Spaces:
Paused
Paused
Update app.py
Browse files
app.py
CHANGED
|
@@ -4,6 +4,9 @@ import numpy as np
|
|
| 4 |
import os
|
| 5 |
import pickle
|
| 6 |
import torch
|
|
|
|
|
|
|
|
|
|
| 7 |
from grobidmonkey import reader
|
| 8 |
|
| 9 |
from transformers import pipeline
|
|
@@ -139,4 +142,39 @@ with col3:
|
|
| 139 |
# Display HTML box
|
| 140 |
st.markdown(st.session_state.current_text)
|
| 141 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 142 |
|
|
|
|
| 4 |
import os
|
| 5 |
import pickle
|
| 6 |
import torch
|
| 7 |
+
import markdown
|
| 8 |
+
import pdfkit
|
| 9 |
+
import io
|
| 10 |
from grobidmonkey import reader
|
| 11 |
|
| 12 |
from transformers import pipeline
|
|
|
|
| 142 |
# Display HTML box
|
| 143 |
st.markdown(st.session_state.current_text)
|
| 144 |
|
| 145 |
+
def render_markdown_to_html(markdown_str):
|
| 146 |
+
return markdown.markdown(markdown_str)
|
| 147 |
+
|
| 148 |
+
def create_pdf_from_markdown_strings(markdown_strings, output_file):
|
| 149 |
+
html_pages = [render_markdown_to_html(md) for md in markdown_strings]
|
| 150 |
+
|
| 151 |
+
# Combine HTML content with page breaks and add a style section for font size and margins
|
| 152 |
+
combined_html = '''
|
| 153 |
+
<html>
|
| 154 |
+
<head>
|
| 155 |
+
<style>
|
| 156 |
+
.page {
|
| 157 |
+
font-size: 16pt; /* Adjust the font size as needed */
|
| 158 |
+
margin: 20mm; /* Set margins for top, right, bottom, and left */
|
| 159 |
+
}
|
| 160 |
+
</style>
|
| 161 |
+
</head>
|
| 162 |
+
<body>
|
| 163 |
+
'''
|
| 164 |
+
for i, page in enumerate(html_pages):
|
| 165 |
+
combined_html += f'<div class="page">{page}</div>'
|
| 166 |
+
if i < len(html_pages) - 1: # Only add page break after if it's not the last page
|
| 167 |
+
combined_html += '<div style="page-break-after: always;"></div>'
|
| 168 |
+
combined_html += '</body></html>'
|
| 169 |
+
|
| 170 |
+
# PDF options: landscape orientation and page size
|
| 171 |
+
options = {
|
| 172 |
+
'page-size': 'A4',
|
| 173 |
+
'orientation': 'Landscape'
|
| 174 |
+
}
|
| 175 |
+
|
| 176 |
+
# Convert combined HTML to PDF directly into a file
|
| 177 |
+
pdfkit.from_string(combined_html, output_file, options=options)
|
| 178 |
+
|
| 179 |
+
create_pdf_from_markdown_strings(st.session_state.summ_text, '/tmp/output.pdf')
|
| 180 |
|