Simplifying and also improving the speed (less regex comparisons!)

wani.py

@@ -109,21 +109,6 @@ TAGSET = {
     "X": ['type']
 }
 
-CATEGORY_BASES = {
-    "N": ['.'] * 5,
-    "V": ['.'] * 7,
-    "A": ['.'] * 6,
-    "R": ['.'] * 2,
-    "P": ['.'] * 6,
-    "M": ['.'] * 6,
-    "S": ['.'] * 1,
-    "C": ['.'] * 1,
-    "Q": [],
-    "I": [],
-    "Y": [],
-    "X": ['.'] * 1
-}
-
 class ComponentType(Enum):
     Other = 0
     Core = 2
@@ -431,7 +416,9 @@ def determine_ppb(rgx):
     if rgx[0] in ("A", "N", "R"):
         return 0
     elif rgx[0] == "V":
-        if 'a' in rgx[1]:
+        if len(rgx) == 1:
+            return 2
+        elif 'a' in rgx[1]:
             return 3
         elif 'm' in rgx[1]:
             return 1
@@ -461,7 +448,7 @@ class MorphologyRegex:
         assert 'POS' in restr_dict
         category = restr_dict['POS'][0].capitalize()
         cat_code = CODES[category]
-        rgx = [cat_code] + CATEGORY_BASES[cat_code]
+        rgx = [cat_code] + ['.'] * 10
 
         del restr_dict['POS']
 
@@ -480,6 +467,13 @@ class MorphologyRegex:
             if typ:
                 self.min_msd_length = max(index + 1, self.min_msd_length)
 
+        # strip rgx
+        for i in reversed(range(len(rgx))):
+            if rgx[i] == '.':
+                rgx = rgx[:-1]
+            else:
+                break
+
         self.re_objects = [re.compile(r) for r in rgx]
         self.rgx = rgx