Merge pull request #99 from mailgun/sergey/capitalized

consider word capitilized only if it is camel case - not all upper case

setup.py

@@ -29,7 +29,7 @@ class InstallCommand(install):
 
 
 setup(name='talon',
-      version='1.2.11',
+      version='1.2.12',
       description=("Mailgun library "
                    "to extract message quotations and signatures."),
       long_description=open("README.rst").read(),

talon/signature/learning/helpers.py

@@ -185,12 +185,13 @@ def capitalized_words_percent(s):
     s = to_unicode(s, precise=True)
     words = re.split('\s', s)
     words = [w for w in words if w.strip()]
+    words = [w for w in words if len(w) > 2]    
     capitalized_words_counter = 0
     valid_words_counter = 0
     for word in words:
         if not INVALID_WORD_START.match(word):
             valid_words_counter += 1
-            if word[0].isupper():
+            if word[0].isupper() and not word[1].isupper():
                 capitalized_words_counter += 1
     if valid_words_counter > 0 and len(words) > 1:
         return 100 * float(capitalized_words_counter) / valid_words_counter

tests/signature/extraction_test.py

@@ -77,6 +77,31 @@ def test_basic():
         signature.extract(msg_body, 'Sergey'))
 
 
+def test_capitalized():
+    msg_body = """Hi Mary,
+
+Do you still need a DJ for your wedding? I've included a video demo of one of our DJs available for your wedding date.
+
+DJ Doe 
+http://example.com
+Password: SUPERPASSWORD
+
+Would you like to check out more?
+
+
+At your service,
+
+John Smith
+Doe Inc
+555-531-7967"""
+
+    sig = """John Smith
+Doe Inc
+555-531-7967"""
+
+    eq_(sig, signature.extract(msg_body, 'Doe')[1])
+
+
 def test_over_2_text_lines_after_signature():
     body = """Blah
 

tests/signature/learning/helpers_test.py

@@ -192,10 +192,11 @@ def test_punctuation_percent(categories_percent):
 def test_capitalized_words_percent():
     eq_(0.0, h.capitalized_words_percent(''))
     eq_(100.0, h.capitalized_words_percent('Example Corp'))
-    eq_(50.0, h.capitalized_words_percent('Qqq qqq QQQ 123 sss'))
+    eq_(50.0, h.capitalized_words_percent('Qqq qqq Aqs 123 sss'))
     eq_(100.0, h.capitalized_words_percent('Cell 713-444-7368'))
     eq_(100.0, h.capitalized_words_percent('8th Floor'))
     eq_(0.0, h.capitalized_words_percent('(212) 230-9276'))
+    eq_(50.0, h.capitalized_words_percent('Password: REMARKABLE'))
 
 
 def test_has_signature():