unicode...

[iramuteq] / corpus.py

diff --git a/corpus.py b/corpus.py
index 76ebb66..5a18f62 100644 (file)
--- a/corpus.py
+++ b/corpus.py
@@ -77,10 +77,10 @@ class Corpus :
                 gramtype = self.parent.lexique[word][1]
                 lem = self.parent.lexique[word][0]
             elif word.isdigit() :
-                gramtype = 'num'
+                gramtype = u'num'
                 lem = word
             else :
-                gramtype = 'nr'
+                gramtype = u'nr'
                 lem = word
             self.formes[word] =  Word(word, gramtype, len(self.formes), lem)
             self.idformesuces[self.formes[word].ident] = {self.ucis[-1].uces[-1].ident : 1}
@@ -514,7 +514,7 @@ class Corpus :
                 self.lems[lem].act = 2
             elif self.lems[lem].gram in gramact :
                 self.lems[lem].act = 1
-            elif gramsup is not None :
+            elif gramsup is not None and self.lems[lem].gram not in gramact:
                 if self.lems[lem].gram in gramsup :
                     self.lems[lem].act = 2
                 else :
@@ -1081,7 +1081,7 @@ class BuildCorpus :
     def firstclean(self, txt) :
         txt = txt.replace(u'’',"'")
         txt = txt.replace(u'œ', u'oe')
-        return txt.replace('...',u' £$£ ').replace('?',' ? ').replace('.',' . ').replace('!', ' ! ').replace(',',' , ').replace(';', ' ; ').replace(':',' : ').replace(u'…', ' £$£ ')
+        return txt.replace('...',u' £$£ ').replace('?',' ? ').replace('.',' . ').replace('!', ' ! ').replace(',',' , ').replace(';', ' ; ').replace(':',' : ').replace(u'…', u' £$£ ')
 
     def make_cleans(self, txt) :
         for clean in self.cleans :