pattern.ru spelling, lexicon

Author: Xsardas1000

examples/03-en/05-tagset.py

@@ -34,7 +34,7 @@
 print(parse_fr("les chats noirs", chunks=False))      # les/DT chats/NNS noirs/JJ
 print(parse_it("i gatti neri", chunks=False))         # i/DT gatti/NNS neri/JJ
 print(parse_nl("de zwarte katten", chunks=False))     # de/DT zwarte/JJ katten/NNS
-print(parse_ru("Да здравствует Сталин!", chunks=False))
+print(parse_ru("какой сегодня хороший день!", chunks=False)) # какой/DT сегодня/RB хороший/JJ день/NN !/.
 print("")
 
 # In some cases, this means the original tagset is mapped to Penn Treebank:

pattern/text/en/__init__.py

@@ -153,7 +153,7 @@ def load(self, path=None):
 )
 
 spelling = Spelling(
-        path = os.path.join(MODULE, "en-spelling.txt")
+    path = os.path.join(MODULE, "en-spelling.txt")
 )
 
 

pattern/text/ru/__init__.py

@@ -42,6 +42,11 @@
     SLASH, WORD, POS, CHUNK, PNP, REL, ANCHOR, LEMMA, AND, OR
 )
 
+# Import spelling base class.
+from pattern.text import (
+    Spelling
+)
+
 sys.path.pop(0)
 
 #--- Russian PARSER --------------------------------------------------------------------------------
@@ -68,6 +73,11 @@ def find_tags(self, tokens, **kwargs):
 )
 
 
+spelling = Spelling(
+    path=os.path.join(MODULE, "ru-spelling.txt"),
+    alphabet='CYRILLIC'
+)
+
 
 def tokenize(s, *args, **kwargs):
     """ Returns a list of sentences, where punctuation marks have been split from words.
@@ -84,4 +94,10 @@ def parse(s, *args, **kwargs):
 def parsetree(s, *args, **kwargs):
     """ Returns a parsed Text from the given string.
     """
-    return Text(parse(s, *args, **kwargs))
+    return Text(parse(s, *args, **kwargs))
+
+
+def suggest(w):
+    """ Returns a list of (word, confidence)-tuples of spelling corrections.
+    """
+    return spelling.suggest(w)