From 57ed5537cf1351ff307e2fc706528240b089cb29 Mon Sep 17 00:00:00 2001
From: Xianjun Zhu <xianjun@us.ibm.com>
Date: Fri, 8 Feb 2019 23:24:38 -0500
Subject: [PATCH] fix: support unicode in markdown

---
 .../plugins/common/ini_file_parser.py         | 27 +++++----
 test_data/config.md                           | 10 ++++
 tests/plugins/high_entropy_strings_test.py    | 56 ++++++++++++-------
 3 files changed, 61 insertions(+), 32 deletions(-)
 create mode 100644 test_data/config.md

diff --git a/detect_secrets/plugins/common/ini_file_parser.py b/detect_secrets/plugins/common/ini_file_parser.py
index 737b7b0a6..c7918f3fe 100644
--- a/detect_secrets/plugins/common/ini_file_parser.py
+++ b/detect_secrets/plugins/common/ini_file_parser.py
@@ -17,23 +17,26 @@ def __init__(self, file, add_header=False, exclude_lines_regex=None):
         :param exclude_lines_regex: optional regex for ignored lines.
         """
         self.parser = configparser.ConfigParser()
-        self.parser.optionxform = str
+        try:
+            # python2.7 compatible
+            self.parser.optionxform = unicode
+        except NameError:
+            self.parser.optionxform = str
 
         self.exclude_lines_regex = exclude_lines_regex
 
-        if not add_header:
-            self.parser.read_file(file)
-        else:
+        content = file.read()
+        if add_header:
             # This supports environment variables, or other files that look
             # like config files, without a section header.
-            content = '[global]\n' + file.read()
-
-            try:
-                # python2.7 compatible
-                self.parser.read_string(unicode(content))
-            except NameError:
-                # python3 compatible
-                self.parser.read_string(content)
+            content = '[global]\n' + content
+
+        try:
+            # python2.7 compatible
+            self.parser.read_string(unicode(content))
+        except NameError:
+            # python3 compatible
+            self.parser.read_string(content)
 
         # Hacky way to keep track of line location
         file.seek(0)
diff --git a/test_data/config.md b/test_data/config.md
new file mode 100644
index 000000000..e3c170e71
--- /dev/null
+++ b/test_data/config.md
@@ -0,0 +1,10 @@
+# Sample markdown file
+
+[guides](http://localhost/guilds)
+
+Test Unicode in non ini file would not fail on python 2.7.
+
+╭─ diagnose
+╰» ssh to server x:22324241234423414
+
+key="ToCynx5Se4e2PtoZxEhW7lUJcOX15c54"
diff --git a/tests/plugins/high_entropy_strings_test.py b/tests/plugins/high_entropy_strings_test.py
index 7cfe57f2a..9a1474210 100644
--- a/tests/plugins/high_entropy_strings_test.py
+++ b/tests/plugins/high_entropy_strings_test.py
@@ -1,6 +1,7 @@
 from __future__ import absolute_import
 from __future__ import unicode_literals
 
+import codecs
 import string
 
 import pytest
@@ -145,39 +146,54 @@ def setup(self):
             secret_string='c3VwZXIgbG9uZyBzdHJpbmcgc2hvdWxkIGNhdXNlIGVub3VnaCBlbnRyb3B5',
         )
 
-    def test_ini_file(self):
+    @pytest.mark.parametrize(
+        'filename, secrets',
+        [
+            (
+                'test_data/config.ini',
+                [
+                    'Location:    test_data/config.ini:2',
+                    'Location:    test_data/config.ini:6',
+                    'Location:    test_data/config.ini:10',
+                    'Location:    test_data/config.ini:15',
+                    'Location:    test_data/config.ini:21',
+                    'Location:    test_data/config.ini:22', ],
+            ),
+            (
+                'test_data/files/file_with_secrets.py',
+                ['Location:    test_data/files/file_with_secrets.py:3', ],
+            ),
+            # Mark down files with colons and unicode charaters preceding the
+            # colon on the line would cause the scanner to fail and exit on
+            # 2.7 due to ini parser being used on non-markdown files
+            # this test case ensure that scanning can complete and find
+            # high entropy issues
+            (
+                'test_data/config.md',
+                ['Location:    test_data/config.md:10', ],
+            ),
+        ],
+    )
+    def test_ini_file(self, filename, secrets):
         # We're testing two files here, because we want to make sure that
         # the HighEntropyStrings regex is reset back to normal after
         # scanning the ini file.
-        filenames = [
-            'test_data/config.ini',
-            'test_data/files/file_with_secrets.py',
-        ]
 
         plugin = Base64HighEntropyString(3)
 
         accumulated_secrets = {}
-        for filename in filenames:
-            with open(filename) as f:
-                accumulated_secrets.update(
-                    plugin.analyze(f, filename),
-                )
+        with codecs.open(filename, encoding='utf-8') as f:
+            accumulated_secrets.update(
+                plugin.analyze(f, filename),
+            )
 
         count = 0
         for secret in accumulated_secrets.values():
             location = str(secret).splitlines()[1]
-            assert location in (
-                'Location:    test_data/config.ini:2',
-                'Location:    test_data/config.ini:6',
-                'Location:    test_data/config.ini:10',
-                'Location:    test_data/config.ini:15',
-                'Location:    test_data/config.ini:21',
-                'Location:    test_data/config.ini:22',
-                'Location:    test_data/files/file_with_secrets.py:3',
-            )
+            assert location in secrets
             count += 1
 
-        assert count == 7
+        assert count == len(secrets)
 
     def test_yaml_file(self):
         plugin = Base64HighEntropyString(