fix assets generation when selecting extractor

app/pages/0_Import_File.py

@@ -7,7 +7,7 @@
 import copy
 from menu import display_pages_menu, display_config
 from pypdf import PdfReader
-from utils import get_pdf_iframe, set_state
+from utils import get_pdf_iframe, set_state, generate_assets
 
 from country_by_country.processor import ReportProcessor
 
@@ -27,28 +27,6 @@ def initiate_configuration() -> None:
         "pagefilter"
     ]["type"]
 
-def generate_assets() -> None:
-    assets = {
-        "pagefilter": {},
-        "table_extractors": [],
-    }
-
-    # Filtering the pages
-    st.session_state["proc"].page_filter(
-        st.session_state["working_file_pdf"].name,
-        assets,
-    )
-
-    logging.info(f"Assets : {assets}")
-
-    if len(assets["pagefilter"]["selected_pages"]) == 0:
-        # No page has been automatically selected by the page filter
-        # Hence, we display the full pdf, letting the user select the pages
-        pdfreader = PdfReader(st.session_state["working_file_pdf"])
-        number_pages = len(PdfReader(st.session_state["working_file_pdf"]).pages)
-        assets["pagefilter"]["selected_pages"] = list(range(number_pages))
-    st.session_state["assets"] = assets
-
 def on_pdf_file_upload() -> None:
     # Change states related to the pdf file upload
     mytmpfile.write(st.session_state.original_pdf.read())

app/pages/1_Selected_Pages.py

@@ -1,6 +1,6 @@
 import streamlit as st
 from country_by_country.processor import ReportProcessor
-from utils import get_pdf_iframe, set_state
+from utils import get_pdf_iframe, set_state, generate_assets
 from country_by_country.utils.utils import keep_pages
 from pypdf import PdfReader
 from menu import display_pages_menu, display_config
@@ -29,7 +29,7 @@ def set_extractors() -> None:
     ]
     set_state(["config", "table_extraction"], selected_extractors_dict)
     st.session_state["proc"] = ReportProcessor(st.session_state["config"])
-
+    generate_assets()
 
 st.set_page_config(layout="wide", page_title="Pages selection")  # page_icon="📈"
 st.title("Country by Country Tax Reporting analysis : Selected Pages")

app/utils.py

@@ -1,8 +1,10 @@
 import base64
+import logging
 from pathlib import Path
 from typing import Any
 
 import pandas as pd
+from pypdf import PdfReader
 import streamlit as st
 
 
@@ -41,7 +43,6 @@ def to_csv_file(df: pd.DataFrame) -> bytes:
 
     return df.to_csv(index=False).encode("utf-8")
 
-
 def set_state(key: Any, value: Any) -> None:
     """
     Sets the session_state[key] to value.
@@ -61,3 +62,24 @@ def set_state(key: Any, value: Any) -> None:
         nested_value[key_list[-1]] = value
     else:
         st.session_state[key] = value
+
+def generate_assets() -> None:
+    assets = {
+        "pagefilter": {},
+        "table_extractors": [],
+    }
+
+    # Filtering the pages
+    st.session_state["proc"].page_filter(
+        st.session_state["working_file_pdf"].name,
+        assets,
+    )
+
+    logging.info(f"Assets : {assets}")
+
+    if len(assets["pagefilter"]["selected_pages"]) == 0:
+        # No page has been automatically selected by the page filter
+        # Hence, we display the full pdf, letting the user select the pages
+        number_pages = len(PdfReader(st.session_state["working_file_pdf"]).pages)
+        assets["pagefilter"]["selected_pages"] = list(range(number_pages))
+    st.session_state["assets"] = assets