Add support for loading yaml strings as unicode.

CHANGES.rst

@@ -1,10 +1,15 @@
 Changelog for Charlatan
 =======================
 
+0.3.5 (unreleased)
+------------------
+
+- Support loading all strings as unicode
+
 0.3.4 (2014-01-21)
 ------------------
 
-- Fix getting attribute from relationhips
+- Fix getting attribute from relationships
 
 0.3.3 (2014-01-18)
 ------------------

charlatan/file_format.py

@@ -1,8 +1,8 @@
 from __future__ import absolute_import
-from __future__ import unicode_literals
 import datetime
 
 import yaml
+from yaml.constructor import Constructor
 
 from charlatan.utils import apply_delta, datetime_to_epoch_timestamp
 
@@ -57,7 +57,19 @@ def relationship_constructor(loader, node):
     yaml.add_constructor(u'!rel', relationship_constructor)
 
 
-def load_file(filename):
+def configure_output(use_unicode=False):
+    """Configure output options of the values loaded by pyyaml
+
+        :param boolean use_unicode: Use unicode constructor for loading strings
+    """
+    if use_unicode:
+        yaml.add_constructor(
+            u'tag:yaml.org,2002:str',
+            Constructor.construct_python_unicode,
+        )
+
+
+def load_file(filename, unicode_output=False):
     """Load fixtures definition from file.
 
     :param str filename:
@@ -69,6 +81,7 @@ def load_file(filename):
     if filename.endswith(".yaml"):
         # Load the custom YAML tags
         configure_yaml()
+        configure_output(use_unicode=unicode_output)
         content = yaml.load(content)
     else:
         raise ValueError("Unsupported filetype: '%s'" % filename)

charlatan/fixtures_manager.py

@@ -47,10 +47,11 @@ class FixturesManager(object):
 
     """
 
-    def __init__(self, db_session=None):
+    def __init__(self, db_session=None, unicode_output=False):
         self.hooks = {}
         self.session = db_session
         self.installed_keys = []
+        self.unicode_output = unicode_output
 
     def load(self, filename, models_package=""):
         """Pre-load the fixtures.
@@ -87,7 +88,7 @@ def _load_fixtures(self, filename):
         :param str filename: file that holds the fixture data
         """
 
-        content = load_file(filename)
+        content = load_file(filename, self.unicode_output)
 
         fixtures = {}
         for k, v in _compat.iteritems(content):
@@ -421,7 +422,7 @@ def set_hook(self, hookname, func):
         :param function func:
         """
 
-        if not hookname in ALLOWED_HOOKS:
+        if hookname not in ALLOWED_HOOKS:
             raise KeyError("'%s' is not an allowed hook." % hookname)
 
         self.hooks[hookname] = func

charlatan/tests/data/strings.yaml

@@ -0,0 +1,2 @@
+foo: bar
+baz: foobar

charlatan/tests/test_strings.py

@@ -0,0 +1,54 @@
+from __future__ import absolute_import
+from sys import version_info
+
+import unittest
+from yaml import add_constructor
+from yaml.constructor import SafeConstructor
+
+from charlatan import file_format, testing
+
+
+class TestUnicodeLoad(testing.TestCase):
+
+    def setUp(self):
+        # preserve the original constructor for strings
+        self.str_constructor = SafeConstructor.yaml_constructors[
+            u'tag:yaml.org,2002:str'
+        ]
+        self.yaml = file_format.load_file(
+            './charlatan/tests/data/strings.yaml',
+            unicode_output=True,
+        )
+
+    def tearDown(self):
+        # reset the constructor
+        add_constructor(u'tag:yaml.org,2002:str', self.str_constructor)
+
+    @unittest.skipIf(version_info[0] == 3, 'Unicode is undefined in Python 3')
+    def test_strings_are_unicode(self):
+        """Assert all strings are loaded as unicode"""
+        for key, val in self.yaml.items():
+            self.assertTrue(isinstance(key, unicode))  # noqa
+            self.assertTrue(isinstance(val, unicode))  # noqa
+
+
+class TestStringLoad(testing.TestCase):
+
+    def setUp(self):
+        self.yaml = file_format.load_file(
+            './charlatan/tests/data/strings.yaml',
+        )
+
+    @unittest.skipIf(version_info[0] == 3, 'Iteration has changed in Python 3')
+    def test_strings_are_strings(self):
+        """Assert all strings are loaded as strings"""
+        for key, val in self.yaml.items():
+            self.assertTrue(isinstance(key, str))
+            self.assertTrue(isinstance(val, str))
+
+    @unittest.skipIf(version_info[0] == 2, 'Iteration has changed in Python 3')
+    def test_strings_are_strings_python3(self):
+        """Assert all strings are loaded as strings"""
+        for key, val in list(self.yaml.items()):
+            self.assertTrue(isinstance(key, str))
+            self.assertTrue(isinstance(val, str))

docs/file-format.rst

@@ -275,3 +275,12 @@ All the same time deltas work.
 
 .. versionadded:: 0.2.9
     It is now possible to use times in seconds since the epoch
+
+Unicode Strings
+---------------
+
+.. versionadded:: 0.3.5
+
+In python 2 strings are not, by default, loaded as unicode.  To load all the 
+strings from the yaml files as unicode strings, pass the option 
+`unicode_output` as `True` when you instantiate your fixture manager.