surinala commited on
Commit
0b45fc7
·
1 Parent(s): 5187edf

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +6 -1
app.py CHANGED
@@ -27,7 +27,7 @@ def process_pdf(doc, diff_line):
27
  data_page = []
28
 
29
  for page_num, page in enumerate(doc.pages()):
30
- for block in page.get_text("blocks"):
31
  for line in block["lines"]:
32
  temp = []
33
  data_page.append(page_num)
@@ -94,6 +94,11 @@ def process_pdf(doc, diff_line):
94
  return pd.DataFrame(data)
95
 
96
  def highlight_pdf(doc, delete_df):
 
 
 
 
 
97
  for i in range(delete_df.shape[0]):
98
  norm_y1 = delete_df["y1_coor"].iloc[i]
99
  norm_y2 = delete_df["y2_coor"].iloc[i]
 
27
  data_page = []
28
 
29
  for page_num, page in enumerate(doc.pages()):
30
+ for block in page["blocks"]:
31
  for line in block["lines"]:
32
  temp = []
33
  data_page.append(page_num)
 
94
  return pd.DataFrame(data)
95
 
96
  def highlight_pdf(doc, delete_df):
97
+ first_page = doc[0]
98
+ rect = first_page.rect
99
+ width = rect.width
100
+ height = rect.height
101
+
102
  for i in range(delete_df.shape[0]):
103
  norm_y1 = delete_df["y1_coor"].iloc[i]
104
  norm_y2 = delete_df["y2_coor"].iloc[i]