Update TextProcessor.py

TextProcessor.py

@@ -1,3 +1,4 @@
+# This file converts the images into text
 import pandas as pd
 import numpy as np
 import fitz
@@ -77,11 +78,13 @@ def paragraphs(folderpath):
 
 def process_file(folderpath):
     paras_df = paragraphs(folderpath)
-    paras_df.to_csv(folderpath + '/paragraphs.csv', index=True)
+    paras_df.to_csv(folderpath + '/paragraphs.csv', index=False)
     data_df = pd.read_csv(folderpath + '/data.csv')
     paras_lines = paras_df['Lines'].tolist()
     indents = []
     for (i, para_lines) in enumerate(paras_lines):
+        if para_lines == []:
+            continue
         para = []
         para_start_pg_ind, para_start_line_ind, _, para_first_line = para_lines[0]
         page_df = data_df[data_df['Pg Ind'] == para_start_pg_ind]
@@ -94,3 +97,4 @@ def process_file(folderpath):
         image = cv2.imread(folderpath + '/' + str(pg_ind) + '-processed.png')
         cv2.circle(image, (x1 - 15, int(0.5 * (y1 + y2))), radius=1, color=(240, 32, 160), thickness=2)
         cv2.imwrite(folderpath + '/' + str(pg_ind) + '-processed.png', image)
+