strip utf-8 bom.

3 files changed, 22 insertions, 1 deletions

diff --git a/simplejson/__init__.py b/simplejson/__init__.py
index a5c0137..bc5c93a 100644
--- a/simplejson/__init__.py
+++ b/simplejson/__init__.py
@@ -437,7 +437,16 @@ def load(fp, encoding=None, cls=None, object_hook=None, parse_float=None,
     of subclassing whenever possible.
 
     """
-    return loads(fp.read(),
+    # Strip the UTF-8 BOM
+    contents = fp.read()
+    ord0 = ord(contents[0])
+    if ord0 in (0xef, 0xfeff):
+        if ord0 == 0xfeff:
+            contents = contents[1:]
+        elif contents[:3] == '\xef\xbb\xbf':
+            contents = contents[3:]
+
+    return loads(contents,
         encoding=encoding, cls=cls, object_hook=object_hook,
         parse_float=parse_float, parse_int=parse_int,
         parse_constant=parse_constant, object_pairs_hook=object_pairs_hook,
diff --git a/simplejson/tests/test_unicode.py b/simplejson/tests/test_unicode.py
index f04cc5c..60492f7 100644
--- a/simplejson/tests/test_unicode.py
+++ b/simplejson/tests/test_unicode.py
@@ -1,4 +1,5 @@
 import sys
+import os.path
 from unittest import TestCase
 
 import simplejson as json
@@ -143,3 +144,11 @@ class TestUnicode(TestCase):
         self.assertEqual(
             json.dumps(c, ensure_ascii=False),
             '"' + c + '"')
+
+    def test_strip_bom(self):
+        thisdir = os.path.dirname(__file__)
+        json_file = os.path.join(thisdir, "utf-8-bom.json")
+        doc_ascii = {
+            u"content": u"\u3053\u3093\u306b\u3061\u308f"
+        }
+        self.assertEqual(json.load(open(json_file)), doc_ascii)
diff --git a/simplejson/tests/utf-8-bom.json b/simplejson/tests/utf-8-bom.json
new file mode 100644
index 0000000..1791beb
--- /dev/null
+++ b/simplejson/tests/utf-8-bom.json
@@ -0,0 +1,3 @@
+{
+    "content": "こんにちわ"
+}