Merge branch 'feature/russian_ocr' into development

apps/ocr/api.py

@@ -31,7 +31,7 @@ def get_language_backend():
     try:
         module = import_module(u'.'.join([u'ocr', u'lang', TESSERACT_LANGUAGE]))
     except ImportError:
-        sys.stderr.write(u'\nError: No OCR app language backend for language: %s\n\n' % TESSERACT_LANGUAGE)
+        sys.stderr.write(u'\nWarning: No OCR app language backend for language: %s\n\n' % TESSERACT_LANGUAGE)
         return None
     return module
 

apps/ocr/lang/rus.py

@@ -0,0 +1,39 @@
+# -*- coding: utf-8 -*-
+import re
+
+
+def check_word(word):
+    ALL_ALPHANUM = re.compile('([0-9ёйцукенгшщзхъфывапролджэячсмитьбю])', re.I)
+    NON_ALPHANUM = re.compile('([^0-9ёйцукенгшщзхъфывапролджэячсмитьбю])', re.I)
+
+    TOO_MANY_VOWELS = re.compile('[ёуеыаоэяию]{3}', re.I)
+    TOO_MANY_CONSONANTS = re.compile('[йцкнгшщзхъфвпрлджчсмтьб{5}', re.I)
+    ALL_ALPHA = re.compile('^[ёйцукенгшщзхъфывапролджэячсмитьбю]+$', re.I)
+    SINGLE_LETTER_WORDS = re.compile('^[уквояси]$', re.I)
+
+    #(L) If a string is longer than 25 characters, it is garbage
+    if len(word) > 25:
+        return None
+
+    #(A) If a string's ratio of alphanumeric characters to total
+    #characters is less than 50%, the string is garbage
+    if len(ALL_ALPHANUM.findall(word)) < len(word) / 2:
+        return None
+
+    #Remove word if all the letters in the word are non alphanumeric
+    if len(NON_ALPHANUM.findall(word)) == len(word):
+        return None
+
+    #Removed words with too many consecutie vowels
+    if TOO_MANY_VOWELS.findall(word):
+        return None
+
+    #Removed words with too many consecutie consonants
+    if TOO_MANY_CONSONANTS.findall(word):
+        return None
+
+    #Only allow specific single letter words
+    if len(word) == 1 and not SINGLE_LETTER_WORDS.findall(word):
+        return None
+
+    return word

docs/contributors.rst

@@ -33,4 +33,4 @@ Translations
 ------------
 * Emerson Soares (http://emersonsoares.com)
 * Renata Oliveira (https://twitter.com/#!/rnataoliveira)
-* Sergey Glita (s.v.glita@gmail.com)
+* Сергей Глита [Sergey Glita] (s.v.glita@gmail.com)