add support for emoji

afinn/afinn.py

@@ -9,11 +9,11 @@
 
 from os.path import dirname, join
 
-
 LANGUAGE_TO_FILENAME = {
     'da': 'AFINN-da-32.txt',
     'en': 'AFINN-en-165.txt',
     'emoticons': 'AFINN-emoticon-8.txt',
+    'emoji': 'AFINN-emoji.txt',
     'fr': 'AFINN-fr-165.txt',
     'sv': 'AFINN-sv-165.txt',
     }
@@ -56,7 +56,8 @@ class Afinn(object):
 
     """
 
-    def __init__(self, language="en", emoticons=False, word_boundary=True):
+    def __init__(self, language="en", emoticons=False, emoji=False,
+                 word_boundary=True):
         """Setup dictionary from data file.
 
         The language parameter can be set to English (en) or Danish (da).
@@ -67,31 +68,46 @@ def __init__(self, language="en", emoticons=False, word_boundary=True):
             Specify language dictionary.
         emoticons : bool, optional
             Includes emoticons in the token list
+        emoji : bool, optional
+            Includes emoji in the token list
         word_boundary : bool, optional
             Use word boundary match in the regular expression.
 
         """
         filename = LANGUAGE_TO_FILENAME[language]
         full_filename = self.full_filename(filename)
-        if emoticons:
-            # Words
+
+        if emoticons or emoji:
             self._dict = self.read_word_file(full_filename)
-            regex_words = self.regex_from_tokens(
+            regexp_words = self.regex_from_tokens(
                 list(self._dict),
-                word_boundary=True, capture=False)
-
-            # Emoticons
-            filename_emoticons = LANGUAGE_TO_FILENAME['emoticons']
-            full_filename_emoticons = self.full_filename(filename_emoticons)
-            emoticons_and_score = self.read_word_file(full_filename_emoticons)
-            self._dict.update(emoticons_and_score)
-            regex_emoticons = self.regex_from_tokens(
-                list(emoticons_and_score), word_boundary=False,
-                capture=False)
+                word_boundary=True,
+                capture=False
+            )
+
+            words = []
+            if emoticons:
+                fname = LANGUAGE_TO_FILENAME['emoticons']
+                full_fname = self.full_filename(fname)
+                emoticon_score_dct = self.read_word_file(full_fname)
+                self._dict.update(emoticon_score_dct)
+                words.extend(list(emoticon_score_dct))
+
+            if emoji:
+                full_fname = self.full_filename(LANGUAGE_TO_FILENAME['emoji'])
+                emoji_score_dct = self.read_word_file(full_fname)
+                self._dict.update(emoji_score_dct)
+                words.extend(list(emoji_score_dct))
+
+            regexp_extra = self.regex_from_tokens(
+                words,
+                word_boundary=False,
+                capture=False
+            )
 
             # Combined words and emoticon regular expression
-            regex = '(' + regex_words + '|' + regex_emoticons + ')'
-            self._setup_pattern_from_regex(regex)
+            regexp = '(%s|%s)' % (regexp_words, regexp_extra)
+            self._setup_pattern_from_regex(regexp)
 
         else:
             self.setup_from_file(full_filename, word_boundary=word_boundary)

afinn/data/AFINN-emoji.txt

@@ -0,0 +1,76 @@
+😀	3
+😬	-2
+😁	-2
+😂	2
+😃	3
+😄	3
+😅	3
+😆	3
+😇	1
+😉	1
+😊	2
+🙂	2
+🙃	1
+☺️	2
+😋	1
+😌	1
+😍	3
+😘	2
+😗	1
+😙	1
+😚	1
+😜	1
+😝	1
+😛	1
+🤑	1
+🤓	1
+😎	1
+🤗	2
+😏	2
+😶	-1
+😐	-2
+😑	-2
+😒	-2
+🙄	-1
+🤔	-1
+😳	-1
+😞	-2
+😟	-2
+😠	-3
+😡	-3
+😔	-2
+😕	-1
+🙁	-2
+☹️	-3
+😣	-2
+😖	-2
+😫	-3
+😩	-3
+😤	-2
+😮	-1
+😱	-1
+😨	-1
+😰	-1
+😯	-1
+😦	-1
+😧	-1
+😢	-2
+😥	-2
+😪	-2
+😓	-1
+😭	-3
+😵	-3
+😲	-3
+🤐	-1
+😴	-1
+👍	3
+👎	-3
+💋	1
+❤️	3
+💛	3
+💚	3
+💙	3
+💜	3
+💔	-2
+💯	2
+🖕	-3

setup.py

@@ -6,7 +6,7 @@
 setup(
     name='afinn',
     packages=['afinn'],
-    version='0.2dev',
+    version='0.3',
     author='Finn Aarup Nielsen',
     author_email='[email protected]',
     description='AFINN sentiment analysis',