Merge pull request #1991 from wjt/empty-md-file

Handle empty Markdown files more gracefully.

pelican/readers.py

@@ -288,7 +288,10 @@ class MarkdownReader(BaseReader):
         with pelican_open(source_path) as text:
             content = self._md.convert(text)
 
-        metadata = self._parse_metadata(self._md.Meta)
+        if hasattr(self._md, 'Meta'):
+            metadata = self._parse_metadata(self._md.Meta)
+        else:
+            metadata = {}
         return content, metadata
 
 

pelican/tests/content/empty_with_bom.md

@@ -0,0 +1 @@
+

pelican/tests/test_cache.py

@@ -56,12 +56,15 @@ class TestCache(unittest.TestCase):
         generator.readers.read_file = MagicMock()
         generator.generate_context()
         """
-        3 Files don't get cached because they were not valid
+        6 files don't get cached because they were not valid
+        - article_with_attributes_containing_double_quotes.html
         - article_with_comments.html
         - article_with_null_attributes.html
         - 2012-11-30_md_w_filename_meta#foo-bar.md
+        - empty.md
+        - empty_with_bom.md
         """
-        self.assertEqual(generator.readers.read_file.call_count, 4)
+        self.assertEqual(generator.readers.read_file.call_count, 6)
 
     @unittest.skipUnless(MagicMock, 'Needs Mock module')
     def test_article_reader_content_caching(self):

pelican/tests/test_readers.py

@@ -546,6 +546,22 @@ class MdReaderTest(ReaderTest):
         }
         self.assertDictHasSubset(metadata, expected)
 
+    def test_empty_file(self):
+        reader = readers.MarkdownReader(settings=get_settings())
+        content, metadata = reader.read(
+            _path('empty.md'))
+
+        self.assertEqual(metadata, {})
+        self.assertEqual(content, '')
+
+    def test_empty_file_with_bom(self):
+        reader = readers.MarkdownReader(settings=get_settings())
+        content, metadata = reader.read(
+            _path('empty_with_bom.md'))
+
+        self.assertEqual(metadata, {})
+        self.assertEqual(content, '')
+
 
 class HTMLReaderTest(ReaderTest):
     def test_article_with_comments(self):

pelican/utils.py

@@ -244,7 +244,7 @@ def pelican_open(filename, mode='rb', strip_crs=(sys.platform == 'win32')):
 
     with codecs.open(filename, mode, encoding='utf-8') as infile:
         content = infile.read()
-    if content[0] == codecs.BOM_UTF8.decode('utf8'):
+    if content[:1] == codecs.BOM_UTF8.decode('utf8'):
         content = content[1:]
     if strip_crs:
         content = content.replace('\r\n', '\n')