Add code to check for typos

1 files changed, 23 insertions, 2 deletions

diff --git a/monkey/monkey.py b/monkey/monkey.py
index 99f5a2a..02048de 100755
--- a/monkey/monkey.py
+++ b/monkey/monkey.py
@@ -2,11 +2,11 @@
 
 import math
 import time
-
 import prolog.engine
+
 from .edits import classify_edits
 from prolog.util import Token, annotate, compose, map_vars, normalized, rename_vars, stringify
-from .util import PQueue
+from .util import damerau_levenshtein, PQueue
 
 # Check whether all tests for problem [name] succeed.
 def test(name, code):
@@ -311,3 +311,24 @@ def fix_hints(code, path):
 
         program[idx:idx+len(a)] = [t.clone(pos=program[idx].pos) for t in b]
         yield fix_type, start, end, msg
+
+
+# Checks for typos in the code and suggest the nearst uploaded term by other users.
+def check_typos(code, names):
+    for token in annotate(code):
+        if token.type == 'NAME':
+            nearest_name = ' '
+            nearest_dist = 1000
+            own_count = names.get(token.val, 0) # count of the token.val which is compared with the
+                                                # each name in the names
+            for name in names.items():
+                if name[0] == token.val:        # If the names are the skip the code
+                    continue
+
+                distance = damerau_levenshtein(token.val, name[0])
+
+                if distance < nearest_dist and distance > 0  and own_count < name[1]:
+                    nearest_dist = distance     # Set best_dist and best_name if the less one is found
+                    nearest_name = name[0]
+            if nearest_dist > 0 and nearest_dist/len(nearest_name) <= 1/3:
+                yield  'typo', token.pos, token.pos + len(token.val) , 'Did you mean "{}"?'.format(nearest_name)