Added Abbreviation Extension. Tests included

4 files changed, 113 insertions, 0 deletions

diff --git a/markdown_extensions/abbr.py b/markdown_extensions/abbr.py
new file mode 100644
index 0000000..551e4f5
--- /dev/null
+++ b/markdown_extensions/abbr.py
@@ -0,0 +1,95 @@
+'''
+Abbreviation Extension for Python-Markdown
+==========================================
+
+This extension adds abbreviation handling to Python-Markdown.
+
+Simple Usage:
+
+    >>> import markdown
+    >>> text = """
+    ... Some text with an ABBR and a REF. Ignore REFERENCE and ref.
+    ...
+    ... *[ABBR]: Abbreviation
+    ... *[REF]: Abbreviation Reference
+    ... """
+    >>> markdown.markdown(text, ['abbr'])
+    u'<p>Some text with an <abbr title="Abbreviation">ABBR</abbr> and a <abbr title="Abbreviation Reference">REF</abbr>. Ignore REFERENCE and ref.</p>'
+
+Copyright 2007-2008
+* [Waylan Limberg](http://achinghead.com/)
+* [Seemant Kulleen](http://www.kulleen.org/)
+	
+
+'''
+
+import markdown, re
+from markdown import etree
+
+# Global Vars
+ABBR_REF_RE = re.compile(r'[*]\[(?P<abbr>[^\]]*)\][ ]?:\s*(?P<title>.*)')
+
+class AbbrExtension(markdown.Extension):
+    """ Abbreviation Extension for Python-Markdown. """
+
+    def extendMarkdown(self, md, md_globals):
+        """ Insert AbbrPreprocessor before ReferencePreprocessor. """
+        md.preprocessors.add('abbr', AbbrPreprocessor(md), '<reference')
+        
+           
+class AbbrPreprocessor(markdown.Preprocessor):
+    """ Abbreviation Preprocessor - parse text for abbr references. """
+
+    def run(self, lines):
+        '''
+        Find and remove all Abbreviation references from the text.
+        Each reference is set as a new AbbrPattern in the markdown instance.
+        
+        '''
+        new_text = []
+        for line in lines:
+            m = ABBR_REF_RE.match(line)
+            if m:
+                abbr = m.group('abbr').strip()
+                title = m.group('title').strip()
+                self.markdown.inlinePatterns['abbr-%s'%abbr] = \
+                    AbbrPattern(self._generate_pattern(abbr), title)
+            else:
+                new_text.append(line)
+        return new_text
+    
+    def _generate_pattern(self, text):
+        '''
+        Given a string, returns an regex pattern to match that string. 
+        
+        'HTML' -> r'(?P<abbr>[H][T][M][L])' 
+        
+        Note: we force each char as a literal match (in brackets) as we don't 
+        know what they will be beforehand.
+
+        '''
+        chars = list(text)
+        for i in range(len(chars)):
+            chars[i] = r'[%s]' % chars[i]
+        return r'(?P<abbr>\b%s\b)' % (r''.join(chars))
+
+
+class AbbrPattern(markdown.Pattern):
+    """ Abbreviation inline pattern. """
+
+    def __init__(self, pattern, title):
+        markdown.Pattern.__init__(self, pattern)
+        self.title = title
+
+    def handleMatch(self, m):
+        abbr = etree.Element('abbr')
+        abbr.text = m.group('abbr')
+        abbr.set('title', self.title)
+        return abbr
+
+def makeExtension(configs=None):
+    return AbbrExtension(configs=configs)
+
+if __name__ == "__main__":
+    import doctest
+    doctest.testmod()
diff --git a/test-markdown.py b/test-markdown.py
index 1445799..3d9ef66 100755
--- a/test-markdown.py
+++ b/test-markdown.py
@@ -376,6 +376,7 @@ testDirectory("tests/extensions-x-codehilite")
 testDirectory("tests/extensions-x-wikilinks")
 testDirectory("tests/extensions-x-toc")
 testDirectory("tests/extensions-x-def_list")
+testDirectory("tests/extensions-x-abbr")
 
 print "\n### Final result ###"
 if len(failedTests):
diff --git a/tests/extensions-x-abbr/abbr.html b/tests/extensions-x-abbr/abbr.html
new file mode 100644
index 0000000..456524e
--- /dev/null
+++ b/tests/extensions-x-abbr/abbr.html
@@ -0,0 +1,4 @@
+<p>An <abbr title="Abbreviation">ABBR</abbr>: "<abbr title="Reference">REF</abbr>".
+ref and REFERENCE should be ignored.</p>
+<p>The <abbr title="Hyper Text Markup Language">HTML</abbr> specification
+is maintained by the <abbr title="World Wide Web Consortium">W3C</abbr>.</p>
\ No newline at end of file
diff --git a/tests/extensions-x-abbr/abbr.txt b/tests/extensions-x-abbr/abbr.txt
new file mode 100644
index 0000000..991bf15
--- /dev/null
+++ b/tests/extensions-x-abbr/abbr.txt
@@ -0,0 +1,13 @@
+An ABBR: "REF".
+ref and REFERENCE should be ignored.
+
+*[REF]: Reference
+*[ABBR]: This gets overriden by the next one.
+*[ABBR]: Abbreviation
+
+The HTML specification
+is maintained by the W3C.
+
+*[HTML]: Hyper Text Markup Language
+*[W3C]:  World Wide Web Consortium
+