Fixing for run.sh and adding run.sh

.gitignore

@@ -10,3 +10,4 @@ __pycache__
 prev
 old
 data
+venv

run.sh

@@ -0,0 +1 @@
+pypy3 src/wani.py data/Kolokacije_strukture_JOS-32-representation_3D_08_1.xml data/input --out data/output --sloleks_db 'superbaza:A)2U&+3Vfd$Fg]Gb:kolokacije:127.0.0.1' --collocation_sentence_map_dest data/collocation_sentence_mapper --db /mnt/tmp/mysql-wani --multiple-output 

src/loader.py

@@ -20,7 +20,7 @@ def load_files(args, database):
     do_msd_translate = not args.no_msd_translate
 
     if len(filenames) == 1 and os.path.isdir(filenames[0]):
-        filenames = [os.path.join(filenames[0], file) for file in os.listdir(filenames[0])]
+        filenames = [os.path.join(filenames[0], file) for file in os.listdir(filenames[0]) if file[-5:] != '.zstd']
 
     database.init("CREATE TABLE Files ( filename varchar(2048) )")
 

src/representation.py

@@ -4,7 +4,7 @@ from collections import Counter
 from codes_tagset import TAGSET, CODES
 from word import WordMsdOnly
 
-from src.word import WordDummy
+from word import WordDummy
 
 
 class ComponentRepresentation:

src/sloleks_db.py

@@ -1,8 +1,11 @@
+from psycopg2cffi import compat
+compat.register()
+
 from sqlalchemy.ext.declarative import declarative_base
 from sqlalchemy.orm import Session, aliased
 from sqlalchemy import create_engine
 
-from src.codes_tagset import TAGSET, CODES, CODES_TRANSLATION
+from codes_tagset import TAGSET, CODES, CODES_TRANSLATION
 
 
 class SloleksDatabase:

src/wani.py

@@ -21,7 +21,7 @@ from loader import load_files
 from database import Database
 from time_info import TimeInfo
 
-from src.postprocessor import Postprocessor
+from postprocessor import Postprocessor
 
 
 def match_file(words, structures, postprocessor):