py3: decode bytes for json.loads()

In py 3.5 json.loads requires to have string as input, all bytes must be
decoded.

Note: python 3.6 supports bytes for json.loads()

https://fedorahosted.org/freeipa/ticket/4985

Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>

1 files changed, 50 insertions, 0 deletions

diff --git a/ipapython/ipautil.py b/ipapython/ipautil.py
index f2b3d7419..c8f87ef5b 100644
--- a/ipapython/ipautil.py
+++ b/ipapython/ipautil.py
@@ -19,6 +19,7 @@
 
 from __future__ import print_function
 
+import codecs
 import string
 import tempfile
 import subprocess
@@ -1361,6 +1362,55 @@ def escape_seq(seq, *args):
     return tuple(a.replace(seq, u'\\{}'.format(seq)) for a in args)
 
 
+def decode_json(data):
+    """Decode JSON bytes to string with proper encoding
+
+    Only for supporting Py 3.5
+
+    Py 3.6 supports bytes as parameter for json.load, we can drop this when
+    there is no need for python 3.5 anymore
+
+    Code from:
+        https://bugs.python.org/file43513/json_detect_encoding_3.patch
+
+    :param data: JSON bytes
+    :return: return JSON string
+    """
+
+    def detect_encoding(b):
+        bstartswith = b.startswith
+        if bstartswith((codecs.BOM_UTF32_BE, codecs.BOM_UTF32_LE)):
+            return 'utf-32'
+        if bstartswith((codecs.BOM_UTF16_BE, codecs.BOM_UTF16_LE)):
+            return 'utf-16'
+        if bstartswith(codecs.BOM_UTF8):
+            return 'utf-8-sig'
+
+        if len(b) >= 4:
+            if not b[0]:
+                # 00 00 -- -- - utf-32-be
+                # 00 XX -- -- - utf-16-be
+                return 'utf-16-be' if b[1] else 'utf-32-be'
+            if not b[1]:
+                # XX 00 00 00 - utf-32-le
+                # XX 00 XX XX - utf-16-le
+                return 'utf-16-le' if b[2] or b[3] else 'utf-32-le'
+        elif len(b) == 2:
+            if not b[0]:
+                # 00 XX - utf-16-be
+                return 'utf-16-be'
+            if not b[1]:
+                # XX 00 - utf-16-le
+                return 'utf-16-le'
+        # default
+        return 'utf-8'
+
+    if isinstance(data, six.text_type):
+        return data
+
+    return data.decode(detect_encoding(data), 'surrogatepass')
+
+
 class APIVersion(tuple):
     """API version parser and handler