🐛 Handle text with only punctuations

Closes #36

Author: nipunsadvilkar

pysbd/lang/standard.py

@@ -39,6 +39,7 @@ class DoublePunctuationRules(object):
     SecondRule = Rule(r'!\?', '☈')
     ThirdRule = Rule(r'\?\?', '☇')
     ForthRule = Rule(r'!!', '☄')
+    DoublePunctuation = r'\?!|!\?|\?\?|!!'
     All = [FirstRule, SecondRule, ThirdRule, ForthRule]
 
 

pysbd/processor.py

@@ -143,8 +143,10 @@ def process_text(self, txt):
             txt += 'ȸ'
         txt = ExclamationWords.apply_rules(txt)
         txt = BetweenPunctuation(txt).replace()
-        txt = Text(txt).apply(*DoublePunctuationRules.All,
-                              Standard.QuestionMarkInQuotationRule,
+        # handle text having only doublepunctuations
+        if not re.match(DoublePunctuationRules.DoublePunctuation, txt):
+            txt = Text(txt).apply(*DoublePunctuationRules.All)
+        txt = Text(txt).apply(Standard.QuestionMarkInQuotationRule,
                               *ExclamationPointRules.All)
         txt = ListItemReplacer(txt).replace_parens()
         txt = self.sentence_boundary_punctuation(txt)

tests/regression/test_issues.py

@@ -20,6 +20,7 @@
     ('#34', '..', ['..']),
     ('#34', '. . .', ['. . .']),
     ('#34', '! ! !', ['! ! !']),
+    ('#36', '??', ['??'])
 ]
 
 @pytest.mark.parametrize('issue_no,text,expected_sents', TEST_ISSUE_DATA)