Initial implementation of nose testing. Still some cleanup to do, but this shows the differances between the old and the new. Also left one test failing (unsignificant white space only) to demonstrate what a failing test looks like.

1 files changed, 62 insertions, 0 deletions

diff --git a/tests/base.py b/tests/base.py
new file mode 100644
index 0000000..9e70996
--- /dev/null
+++ b/tests/base.py
@@ -0,0 +1,62 @@
+import os, codecs
+import markdown
+from nose.tools import assert_equal
+import difflib
+from plugins import MdSyntaxError
+
+class SyntaxBase:
+    """
+    Generator that steps through all files in a dir and runs each text file
+    (*.txt) as a seperate unit test.
+
+    Each subclass of SyntaxBase should define `dir` as a string containing the 
+    name of the directory which holds the test files. For example:
+
+        dir = "path/to/mytests"
+
+    A subclass may redefine the `setUp` method to create a custom `Markdown`
+    instance specific to that batch of tests.
+    
+    """
+    
+    dir = ""
+
+    def __init__(self):
+        self.files = [x.replace(".txt", "")
+                      for x in os.listdir(self.dir) if x.endswith(".txt")]
+
+    def setUp(self):
+        """ 
+        Create  Markdown instance. 
+        
+        Override this method to create a custom `Markdown` instance assigned to 
+        `self.md`. For example:
+
+            self.md = markdown.Markdown(extensions=["footnotes"], safe_mode="replace")
+
+        """
+        self.md = markdown.Markdown()
+
+    def tearDown(self):
+        """ tearDown is not implemented. """
+        pass
+
+    def test_syntax(self):
+        for file in self.files:
+            yield self.check_syntax, file
+
+    def check_syntax(self, file):
+        input_file = os.path.join(self.dir, file + ".txt")
+        input = codecs.open(input_file, encoding="utf-8").read()
+        output_file = os.path.join(self.dir, file + ".html")
+        expected_output = codecs.open(output_file, encoding="utf-8").read()
+        output = self.normalize(self.md.convert(input))
+        diff = [l for l in difflib.unified_diff(self.normalize(expected_output),
+                                                output, output_file, 
+                                                'actual_output.html', n=3)]
+        if diff:
+            #assert False, 
+            raise MdSyntaxError('Output from "%s" failed to match expected output.\n\n%s' % (input_file, ''.join(diff)))
+
+    def normalize(self, text):
+        return ['%s\n' % l for l in text.strip().split('\n')]