From 555659cdc32c1a52344b7b00a2f3cace8ce3e68a Mon Sep 17 00:00:00 2001
From: Waylan Limberg <waylan.limberg@icloud.com>
Date: Mon, 5 Jun 2017 19:49:59 -0400
Subject: [PATCH] Open files with `utf-8-sig` to account for BOM.

Python simply discards the BOM with `utf-8-sig`. This way, users of
Microsoft text editors can have their files properly parsered. In all other
ways behaves as reading files using `utf-8` encoding. For more info see:
https://docs.python.org/2/library/codecs.html#encodings-and-unicode

Fixes #1186.
---
 mkdocs/nav.py               |  2 +-
 mkdocs/tests/build_tests.py | 30 ++++++++++++++++++++++++++++++
 2 files changed, 31 insertions(+), 1 deletion(-)

diff --git a/mkdocs/nav.py b/mkdocs/nav.py
index b73879d5..7240ef22 100644
--- a/mkdocs/nav.py
+++ b/mkdocs/nav.py
@@ -248,7 +248,7 @@ class Page(object):
             'page_read_source', None, config=config, page=self)
         if source is None:
             try:
-                with io.open(self.abs_input_path, 'r', encoding='utf-8') as f:
+                with io.open(self.abs_input_path, 'r', encoding='utf-8-sig') as f:
                     source = f.read()
             except IOError:
                 log.error('File not found: %s', self.abs_input_path)
diff --git a/mkdocs/tests/build_tests.py b/mkdocs/tests/build_tests.py
index 4820e92d..ac5f5116 100644
--- a/mkdocs/tests/build_tests.py
+++ b/mkdocs/tests/build_tests.py
@@ -7,6 +7,7 @@ import shutil
 import tempfile
 import unittest
 import mock
+import io
 
 try:
     from itertools import izip as zip
@@ -464,3 +465,32 @@ class BuildTests(unittest.TestCase):
         context = build.get_context(mock.Mock(), cfg)
 
         self.assertEqual(context['config']['extra']['a'], 1)
+
+    def test_BOM(self):
+        docs_dir = tempfile.mkdtemp()
+        site_dir = tempfile.mkdtemp()
+        try:
+            # Create an UTF-8 Encoded file with BOM (as Micorsoft editors do). See #1186.
+            f = io.open(os.path.join(docs_dir, 'index.md'), 'w', encoding='utf-8-sig')
+            f.write('# An UTF-8 encoded file with a BOM')
+            f.close()
+
+            cfg = load_config(
+                docs_dir=docs_dir,
+                site_dir=site_dir
+            )
+            build.build(cfg)
+
+            # Verify that the file was generated properly.
+            # If the BOM is not removed, Markdown will return:
+            # `<p>\ufeff# An UTF-8 encoded file with a BOM</p>`.
+            f = io.open(os.path.join(site_dir, 'index.html'), 'r', encoding='utf-8')
+            output = f.read()
+            f.close()
+            self.assertTrue(
+                '<h1 id="an-utf-8-encoded-file-with-a-bom">An UTF-8 encoded file with a BOM</h1>' in output
+            )
+
+        finally:
+            shutil.rmtree(docs_dir)
+            shutil.rmtree(site_dir)