fix page 6-7 extraction issue

app.py

@@ -159,6 +159,12 @@ if 'page_count' in st.session_state:
     else:
         st.session_state.num_pages_to_extract = st.slider('Number of pages to extract:', min_value=1, max_value=st.session_state.page_count, value=st.session_state.num_pages_to_extract , key='num_pages_to_extract_slider')
 
+    if 'num_pages_to_extract2'not in st.session_state:
+        st.session_state.num_pages_to_extract2 = st.slider('Number of pages to extract:', min_value=1, max_value=st.session_state.page_count, value=6, key='num_pages_to_extract_slider2')
+    else:
+        st.session_state.num_pages_to_extract2 = st.slider('Number of pages to extract:', min_value=1, max_value=st.session_state.page_count, value=st.session_state.num_pages_to_extract2 , key='num_pages_to_extract_slider2')
+
+
     st.write(f"num of pages to extract {st.session_state.num_pages_to_extract}")
     if 'run_button' in st.session_state and st.session_state.run_button == True:
         st.session_state.running = True
@@ -175,16 +181,17 @@ if 'page_count' in st.session_state:
         pdf_tables_image_list=[]
         st.session_state.pdf_text_list=[]
 
-        for page_number in range(st.session_state.num_pages_to_extract):
+        for page_number in range(st.session_state.num_pages_to_extract,
+                                 st.session_state.num_pages_to_extract2+1):
              image = pdf2image.convert_from_path(st.session_state.uploaded_pdf_path, first_page=page_number+1, last_page=page_number+1)
              st.session_state.color_image_list.append(image[0])
-             progress_percentage = (page_number) / (st.session_state.num_pages_to_extract-1)
+             progress_percentage = (page_number-st.session_state.num_pages_to_extract) / (st.session_state.num_pages_to_extract2-st.session_state.num_pages_to_extract)
              read_pdf_progress_bar.progress(progress_percentage)
         read_pdf_progress_bar.progress(0)
         for index, image in enumerate(st.session_state.color_image_list):
             image_np = np.array(image)
             st.session_state.gray_image_np_list.append(cv2.cvtColor(np.array(image_np), cv2.COLOR_BGR2GRAY))
-            progress_percentage = (index) / (st.session_state.num_pages_to_extract - 1)
+            progress_percentage = (index) / len(st.session_state.color_image_list)
             read_pdf_progress_bar.progress(progress_percentage)
         st.session_state.extracted_text = ""
 
@@ -195,34 +202,38 @@ if 'page_count' in st.session_state:
 
 
           figures_image_list,tables_image_list,textbox_image_list,text=utils.gray_pdf_image_np_to_text(index,gray_pdf_image_np, debug=True)
-          print("index="+str(index)+" txt book " + str(len(textbox_image_list)))
+          if textbox_image_list:
+            print("index="+str(index)+" txt book " + str(len(textbox_image_list)))
           st.session_state.pdf_figures_image_list.append(figures_image_list)
           st.session_state.pdf_tables_image_list.append(tables_image_list)
           st.session_state.pdf_textbox_image_list.append(textbox_image_list)
           if st.session_state.pdf_figures_image_list[index]:
-              for pdf_figure_text_image in st.session_state.pdf_figures_image_list[index]:
-                  raw_image_file_name = f"page_{index+1}_{pdf_figure_text_image[0]}.png"
-                  cleaned_image_file_name = clean_filename(raw_image_file_name)
-                  Image.fromarray(pdf_figure_text_image[1]).save(temp_figure_dir+cleaned_image_file_name)
+              if st.session_state.pdf_figures_image_list[index]:
+                  for pdf_figure_text_image in st.session_state.pdf_figures_image_list[index]:
+                      raw_image_file_name = f"page_{index+1}_{pdf_figure_text_image[0]}.png"
+                      cleaned_image_file_name = clean_filename(raw_image_file_name)
+                      Image.fromarray(pdf_figure_text_image[1]).save(temp_figure_dir+cleaned_image_file_name)
           if st.session_state.pdf_tables_image_list:
-              for pdf_table_text_image in st.session_state.pdf_tables_image_list[index]:
-                  raw_image_file_name = f"page_{index+1}_{pdf_table_text_image[0]}.png"
-                  cleaned_image_file_name = clean_filename(raw_image_file_name)
-                  Image.fromarray(pdf_table_text_image[1]).save(temp_table_dir + cleaned_image_file_name)
+              if st.session_state.pdf_tables_image_list[index]:
+                  for pdf_table_text_image in st.session_state.pdf_tables_image_list[index]:
+                      raw_image_file_name = f"page_{index+1}_{pdf_table_text_image[0]}.png"
+                      cleaned_image_file_name = clean_filename(raw_image_file_name)
+                      Image.fromarray(pdf_table_text_image[1]).save(temp_table_dir + cleaned_image_file_name)
           if st.session_state.pdf_textbox_image_list:
               textbox_index = 1
-              for pdf_textbox_image in st.session_state.pdf_textbox_image_list[index]:
-                  raw_image_file_name = f"page_{index+1}_textbox_{textbox_index}.png"
-                  cleaned_image_file_name = clean_filename(raw_image_file_name)
-                  Image.fromarray(pdf_textbox_image).save(temp_textbox_dir + cleaned_image_file_name)
-                  textbox_index = textbox_index + 1
+              if st.session_state.pdf_textbox_image_list[index]:
+                  for pdf_textbox_image in st.session_state.pdf_textbox_image_list[index]:
+                      raw_image_file_name = f"page_{index+1}_textbox_{textbox_index}.png"
+                      cleaned_image_file_name = clean_filename(raw_image_file_name)
+                      Image.fromarray(pdf_textbox_image).save(temp_textbox_dir + cleaned_image_file_name)
+                      textbox_index = textbox_index + 1
 
 
-          st.session_state.pdf_text_list.append(text)
-          st.session_state.extracted_text=st.session_state.extracted_text+f"<Page {index+1} start>\n" + text + f"\n<Page {index+1} end>\n>"
+              st.session_state.pdf_text_list.append(text)
+              st.session_state.extracted_text=st.session_state.extracted_text+f"<Page {index+1} start>\n" + text + f"\n<Page {index+1} end>\n>"
           # st.write(text)
           # print(text)
-          progress_percentage = (index) / (st.session_state.num_pages_to_extract - 1)
+          progress_percentage = (index) / len(st.session_state.color_image_list)
           read_pdf_progress_bar.progress(progress_percentage)
         st.session_state.figure_zip_bytes=zip_directory(temp_figure_dir)
         st.session_state.table_zip_bytes = zip_directory(temp_table_dir)

utils.py

@@ -393,7 +393,7 @@ def gray_pdf_image_np_to_text(image_index,gray_pdf_image_np, debug=False):
           print("Table of Contents")
           # display_image_np(max_height_image)
           #print(text)
-          return("Table of Contents")
+          return None, None, None, "Table of Contents"
         else:
           print("not Table of Contents")
           indeces_stop=get_where_image_np_two_columns_stops(max_height_image,20,10)
@@ -420,7 +420,8 @@ def gray_pdf_image_np_to_text(image_index,gray_pdf_image_np, debug=False):
             print("no left over end")
           text=extract_two_columns_text(max_height_image_copy,debug)
           if text == "error":
-            return("error")
+            return None, None, None, "error"
+            # return()
           else:
             return figures_image_list,tables_image_list,text_box_list,text
     else: