Support multiple languages

Support english and russian for now.
Change UI according to selected language.

app.py

@@ -2,6 +2,12 @@ import streamlit as st
 
 from category_classification.models import models as class_models
 from common import Input
+from languages import *
+
+title = {en: "Papers classification", ru: "Классификация статей"}
+model_label = {en: "Select model", ru: "Выберете модель"}
+abstract_label = {en: "Abstract", ru: "Аннотация"}
+
 
 @st.cache_data
 def load_class_model(name):
@@ -9,12 +15,12 @@ def load_class_model(name):
     return model
 
 
+lang = st.pills(label=langs_str, options=langs, default=en)
+st.title(title[lang])
 model_name = st.selectbox(
-    "Select model",
-    options=class_models.get_all_model_names()
+    model_label[lang], options=class_models.get_model_names_by_lang(lang)
 )
-title = st.text_area("Paper title")
-abstract = st.text_area("Paper abstract")
+abstract = st.text_area(abstract_label[lang])
 
 if title:
     input = Input(title=title, abstract=abstract)

category_classification/models/models.py

@@ -1,11 +1,15 @@
 import importlib.util
 import os
 import sys
+import typing as tp
 import warnings
 from pathlib import Path
 
+
 def import_model_module(file_path: os.PathLike):
-    module_name = str(Path(file_path).relative_to(os.getcwd())).replace(os.path.sep, ".")
+    module_name = str(Path(file_path).relative_to(os.getcwd())).replace(
+        os.path.sep, "."
+    )
     spec = importlib.util.spec_from_file_location(module_name, file_path)
     module = importlib.util.module_from_spec(spec)
     sys.modules[module_name] = module
@@ -14,6 +18,7 @@ def import_model_module(file_path: os.PathLike):
 
 
 models = {}
+language_to_models = {}
 
 file_dir = Path(__file__).parents[0]
 
@@ -25,8 +30,10 @@ for path in file_dir.glob("*"):
         module = import_model_module(model_file_path)
         name_key = "name"
         get_model_key = "get_model"
+        supported_langs_key = "supported_langs"
         name = getattr(module, name_key, None)
         get_model = getattr(module, get_model_key, None)
+        supported_langs = getattr(module, supported_langs_key, None)
 
         def check_attr_exists(attr_name, attr):
             if attr is None:
@@ -54,12 +61,19 @@ for path in file_dir.glob("*"):
             continue
         if not check_attr_exists(get_model_key, get_model):
             continue
+        if not check_attr_exists(supported_langs_key, supported_langs):
+            continue
         if not check_attr_type(name_key, name, str):
             continue
         if not check_attr_callable(get_model_key, get_model):
             continue
+        if not check_attr_type(supported_langs_key, supported_langs, tp.Iterable):
+            continue
 
         models[name] = get_model
+        for lang in supported_langs:
+            language_to_models.setdefault(lang, {})
+        language_to_models[lang][name] = get_model
 
 
 def get_model(name: str):
@@ -70,3 +84,8 @@ def get_model(name: str):
 
 def get_all_model_names():
     return list(models.keys())
+
+def get_model_names_by_lang(lang):
+    if lang not in language_to_models:
+        return []
+    return language_to_models[lang]

category_classification/models/oracat__bert_paper_classifier/model.py

@@ -15,3 +15,5 @@ class BertPaperClassifierModel:
 
 def get_model():
     return BertPaperClassifierModel()
+
+supported_langs = ['en']

category_classification/models/oracat__bert_paper_classifier_arxiv/model.py

@@ -12,3 +12,5 @@ class BertPaperClassifierArxivModel:
 
 def get_model():
     return BertPaperClassifierArxivModel()
+
+supported_langs = ['en']

languages.py

@@ -0,0 +1,6 @@
+en = "en"
+ru = "ru"
+
+langs = [en, ru]
+
+langs_str = "Language/Язык"