Added fix for punctuations

luscenje_struktur/codes_tagset.py

@@ -133,6 +133,7 @@ CODES = {
     "Interjection": "I",
     "Abbreviation": "Y",
     "Residual": "X",
+    "Punctuation": "Z",
 
     'common': 'c',
     'proper': 'p',

luscenje_struktur/component.py

@@ -147,9 +147,6 @@ class Component:
                 if not order.match(word, next_word):
                     continue
 
-                if word.lemma == 'aktivirati' and next_word.text == 'potomcih':
-                    a = 0
-
                 match = next.match(next_word)
 
                 if match is not None:

luscenje_struktur/syntactic_structure.py

@@ -12,7 +12,7 @@ class SyntacticStructure:
         self.components = []
 
     @staticmethod
-    def from_xml(xml):
+    def from_xml(xml, no_stats):
         st = SyntacticStructure()
         st.id = xml.get('id_nsss')
         st.lbs = xml.get('LBS')
@@ -47,7 +47,8 @@ class SyntacticStructure:
         fake_root_component = Component({'cid': '#', 'type': 'other'})
         st.components = fake_root_component.find_next(deps, comps, restrs, forms)
 
-        st.determine_core2w()
+        if not no_stats:
+            st.determine_core2w()
         return st
 
     def determine_core2w(self):
@@ -98,6 +99,7 @@ class SyntacticStructure:
 
 def build_structures(args):
     filename = args.structures
+    no_stats = args.out is None or args.stats is None
 
     max_num_components = -1
     with open(filename, 'r') as fp:
@@ -105,7 +107,7 @@ def build_structures(args):
 
     structures = []
     for structure in et.iter('syntactic_structure'):
-        to_append = SyntacticStructure.from_xml(structure)
+        to_append = SyntacticStructure.from_xml(structure, no_stats)
         if to_append is None:
             continue