From a0b7f3537f6c6b2de47496a9ca37dd71a7070638 Mon Sep 17 00:00:00 2001
From: Cleber Rosa <crosa@redhat.com>
Date: Fri, 6 Apr 2018 18:02:19 -0400
Subject: [PATCH] String utilities: add checks for data type (text or binary)

To lay some common ground across Python versions, let's define
the criteria we have for binary or text data.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
---
 avocado/utils/astring.py       | 26 ++++++++++++++++++++++++++
 selftests/unit/test_astring.py | 29 +++++++++++++++++++++++++++++
 2 files changed, 55 insertions(+)

diff --git a/avocado/utils/astring.py b/avocado/utils/astring.py
index e4b1e75f..cce7db3f 100644
--- a/avocado/utils/astring.py
+++ b/avocado/utils/astring.py
@@ -27,6 +27,7 @@ And not notice until their code starts failing.
 
 import itertools
 import re
+import sys
 import string
 
 from six import string_types, PY3
@@ -278,3 +279,28 @@ def string_to_safe_path(input_str):
         for bad_chr in FS_UNSAFE_CHARS:
             input_str = input_str.replace(bad_chr, "_")
         return input_str
+
+
+def is_bytes(data):
+    """
+    Checks if the data given is a sequence of bytes
+
+    And not a "text" type, that can be of multi-byte characters.
+    Also, this does NOT mean a bytearray type.
+
+    :param data: the instance to be checked if it falls under the definition
+                 of an array of bytes.
+    """
+    return isinstance(data, bytes)
+
+
+def is_text(data):
+    """
+    Checks if the data given is a suitable for holding text
+
+    That is, if it can hold text that requires more than one byte for
+    each character.
+    """
+    if sys.version_info[0] < 3:
+        return isinstance(data, unicode)
+    return isinstance(data, str)
diff --git a/selftests/unit/test_astring.py b/selftests/unit/test_astring.py
index 7c225a21..56e1e0d9 100644
--- a/selftests/unit/test_astring.py
+++ b/selftests/unit/test_astring.py
@@ -77,6 +77,35 @@ class AstringTest(unittest.TestCase):
         self.assertEqual(astring.string_to_safe_path(avocado),
                          "%s__" % avocado[:-2])
 
+    def test_is_bytes(self):
+        """
+        Verifies what bytes means, basically that they are the same
+        thing accross Python 2 and 3 and can be decoded into "text"
+        """
+        binary = b''
+        text = u''
+        self.assertTrue(astring.is_bytes(binary))
+        self.assertFalse(astring.is_bytes(text))
+        self.assertTrue(hasattr(binary, 'decode'))
+        self.assertTrue(astring.is_text(binary.decode()))
+        # on Python 2, each str member is also a single byte char
+        if sys.version_info[0] < 3:
+            self.assertTrue(astring.is_bytes(str('')))
+        else:
+            self.assertFalse(astring.is_bytes(str('')))
+
+    def test_is_text(self):
+        """
+        Verifies what text means, basically that they can represent
+        extended set of characters and can be encoded into "bytes"
+        """
+        binary = b''
+        text = u''
+        self.assertTrue(astring.is_text(text))
+        self.assertFalse(astring.is_text(binary))
+        self.assertTrue(hasattr(text, 'encode'))
+        self.assertTrue(astring.is_bytes(text.encode()))
+
 
 if __name__ == '__main__':
     unittest.main()
-- 
GitLab