Fix bug in training when function is not found

scripts/gen-training.py

@@ -78,9 +78,13 @@ with open(gname, "r") as f:
 
             dis = get_dis_from_all_dis(int(addr, 16))
             #objs.append((addr, typ, name, dis))
-            if dis is not None:
+            if dis is None:
+                dis = ""
+                typ = "notfound"
+            else:
                 dis = dis.decode("utf8")
-                df = df.append({'Binary': bname, 'Addr': addr, 'Name': name, 'Type': typ, 'Disassembly': dis}, ignore_index=True)
+
+            df = df.append({'Binary': bname, 'Addr': addr, 'Name': name, 'Type': typ, 'Disassembly': dis}, ignore_index=True)
             
             if False:
                 df.to_csv(oname, index=False)