Merged in rogerjhu/coverage.py (pull request #19)

Make UTF-8 detection more robust.

2 files changed, 16 insertions, 2 deletions

diff --git a/coverage/phystokens.py b/coverage/phystokens.py
index 166020e1..df569fc0 100644
--- a/coverage/phystokens.py
+++ b/coverage/phystokens.py
@@ -188,7 +188,7 @@ def source_encoding(source):
         bom_found = True
         first = first[3:]
         default = 'utf-8-sig'
-    if not first:
+    if first is None:
         return default
 
     encoding = find_cookie(first)
diff --git a/tests/test_phystokens.py b/tests/test_phystokens.py
index 6d38a0e5..5a9ddac6 100644
--- a/tests/test_phystokens.py
+++ b/tests/test_phystokens.py
@@ -2,7 +2,7 @@
 
 import os, re
 from tests.coveragetest import CoverageTest
-from coverage.phystokens import source_token_lines
+from coverage.phystokens import source_token_lines, source_encoding
 
 
 SIMPLE = """\
@@ -77,3 +77,17 @@ class PhysTokensTest(CoverageTest):
         self.check_file_tokenization(stress)
         stress = os.path.join(HERE, "stress_phystoken_dos.tok")
         self.check_file_tokenization(stress)
+
+    def test_source_encoding_detect_utf8(self):
+        source = """\
+# coding=utf-8
+"""
+        self.assertEqual(source_encoding(source), 'utf-8')
+
+    def test_source_encoding_second_line_detect_utf8(self):
+        """ Verifies that UTF-8 encoding will still be detected in spite of the newline."""
+        source = """\
+
+# coding=utf-8
+"""
+        self.assertEqual(source_encoding(source), 'utf-8')