 from gzip import GzipFile
+from io import BytesIO
 
 try:
-    from StringIO import StringIO
     from urllib2 import URLError
     from urllib2 import HTTPSHandler, BaseHandler, AbstractHTTPHandler, Request, addinfourl
     from httplib import HTTPException, HTTPConnection, HTTPS_PORT
 except ImportError:
-    from io import StringIO
     from urllib.error import URLError
     from urllib.request import HTTPSHandler, BaseHandler, AbstractHTTPHandler, Request, addinfourl
     from http.client import HTTPException, HTTPConnection, HTTPS_PORT
@@ -118,9 +117,9 @@ class GZIPHandler(BaseHandler):
         if 200 <= resp.code < 300:
             if resp.headers.get('Content-Encoding') == 'gzip':
                 data = resp.read()
-                data = GzipFile(fileobj=StringIO(data), mode='r').read()
+                data = GzipFile(fileobj=BytesIO(data), mode='r').read()
 
-                fp = StringIO(data)
+                fp = BytesIO(data)
                 old_resp = resp
                 resp = addinfourl(fp, old_resp.headers, old_resp.url, old_resp.code)
                 resp.msg = old_resp.msg
@@ -135,13 +134,13 @@ def detect_encoding(data, con=None):
     if con is not None and con.info().get('charset'):
         return con.info().get('charset')
 
-    match = re.search('charset=["\']?([0-9a-zA-Z-]+)', data[:1000])
+    match = re.search(b'charset=["\']?([0-9a-zA-Z-]+)', data[:1000])
     if match:
-        return match.groups()[0]
+        return match.groups()[0].lower().decode()
 
-    match = re.search('encoding=["\']?([0-9a-zA-Z-]+)', data[:100])
+    match = re.search(b'encoding=["\']?([0-9a-zA-Z-]+)', data[:100])
     if match:
-        return match.groups()[0].lower()
+        return match.groups()[0].lower().decode()
 
     return None
 
@@ -157,7 +156,7 @@ class EncodingFixHandler(BaseHandler):
                 data = data.decode(enc, 'replace')
                 data = data.encode(enc)
 
-            fp = StringIO(data)
+            fp = BytesIO(data)
             old_resp = resp
             resp = addinfourl(fp, old_resp.headers, old_resp.url, old_resp.code)
             resp.msg = old_resp.msg
@@ -231,7 +230,7 @@ class MetaRedirectHandler(BaseHandler):
         if 200 <= resp.code < 300 and contenttype.startswith('text/'):
             if contenttype in MIMETYPE['html']:
                 data = resp.read()
-                match = re.search(r'(?i)<meta http-equiv=.refresh[^>]*?url=(http.*?)["\']', data)
+                match = re.search(b'(?i)<meta http-equiv=.refresh[^>]*?url=(http.*?)["\']', data)
                 if match:
                     new_url = match.groups()[0]
                     new_headers = dict((k, v) for k, v in list(req.headers.items())
@@ -243,7 +242,7 @@ class MetaRedirectHandler(BaseHandler):
 
                     return self.parent.open(new, timeout=req.timeout)
                 else:
-                    fp = StringIO(data)
+                    fp = BytesIO(data)
                     old_resp = resp
                     resp = addinfourl(fp, old_resp.headers, old_resp.url, old_resp.code)
                     resp.msg = old_resp.msg
@@ -273,7 +272,7 @@ class EtagHandler(BaseHandler):
             headers.addheader('etag', self.etag)
         if self.lastmodified:
             headers.addheader('last-modified', self.lastmodified)
-        resp = addinfourl(StringIO(self.cache), headers, req.get_full_url(), 200)
+        resp = addinfourl(BytesIO(self.cache), headers, req.get_full_url(), 200)
         return resp
 
     https_request = http_request
diff --git a/morss/feeds.py b/morss/feeds.py
index 9cfcca3..49ed038 100644
--- a/morss/feeds.py
+++ b/morss/feeds.py
@@ -88,9 +88,9 @@ class FeedException(Exception):
 
 def parse(data):
     # encoding
-    match = re.search('encoding=["\']?([0-9a-zA-Z-]+)', data[:100])
+    match = re.search(b'encoding=["\']?([0-9a-zA-Z-]+)', data[:100])
     if match:
-        enc = match.groups()[0].lower()
+        enc = match.groups()[0].lower().decode()
         if isinstance(data, bytes):
             data = data.decode(enc, 'ignore')
         data = data.encode(enc)
diff --git a/morss/morss.py b/morss/morss.py
index 6e0590f..e2bacf4 100644
--- a/morss/morss.py
+++ b/morss/morss.py
@@ -182,6 +182,9 @@ class Cache:
             return None
 
     def set(self, key, content):
+        if sys.version > '3' and isinstance(content, bytes):
+            content = content.decode('utf-8')
+
         self._cache[key] = {'last': time.time(), 'value': content}
 
     __getitem__ = get
@@ -383,8 +386,7 @@ def Fill(item, cache, options, feedurl='/', fast=False):
 
     # download
     try:
-        url = link.encode('utf-8')
-        con = build_opener(*accept_handler(('html', 'text/*'), True)).open(url, timeout=TIMEOUT)
+        con = build_opener(*accept_handler(('html', 'text/*'), True)).open(link, timeout=TIMEOUT)
         data = con.read()
     except (IOError, HTTPException) as e:
         log('http error:  %s' % e.message)
@@ -423,6 +425,9 @@ def Init(url, cache_path, options):
 
     url = url.replace(' ', '%20')
 
+    if isinstance(url, bytes):
+        url = url.decode()
+
     # cache
     cache = Cache(cache_path, url)
     log(cache._hash)
@@ -464,7 +469,7 @@ def Fetch(url, cache, options):
 
         if url.startswith('https://itunes.apple.com/lookup?id='):
             style = 'itunes'
-        elif xml.startswith('<?xml') or contenttype in MIMETYPE['xml']:
+        elif xml.startswith(b'<?xml') or contenttype in MIMETYPE['xml']:
             style = 'normal'
         elif feedify.supported(url):
             style = 'feedify'