From 042b45f6fdf204a1801fafb259ffe25ba0771be7 Mon Sep 17 00:00:00 2001
From: Jere Malinen <jere.malinen@arcusys.fi>
Date: Fri, 10 Oct 2014 12:03:16 +0300
Subject: [PATCH] Fix reading UTF-8 files which have BOM

https://github.com/django-compressor/django-compressor/issues/567
---
 compressor/base.py                             | 3 +++
 compressor/tests/static/css/utf-8_with-BOM.css | 1 +
 compressor/tests/test_base.py                  | 8 ++++++++
 3 files changed, 12 insertions(+)
 create mode 100644 compressor/tests/static/css/utf-8_with-BOM.css

diff --git a/compressor/base.py b/compressor/base.py
index de9c9ce..ae3a2c4 100644
--- a/compressor/base.py
+++ b/compressor/base.py
@@ -140,6 +140,9 @@ class Compressor(object):
         """
         Reads file contents using given `charset` and returns it as text.
         """
+        if charset == 'utf-8':
+            # Removes BOM
+            charset = 'utf-8-sig'
         with codecs.open(filename, 'r', charset) as fd:
             try:
                 return fd.read()
diff --git a/compressor/tests/static/css/utf-8_with-BOM.css b/compressor/tests/static/css/utf-8_with-BOM.css
new file mode 100644
index 0000000..06a6209
--- /dev/null
+++ b/compressor/tests/static/css/utf-8_with-BOM.css
@@ -0,0 +1 @@
+.compress-test {color: red;}
\ No newline at end of file
diff --git a/compressor/tests/test_base.py b/compressor/tests/test_base.py
index c6efe5c..d4de9d1 100644
--- a/compressor/tests/test_base.py
+++ b/compressor/tests/test_base.py
@@ -112,6 +112,14 @@ class CompressorTestCase(SimpleTestCase):
         hunks = '\n'.join([h for h in self.css_node.hunks()])
         self.assertEqual(out, hunks)
 
+    def test_css_output_with_bom_input(self):
+        out = 'body { background:#990; }\n.compress-test {color: red;}'
+        css = ("""<link rel="stylesheet" href="/static/css/one.css" type="text/css" />
+        <link rel="stylesheet" href="/static/css/utf-8_with-BOM.css" type="text/css" />""")
+        css_node_with_bom = CssCompressor(css)
+        hunks = '\n'.join([h for h in css_node_with_bom.hunks()])
+        self.assertEqual(out, hunks)
+
     def test_css_mtimes(self):
         is_date = re.compile(r'^\d{10}[\.\d]+$')
         for date in self.css_node.mtimes: