From 9f846d36870c5d9bd97355ea2f8b927cd6a2d514 Mon Sep 17 00:00:00 2001
From: Cosimo Lupo <cosimo.lupo@daltonmaag.com>
Date: Mon, 27 Jul 2015 19:24:02 +0100
Subject: [PATCH] [xmlReader_test] test newlines gets normalised by expat
 parser, unless escaped

---
 Lib/fontTools/misc/xmlReader_test.py | 30 ++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)

diff --git a/Lib/fontTools/misc/xmlReader_test.py b/Lib/fontTools/misc/xmlReader_test.py
index 2081ba099..e468f8e70 100644
--- a/Lib/fontTools/misc/xmlReader_test.py
+++ b/Lib/fontTools/misc/xmlReader_test.py
@@ -46,6 +46,36 @@ class TestXMLReader(unittest.TestCase):
 		content = strjoin(reader.contents[0]).strip() 
 		self.assertEqual(expected, content)
 
+	def test_normalise_newlines(self):
+
+		class DebugXMLReader(XMLReader):
+
+			def __init__(self, fileName, ttFont, progress=None, quiet=False):
+				super(DebugXMLReader, self).__init__(
+					fileName, ttFont, progress, quiet)
+				self.newlines = []
+
+			def _characterDataHandler(self, data):
+				self.newlines.extend([c for c in data if c in ('\r', '\n')])
+
+		# notice how when CR is escaped, it is not normalised by the XML parser
+		data = (
+			'<ttFont>\r'                                    #        \r -> \n
+			'  <test>\r\n'                                  #      \r\n -> \n 
+			'    a line of text\n'                          #              \n
+			'    escaped CR and unix newline &#13;\n'       #   &#13;\n -> \r\n
+			'    escaped CR and macintosh newline &#13;\r'  #   &#13;\r -> \r\n
+			'    escaped CR and windows newline &#13;\r\n'  # &#13;\r\n -> \r\n
+			'  </test>\n'                                   #              \n
+			'</ttFont>')
+		with tempfile.NamedTemporaryFile(delete=False) as tmp:
+			tmp.write(data.encode('utf-8'))
+		reader = DebugXMLReader(tmp.name, TTFont(), quiet=True)
+		reader.read()
+		os.remove(tmp.name)
+		expected = ['\n'] * 3 + ['\r', '\n'] * 3 + ['\n']
+		self.assertEqual(expected, reader.newlines)
+
 
 if __name__ == '__main__':
 	unittest.main()