exchange math delimiters (#1)

- exchange math delimiters (3dd4fff646571b17c82a9283b8505662ffcaf546)


Co-authored-by: Lukas Blecher <[email protected]>

app.py

@@ -3,6 +3,7 @@ import subprocess
 import uuid
 import os
 import requests
+import re
 
 
 def get_pdf(pdf_link):
@@ -31,7 +32,8 @@ def nougat_ocr(file_name):
       #'--out', unique_filename,
       '--out', 'output',
       'pdf', f'{file_name}',
-      '--checkpoint', 'nougat'
+      '--checkpoint', 'nougat',
+      '--markdown'
   ]
 
   # Run the command and capture its output
@@ -64,6 +66,8 @@ def predict(pdf_file, pdf_link):
   file_name = file_name.split('/')[-1][:-4]
   with open(f'output/{file_name}.mmd', 'r') as file:
       content = file.read()
+  # switch math delimiters
+  content = content.replace(r'\(', '$').replace(r'\)', '$').replace(r'\[', '$$').replace(r'\]', '$$')
   return content
 
 
@@ -76,7 +80,8 @@ def nougat_ocr1(file_name):
       'nougat',
       '--out', 'output',
       'pdf', f'{file_name}',
-      '--checkpoint', 'nougat'
+      '--checkpoint', 'nougat',
+      '--markdown'
   ]
 
   # Run the command and get .mmd file in an output folder