Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,590 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import io
|
2 |
+
import os
|
3 |
+
import re
|
4 |
+
import glob
|
5 |
+
import textwrap
|
6 |
+
from datetime import datetime
|
7 |
+
from pathlib import Path
|
8 |
+
import streamlit as st
|
9 |
+
import pandas as pd
|
10 |
+
from PIL import Image
|
11 |
+
from reportlab.pdfgen import canvas
|
12 |
+
from reportlab.lib.pagesizes import letter, A4
|
13 |
+
from reportlab.lib.utils import ImageReader
|
14 |
+
from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, Table, TableStyle
|
15 |
+
from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
|
16 |
+
from reportlab.lib import colors
|
17 |
+
from reportlab.pdfbase import pdfmetrics
|
18 |
+
from reportlab.pdfbase.ttfonts import TTFont
|
19 |
+
import mistune
|
20 |
+
from gtts import gTTS
|
21 |
+
import fitz
|
22 |
+
import edge_tts
|
23 |
+
import asyncio
|
24 |
+
import base64
|
25 |
+
from urllib.parse import quote
|
26 |
+
|
27 |
+
# Page config
|
28 |
+
st.set_page_config(page_title="PDF & Code Interpreter", layout="wide", page_icon="๐")
|
29 |
+
|
30 |
+
def delete_asset(path):
|
31 |
+
try:
|
32 |
+
os.remove(path)
|
33 |
+
except Exception as e:
|
34 |
+
st.error(f"Error deleting file: {e}")
|
35 |
+
st.rerun()
|
36 |
+
|
37 |
+
async def generate_audio(text, voice, filename):
|
38 |
+
communicate = edge_tts.Communicate(text, voice)
|
39 |
+
await communicate.save(filename)
|
40 |
+
return filename
|
41 |
+
|
42 |
+
def clean_for_speech(text):
|
43 |
+
text = text.replace("#", "")
|
44 |
+
emoji_pattern = re.compile(
|
45 |
+
r"[\U0001F300-\U0001F5FF"
|
46 |
+
r"\U0001F600-\U0001F64F"
|
47 |
+
r"\U0001F680-\U0001F6FF"
|
48 |
+
r"\U0001F700-\U0001F77F"
|
49 |
+
r"\U0001F780-\U0001F7FF"
|
50 |
+
r"\U0001F800-\U0001F8FF"
|
51 |
+
r"\U0001F900-\U0001F9FF"
|
52 |
+
r"\U0001FA00-\U0001FA6F"
|
53 |
+
r"\U0001FA70-\U0001FAFF"
|
54 |
+
r"\u2600-\u26FF"
|
55 |
+
r"\u2700-\u27BF]+", flags=re.UNICODE)
|
56 |
+
return emoji_pattern.sub('', text)
|
57 |
+
|
58 |
+
def detect_and_convert_links(text):
|
59 |
+
md_link_pattern = re.compile(r'\[(.*?)\]\((https?://[^\s\[\]()<>{}]+)\)')
|
60 |
+
text = md_link_pattern.sub(r'<a href="\2" color="blue">\1</a>', text)
|
61 |
+
url_pattern = re.compile(r'(?<!href=")(https?://[^\s<>{}]+)', re.IGNORECASE)
|
62 |
+
text = url_pattern.sub(r'<a href="\1" color="blue">\1</a>', text)
|
63 |
+
return text
|
64 |
+
|
65 |
+
def apply_emoji_font(text, emoji_font):
|
66 |
+
tag_pattern = re.compile(r'(<[^>]+>)')
|
67 |
+
segments = tag_pattern.split(text)
|
68 |
+
result = []
|
69 |
+
emoji_pattern = re.compile(
|
70 |
+
r"([\U0001F300-\U0001F5FF"
|
71 |
+
r"\U0001F600-\U0001F64F"
|
72 |
+
r"\U0001F680-\U0001F6FF"
|
73 |
+
r"\U0001F700-\U0001F77F"
|
74 |
+
r"\U0001F780-\U0001F7FF"
|
75 |
+
r"\U0001F800-\U0001F8FF"
|
76 |
+
r"\U0001F900-\U0001F9FF"
|
77 |
+
r"\U0001FAD0-\U0001FAD9"
|
78 |
+
r"\U0001FA00-\U0001FA6F"
|
79 |
+
r"\U0001FA70-\U0001FAFF"
|
80 |
+
r"\u2600-\u26FF"
|
81 |
+
r"\u2700-\u27BF]+)"
|
82 |
+
)
|
83 |
+
def replace_emoji(match):
|
84 |
+
emoji = match.group(1)
|
85 |
+
return f'<font face="{emoji_font}">{emoji}</font>'
|
86 |
+
for segment in segments:
|
87 |
+
if tag_pattern.match(segment):
|
88 |
+
result.append(segment)
|
89 |
+
else:
|
90 |
+
parts = []
|
91 |
+
last_pos = 0
|
92 |
+
for match in emoji_pattern.finditer(segment):
|
93 |
+
start, end = match.span()
|
94 |
+
if last_pos < start:
|
95 |
+
parts.append(f'<font face="DejaVuSans">{segment[last_pos:start]}</font>')
|
96 |
+
parts.append(replace_emoji(match))
|
97 |
+
last_pos = end
|
98 |
+
if last_pos < len(segment):
|
99 |
+
parts.append(f'<font face="DejaVuSans">{segment[last_pos:]}</font>')
|
100 |
+
result.append(''.join(parts))
|
101 |
+
return ''.join(result)
|
102 |
+
|
103 |
+
def markdown_to_pdf_content(markdown_text, add_space_before_numbered, headings_to_fonts):
|
104 |
+
lines = markdown_text.strip().split('\n')
|
105 |
+
pdf_content = []
|
106 |
+
number_pattern = re.compile(r'^\d+(\.\d+)*\.\s')
|
107 |
+
heading_pattern = re.compile(r'^(#{1,4})\s+(.+)$')
|
108 |
+
first_numbered_seen = False
|
109 |
+
for line in lines:
|
110 |
+
line = line.strip()
|
111 |
+
if not line:
|
112 |
+
continue
|
113 |
+
if headings_to_fonts and line.startswith('#'):
|
114 |
+
heading_match = heading_pattern.match(line)
|
115 |
+
if heading_match:
|
116 |
+
level = len(heading_match.group(1))
|
117 |
+
heading_text = heading_match.group(2).strip()
|
118 |
+
formatted_heading = f"<h{level}>{heading_text}</h{level}>"
|
119 |
+
pdf_content.append(formatted_heading)
|
120 |
+
continue
|
121 |
+
is_numbered_line = number_pattern.match(line) is not None
|
122 |
+
if add_space_before_numbered and is_numbered_line:
|
123 |
+
if first_numbered_seen and not line.startswith("1."):
|
124 |
+
pdf_content.append("")
|
125 |
+
if not first_numbered_seen:
|
126 |
+
first_numbered_seen = True
|
127 |
+
line = detect_and_convert_links(line)
|
128 |
+
line = re.sub(r'\*\*(.+?)\*\*', r'<b>\1</b>', line)
|
129 |
+
line = re.sub(r'\*([^*]+?)\*', r'<b>\1</b>', line)
|
130 |
+
pdf_content.append(line)
|
131 |
+
total_lines = len(pdf_content)
|
132 |
+
return pdf_content, total_lines
|
133 |
+
|
134 |
+
def create_pdf(markdown_texts, image_files, base_font_size=14, num_columns=2, add_space_before_numbered=True, headings_to_fonts=True, doc_title="Combined Document"):
|
135 |
+
if not markdown_texts and not image_files:
|
136 |
+
return None
|
137 |
+
buffer = io.BytesIO()
|
138 |
+
page_width = A4[0] * 2
|
139 |
+
page_height = A4[1]
|
140 |
+
doc = SimpleDocTemplate(
|
141 |
+
buffer,
|
142 |
+
pagesize=(page_width, page_height),
|
143 |
+
leftMargin=36,
|
144 |
+
rightMargin=36,
|
145 |
+
topMargin=36,
|
146 |
+
bottomMargin=36,
|
147 |
+
title=doc_title
|
148 |
+
)
|
149 |
+
styles = getSampleStyleSheet()
|
150 |
+
spacer_height = 10
|
151 |
+
try:
|
152 |
+
pdfmetrics.registerFont(TTFont("DejaVuSans", "DejaVuSans.ttf"))
|
153 |
+
pdfmetrics.registerFont(TTFont("NotoEmoji-Bold", "NotoEmoji-Bold.ttf"))
|
154 |
+
except Exception as e:
|
155 |
+
st.error(f"Font registration error: {e}")
|
156 |
+
return None
|
157 |
+
story = []
|
158 |
+
for markdown_text in markdown_texts:
|
159 |
+
pdf_content, total_lines = markdown_to_pdf_content(markdown_text, add_space_before_numbered, headings_to_fonts)
|
160 |
+
total_chars = sum(len(line) for line in pdf_content)
|
161 |
+
hierarchy_weight = sum(1.5 if line.startswith("<b>") else 1 for line in pdf_content)
|
162 |
+
longest_line_words = max(len(line.split()) for line in pdf_content) if pdf_content else 0
|
163 |
+
content_density = total_lines * hierarchy_weight + total_chars / 50
|
164 |
+
usable_height = page_height - 72 - spacer_height
|
165 |
+
usable_width = page_width - 72
|
166 |
+
avg_line_chars = total_chars / total_lines if total_lines > 0 else 50
|
167 |
+
col_width = usable_width / num_columns
|
168 |
+
min_font_size = 5
|
169 |
+
max_font_size = 16
|
170 |
+
lines_per_col = total_lines / num_columns if num_columns > 0 else total_lines
|
171 |
+
target_height_per_line = usable_height / lines_per_col if lines_per_col > 0 else usable_height
|
172 |
+
estimated_font_size = int(target_height_per_line / 1.5)
|
173 |
+
adjusted_font_size = max(min_font_size, min(max_font_size, estimated_font_size))
|
174 |
+
if avg_line_chars > col_width / adjusted_font_size * 10:
|
175 |
+
adjusted_font_size = int(col_width / (avg_line_chars / 10))
|
176 |
+
adjusted_font_size = max(min_font_size, adjusted_font_size)
|
177 |
+
if longest_line_words > 17 or lines_per_col > 20:
|
178 |
+
font_scale = min(17 / max(longest_line_words, 17), 60 / max(lines_per_col, 20))
|
179 |
+
adjusted_font_size = max(min_font_size, int(base_font_size * font_scale))
|
180 |
+
item_style = ParagraphStyle(
|
181 |
+
'ItemStyle', parent=styles['Normal'], fontName="DejaVuSans",
|
182 |
+
fontSize=adjusted_font_size, leading=adjusted_font_size * 1.15, spaceAfter=1,
|
183 |
+
linkUnderline=True
|
184 |
+
)
|
185 |
+
numbered_bold_style = ParagraphStyle(
|
186 |
+
'NumberedBoldStyle', parent=styles['Normal'], fontName="NotoEmoji-Bold",
|
187 |
+
fontSize=adjusted_font_size, leading=adjusted_font_size * 1.15, spaceAfter=1,
|
188 |
+
linkUnderline=True
|
189 |
+
)
|
190 |
+
section_style = ParagraphStyle(
|
191 |
+
'SectionStyle', parent=styles['Heading2'], fontName="DejaVuSans",
|
192 |
+
textColor=colors.darkblue, fontSize=adjusted_font_size * 1.1, leading=adjusted_font_size * 1.32, spaceAfter=2,
|
193 |
+
linkUnderline=True
|
194 |
+
)
|
195 |
+
columns = [[] for _ in range(num_columns)]
|
196 |
+
lines_per_column = total_lines / num_columns if num_columns > 0 else total_lines
|
197 |
+
current_line_count = 0
|
198 |
+
current_column = 0
|
199 |
+
number_pattern = re.compile(r'^\d+(\.\d+)*\.\s')
|
200 |
+
for item in pdf_content:
|
201 |
+
if current_line_count >= lines_per_column and current_column < num_columns - 1:
|
202 |
+
current_column += 1
|
203 |
+
current_line_count = 0
|
204 |
+
columns[current_column].append(item)
|
205 |
+
current_line_count += 1
|
206 |
+
column_cells = [[] for _ in range(num_columns)]
|
207 |
+
for col_idx, column in enumerate(columns):
|
208 |
+
for item in column:
|
209 |
+
if isinstance(item, str):
|
210 |
+
heading_match = re.match(r'<h(\d)>(.*?)</h\1>', item) if headings_to_fonts else None
|
211 |
+
if heading_match:
|
212 |
+
level = int(heading_match.group(1))
|
213 |
+
heading_text = heading_match.group(2)
|
214 |
+
heading_style = ParagraphStyle(
|
215 |
+
f'Heading{level}Style',
|
216 |
+
parent=styles['Heading1'],
|
217 |
+
fontName="DejaVuSans",
|
218 |
+
textColor=colors.darkblue if level == 1 else (colors.black if level > 2 else colors.blue),
|
219 |
+
fontSize=adjusted_font_size * (1.6 - (level-1)*0.15),
|
220 |
+
leading=adjusted_font_size * (1.8 - (level-1)*0.15),
|
221 |
+
spaceAfter=4 - (level-1),
|
222 |
+
spaceBefore=6 - (level-1),
|
223 |
+
linkUnderline=True
|
224 |
+
)
|
225 |
+
column_cells[col_idx].append(Paragraph(apply_emoji_font(heading_text, "NotoEmoji-Bold"), heading_style))
|
226 |
+
elif item.startswith("<b>") and item.endswith("</b>"):
|
227 |
+
content = item[3:-4].strip()
|
228 |
+
if number_pattern.match(content):
|
229 |
+
column_cells[col_idx].append(Paragraph(apply_emoji_font(content, "NotoEmoji-Bold"), numbered_bold_style))
|
230 |
+
else:
|
231 |
+
column_cells[col_idx].append(Paragraph(apply_emoji_font(content, "NotoEmoji-Bold"), section_style))
|
232 |
+
else:
|
233 |
+
column_cells[col_idx].append(Paragraph(apply_emoji_font(item, "NotoEmoji-Bold"), item_style))
|
234 |
+
else:
|
235 |
+
column_cells[col_idx].append(Paragraph(apply_emoji_font(str(item), "NotoEmoji-Bold"), item_style))
|
236 |
+
max_cells = max(len(cells) for cells in column_cells) if column_cells else 0
|
237 |
+
for cells in column_cells:
|
238 |
+
cells.extend([Paragraph("", item_style)] * (max_cells - len(cells)))
|
239 |
+
table_data = list(zip(*column_cells)) if column_cells else [[]]
|
240 |
+
table = Table(table_data, colWidths=[col_width] * num_columns, hAlign='CENTER')
|
241 |
+
table.setStyle(TableStyle([
|
242 |
+
('VALIGN', (0, 0), (-1, -1), 'TOP'),
|
243 |
+
('ALIGN', (0, 0), (-1, -1), 'LEFT'),
|
244 |
+
('BACKGROUND', (0, 0), (-1, -1), colors.white),
|
245 |
+
('GRID', (0, 0), (-1, -1), 0, colors.white),
|
246 |
+
('LINEAFTER', (0, 0), (num_columns-1, -1), 0.5, colors.grey),
|
247 |
+
('LEFTPADDING', (0, 0), (-1, -1), 2),
|
248 |
+
('RIGHTPADDING', (0, 0), (-1, -1), 2),
|
249 |
+
('TOPPADDING', (0, 0), (-1, -1), 1),
|
250 |
+
('BOTTOMPADDING', (0, 0), (-1, -1), 1),
|
251 |
+
]))
|
252 |
+
story.append(Spacer(1, spacer_height))
|
253 |
+
story.append(table)
|
254 |
+
story.append(Spacer(1, spacer_height * 2))
|
255 |
+
for img_path in image_files:
|
256 |
+
try:
|
257 |
+
img = Image.open(img_path)
|
258 |
+
img_width, img_height = img.size
|
259 |
+
page_width, page_height = A4
|
260 |
+
scale = min((page_width - 40) / img_width, (page_height - 40) / img_height)
|
261 |
+
new_width = img_width * scale
|
262 |
+
new_height = img_height * scale
|
263 |
+
x = (page_width - new_width) / 2
|
264 |
+
y = (page_height - new_height) / 2
|
265 |
+
buffer_img = io.BytesIO()
|
266 |
+
c = canvas.Canvas(buffer_img, pagesize=A4)
|
267 |
+
c.drawImage(img_path, x, y, new_width, new_height)
|
268 |
+
c.showPage()
|
269 |
+
c.save()
|
270 |
+
buffer_img.seek(0)
|
271 |
+
img_pdf = fitz.open(stream=buffer_img, filetype="pdf")
|
272 |
+
story.append(fitz.open(stream=buffer_img, filetype="pdf").load_page(0))
|
273 |
+
except Exception as e:
|
274 |
+
st.warning(f"Could not process image {img_path}: {e}")
|
275 |
+
continue
|
276 |
+
doc.build(story)
|
277 |
+
buffer.seek(0)
|
278 |
+
return buffer.getvalue()
|
279 |
+
|
280 |
+
def pdf_to_image(pdf_bytes):
|
281 |
+
if pdf_bytes is None:
|
282 |
+
return None
|
283 |
+
try:
|
284 |
+
doc = fitz.open(stream=pdf_bytes, filetype="pdf")
|
285 |
+
images = []
|
286 |
+
for page in doc:
|
287 |
+
pix = page.get_pixmap(matrix=fitz.Matrix(2.0, 2.0))
|
288 |
+
img = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
|
289 |
+
images.append(img)
|
290 |
+
doc.close()
|
291 |
+
return images
|
292 |
+
except Exception as e:
|
293 |
+
st.error(f"Failed to render PDF preview: {e}")
|
294 |
+
return None
|
295 |
+
|
296 |
+
def get_video_html(video_path, width="100%"):
|
297 |
+
try:
|
298 |
+
video_url = f"data:video/mp4;base64,{base64.b64encode(open(video_path, 'rb').read()).decode()}"
|
299 |
+
return f'''
|
300 |
+
<video width="{width}" controls autoplay muted loop>
|
301 |
+
<source src="{video_url}" type="video/mp4">
|
302 |
+
Your browser does not support the video tag.
|
303 |
+
</video>
|
304 |
+
'''
|
305 |
+
except Exception as e:
|
306 |
+
st.warning(f"Could not load video {video_path}: {e}")
|
307 |
+
return ""
|
308 |
+
|
309 |
+
def display_glossary_entity(k):
|
310 |
+
search_urls = {
|
311 |
+
"๐๐ArXiv": lambda k: f"https://arxiv.org/search/?query={quote(k)}&searchtype=all",
|
312 |
+
"๐": lambda k: f"https://en.wikipedia.org/wiki/{quote(k)}",
|
313 |
+
"๐": lambda k: f"https://www.google.com/search?q={quote(k)}",
|
314 |
+
"๐ฅ": lambda k: f"https://www.youtube.com/results?search_query={quote(k)}",
|
315 |
+
}
|
316 |
+
links_md = ' '.join([f"[{emoji}]({url(k)})" for emoji, url in search_urls.items()])
|
317 |
+
st.markdown(f"**{k}** <small>{links_md}</small>", unsafe_allow_html=True)
|
318 |
+
|
319 |
+
# Tabs setup
|
320 |
+
tab1, tab2 = st.tabs(["๐ PDF Composer", "๐งช Code Interpreter"])
|
321 |
+
|
322 |
+
with tab1:
|
323 |
+
st.header("๐ PDF Composer & Voice Generator ๐")
|
324 |
+
# Sidebar PDF text settings
|
325 |
+
columns = st.sidebar.slider("Text columns", 1, 3, 2)
|
326 |
+
font_family = st.sidebar.selectbox("Font", ["Helvetica", "Times-Roman", "Courier", "DejaVuSans"])
|
327 |
+
font_size = st.sidebar.slider("Font size", 6, 24, 14)
|
328 |
+
# Markdown input
|
329 |
+
md_file = st.file_uploader("Upload Markdown (.md)", type=["md"])
|
330 |
+
if md_file:
|
331 |
+
md_text = md_file.getvalue().decode("utf-8")
|
332 |
+
stem = Path(md_file.name).stem
|
333 |
+
else:
|
334 |
+
md_text = st.text_area("Or enter markdown text directly", height=200)
|
335 |
+
stem = datetime.now().strftime('%Y%m%d_%H%M%S')
|
336 |
+
# Convert Markdown to plain text
|
337 |
+
renderer = mistune.HTMLRenderer()
|
338 |
+
markdown = mistune.create_markdown(renderer=renderer)
|
339 |
+
html = markdown(md_text or "")
|
340 |
+
plain_text = re.sub(r'<[^>]+>', '', html)
|
341 |
+
# Voice settings
|
342 |
+
languages = {"English (US)": "en", "English (UK)": "en-uk", "Spanish": "es"}
|
343 |
+
voice_choice = st.selectbox("Voice Language", list(languages.keys()))
|
344 |
+
voice_lang = languages[voice_choice]
|
345 |
+
slow = st.checkbox("Slow Speech")
|
346 |
+
VOICES = ["en-US-AriaNeural", "en-US-JennyNeural", "en-GB-SoniaNeural", "en-US-GuyNeural", "en-US-AnaNeural"]
|
347 |
+
selected_voice = st.selectbox("Select Voice for TTS", options=VOICES, index=0)
|
348 |
+
if st.button("๐ Generate & Download Voice MP3 from Text"):
|
349 |
+
if plain_text.strip():
|
350 |
+
voice_file = f"{stem}_{selected_voice}.mp3"
|
351 |
+
try:
|
352 |
+
cleaned_text = clean_for_speech(plain_text)
|
353 |
+
audio_file = asyncio.run(generate_audio(cleaned_text, selected_voice, voice_file))
|
354 |
+
st.audio(audio_file)
|
355 |
+
with open(audio_file, 'rb') as mp3:
|
356 |
+
st.download_button("๐ฅ Download MP3", data=mp3, file_name=voice_file, mime="audio/mpeg")
|
357 |
+
except Exception as e:
|
358 |
+
st.error(f"Error generating voice: {e}")
|
359 |
+
else:
|
360 |
+
st.warning("No text to generate voice from.")
|
361 |
+
# Image uploads and ordering
|
362 |
+
imgs = st.file_uploader("Upload Images for PDF", type=["png", "jpg", "jpeg"], accept_multiple_files=True)
|
363 |
+
ordered_images = []
|
364 |
+
if imgs:
|
365 |
+
df_imgs = pd.DataFrame([{"name": f.name, "order": i} for i, f in enumerate(imgs)])
|
366 |
+
edited = st.data_editor(df_imgs, use_container_width=True, num_rows="dynamic")
|
367 |
+
for _, row in edited.sort_values("order").iterrows():
|
368 |
+
for f in imgs:
|
369 |
+
if f.name == row['name']:
|
370 |
+
ordered_images.append(f)
|
371 |
+
break
|
372 |
+
if st.button("๐๏ธ Generate PDF with Markdown & Images"):
|
373 |
+
if not plain_text.strip() and not ordered_images:
|
374 |
+
st.warning("Please provide some text or upload images to generate a PDF.")
|
375 |
+
else:
|
376 |
+
buf = io.BytesIO()
|
377 |
+
c = canvas.Canvas(buf)
|
378 |
+
if plain_text.strip():
|
379 |
+
page_w, page_h = letter
|
380 |
+
margin = 40
|
381 |
+
gutter = 20
|
382 |
+
col_w = (page_w - 2*margin - (columns-1)*gutter) / columns
|
383 |
+
c.setFont(font_family, font_size)
|
384 |
+
line_height = font_size * 1.2
|
385 |
+
col = 0
|
386 |
+
x = margin
|
387 |
+
y = page_h - margin
|
388 |
+
avg_char_width = font_size * 0.6
|
389 |
+
wrap_width = int(col_w / avg_char_width) if avg_char_width > 0 else 100
|
390 |
+
for paragraph in plain_text.split("\n"):
|
391 |
+
if not paragraph.strip():
|
392 |
+
y -= line_height
|
393 |
+
if y < margin:
|
394 |
+
col += 1
|
395 |
+
if col >= columns:
|
396 |
+
c.showPage()
|
397 |
+
c.setFont(font_family, font_size)
|
398 |
+
col = 0
|
399 |
+
x = margin + col*(col_w+gutter)
|
400 |
+
y = page_h - margin
|
401 |
+
continue
|
402 |
+
for line in textwrap.wrap(paragraph, wrap_width):
|
403 |
+
if y < margin:
|
404 |
+
col += 1
|
405 |
+
if col >= columns:
|
406 |
+
c.showPage()
|
407 |
+
c.setFont(font_family, font_size)
|
408 |
+
col = 0
|
409 |
+
x = margin + col*(col_w+gutter)
|
410 |
+
y = page_h - margin
|
411 |
+
c.drawString(x, y, line)
|
412 |
+
y -= line_height
|
413 |
+
y -= line_height
|
414 |
+
for img_f in ordered_images:
|
415 |
+
try:
|
416 |
+
img = Image.open(img_f)
|
417 |
+
w, h = img.size
|
418 |
+
c.showPage()
|
419 |
+
c.setPageSize((w, h))
|
420 |
+
c.drawImage(ImageReader(img), 0, 0, w, h, preserveAspectRatio=False)
|
421 |
+
except Exception as e:
|
422 |
+
st.warning(f"Could not process image {img_f.name}: {e}")
|
423 |
+
continue
|
424 |
+
c.save()
|
425 |
+
buf.seek(0)
|
426 |
+
pdf_name = f"{stem}.pdf"
|
427 |
+
st.download_button("โฌ๏ธ Download PDF", data=buf, file_name=pdf_name, mime="application/pdf")
|
428 |
+
st.markdown("---")
|
429 |
+
st.subheader("๐ Available Assets")
|
430 |
+
all_assets = glob.glob("*.*")
|
431 |
+
excluded_extensions = ['.py', '.ttf', '.txt']
|
432 |
+
excluded_files = ['README.md', 'index.html']
|
433 |
+
assets = sorted([
|
434 |
+
a for a in all_assets
|
435 |
+
if not (a.lower().endswith(tuple(excluded_extensions)) or a in excluded_files)
|
436 |
+
and a.lower().endswith(('.md', '.png', '.jpg', '.jpeg'))
|
437 |
+
])
|
438 |
+
if 'selected_assets' not in st.session_state:
|
439 |
+
st.session_state.selected_assets = []
|
440 |
+
if not assets:
|
441 |
+
st.info("No available assets found.")
|
442 |
+
else:
|
443 |
+
for a in assets:
|
444 |
+
ext = a.split('.')[-1].lower()
|
445 |
+
cols = st.columns([1, 3, 1, 1])
|
446 |
+
with cols[0]:
|
447 |
+
is_selected = st.checkbox("", key=f"select_{a}", value=a in st.session_state.selected_assets)
|
448 |
+
if is_selected and a not in st.session_state.selected_assets:
|
449 |
+
st.session_state.selected_assets.append(a)
|
450 |
+
elif not is_selected and a in st.session_state.selected_assets:
|
451 |
+
st.session_state.selected_assets.remove(a)
|
452 |
+
cols[1].write(a)
|
453 |
+
try:
|
454 |
+
if ext == 'md':
|
455 |
+
with open(a, 'r', encoding='utf-8') as f:
|
456 |
+
cols[2].download_button("๐ฅ", data=f.read(), file_name=a, mime="text/markdown")
|
457 |
+
elif ext in ['png', 'jpg', 'jpeg']:
|
458 |
+
with open(a, 'rb') as img_file:
|
459 |
+
cols[2].download_button("โฌ๏ธ", data=img_file, file_name=a, mime=f"image/{ext}")
|
460 |
+
cols[3].button("๐๏ธ", key=f"del_{a}", on_click=delete_asset, args=(a,))
|
461 |
+
except Exception as e:
|
462 |
+
cols[3].error(f"Error handling file {a}: {e}")
|
463 |
+
if st.button("๐ Generate PDF from Selected Assets"):
|
464 |
+
if not st.session_state.selected_assets:
|
465 |
+
st.warning("Please select at least one asset to generate a PDF.")
|
466 |
+
else:
|
467 |
+
markdown_texts = []
|
468 |
+
image_files = []
|
469 |
+
for a in st.session_state.selected_assets:
|
470 |
+
ext = a.split('.')[-1].lower()
|
471 |
+
if ext == 'md':
|
472 |
+
with open(a, 'r', encoding='utf-8') as f:
|
473 |
+
markdown_texts.append(f.read())
|
474 |
+
elif ext in ['png', 'jpg', 'jpeg']:
|
475 |
+
image_files.append(a)
|
476 |
+
with st.spinner("Generating PDF from selected assets..."):
|
477 |
+
pdf_bytes = create_pdf(
|
478 |
+
markdown_texts=markdown_texts,
|
479 |
+
image_files=image_files,
|
480 |
+
base_font_size=14,
|
481 |
+
num_columns=2,
|
482 |
+
add_space_before_numbered=True,
|
483 |
+
headings_to_fonts=True,
|
484 |
+
doc_title="Combined_Selected_Assets"
|
485 |
+
)
|
486 |
+
if pdf_bytes:
|
487 |
+
pdf_images = pdf_to_image(pdf_bytes)
|
488 |
+
if pdf_images:
|
489 |
+
st.subheader("Preview of Generated PDF")
|
490 |
+
for i, img in enumerate(pdf_images):
|
491 |
+
st.image(img, caption=f"Page {i+1}", use_container_width=True)
|
492 |
+
prefix = datetime.now().strftime("%Y%m%d_%H%M%S")
|
493 |
+
st.download_button(
|
494 |
+
label="๐พ Download Combined PDF",
|
495 |
+
data=pdf_bytes,
|
496 |
+
file_name=f"{prefix}_combined.pdf",
|
497 |
+
mime="application/pdf"
|
498 |
+
)
|
499 |
+
else:
|
500 |
+
st.error("Failed to generate PDF.")
|
501 |
+
st.markdown("---")
|
502 |
+
st.subheader("๐ผ Image Gallery")
|
503 |
+
image_files = glob.glob("*.png") + glob.glob("*.jpg") + glob.glob("*.jpeg")
|
504 |
+
image_cols = st.slider("Gallery Columns ๐ผ", min_value=1, max_value=15, value=5, key="image_cols")
|
505 |
+
if image_files:
|
506 |
+
cols = st.columns(image_cols)
|
507 |
+
for idx, image_file in enumerate(image_files):
|
508 |
+
with cols[idx % image_cols]:
|
509 |
+
try:
|
510 |
+
img = Image.open(image_file)
|
511 |
+
st.image(img, caption=image_file, use_container_width=True)
|
512 |
+
display_glossary_entity(os.path.splitext(image_file)[0])
|
513 |
+
except Exception as e:
|
514 |
+
st.warning(f"Could not load image {image_file}: {e}")
|
515 |
+
else:
|
516 |
+
st.info("No images found in the current directory.")
|
517 |
+
st.markdown("---")
|
518 |
+
st.subheader("๐ฅ Video Gallery")
|
519 |
+
video_files = glob.glob("*.mp4")
|
520 |
+
video_cols = st.slider("Gallery Columns ๐ฌ", min_value=1, max_value=5, value=3, key="video_cols")
|
521 |
+
if video_files:
|
522 |
+
cols = st.columns(video_cols)
|
523 |
+
for idx, video_file in enumerate(video_files):
|
524 |
+
with cols[idx % video_cols]:
|
525 |
+
st.markdown(get_video_html(video_file, width="100%"), unsafe_allow_html=True)
|
526 |
+
display_glossary_entity(os.path.splitext(video_file)[0])
|
527 |
+
else:
|
528 |
+
st.info("No videos found in the current directory.")
|
529 |
+
|
530 |
+
with tab2:
|
531 |
+
st.header("๐งช Python Code Executor & Demo")
|
532 |
+
import io, sys
|
533 |
+
from contextlib import redirect_stdout
|
534 |
+
DEFAULT_CODE = '''import streamlit as st
|
535 |
+
import random
|
536 |
+
st.title("๐ Demo App")
|
537 |
+
st.markdown("Random number and color demo")
|
538 |
+
col1, col2 = st.columns(2)
|
539 |
+
with col1:
|
540 |
+
num = st.number_input("Number:", 1, 100, 10)
|
541 |
+
mul = st.slider("Multiplier:", 1, 10, 2)
|
542 |
+
if st.button("Calc"):
|
543 |
+
st.write(num * mul)
|
544 |
+
with col2:
|
545 |
+
color = st.color_picker("Pick color","#ff0000")
|
546 |
+
st.markdown(f'<div style="background:{color};padding:10px;">Color</div>', unsafe_allow_html=True)
|
547 |
+
'''
|
548 |
+
def extract_python_code(md: str) -> list:
|
549 |
+
return re.findall(r"```python\s*(.*?)```", md, re.DOTALL)
|
550 |
+
def execute_code(code: str) -> tuple:
|
551 |
+
buf = io.StringIO(); local_vars = {}
|
552 |
+
try:
|
553 |
+
with redirect_stdout(buf):
|
554 |
+
exec(code, {}, local_vars)
|
555 |
+
return buf.getvalue(), None
|
556 |
+
except Exception as e:
|
557 |
+
return None, str(e)
|
558 |
+
up = st.file_uploader("Upload .py or .md", type=['py', 'md'])
|
559 |
+
if 'code' not in st.session_state:
|
560 |
+
st.session_state.code = DEFAULT_CODE
|
561 |
+
if up:
|
562 |
+
text = up.getvalue().decode()
|
563 |
+
if up.type == 'text/markdown':
|
564 |
+
codes = extract_python_code(text)
|
565 |
+
if codes:
|
566 |
+
st.session_state.code = codes[0].strip()
|
567 |
+
else:
|
568 |
+
st.warning("No Python code block found in the markdown file.")
|
569 |
+
st.session_state.code = ''
|
570 |
+
else:
|
571 |
+
st.session_state.code = text.strip()
|
572 |
+
st.code(st.session_state.code, language='python')
|
573 |
+
else:
|
574 |
+
st.session_state.code = st.text_area("๐ป Code Editor", value=st.session_state.code, height=400)
|
575 |
+
c1, c2 = st.columns([1, 1])
|
576 |
+
if c1.button("โถ๏ธ Run Code"):
|
577 |
+
if st.session_state.code.strip():
|
578 |
+
out, err = execute_code(st.session_state.code)
|
579 |
+
if err:
|
580 |
+
st.error(f"Execution Error:\n{err}")
|
581 |
+
elif out:
|
582 |
+
st.subheader("Output:")
|
583 |
+
st.code(out)
|
584 |
+
else:
|
585 |
+
st.success("Executed with no standard output.")
|
586 |
+
else:
|
587 |
+
st.warning("No code to run.")
|
588 |
+
if c2.button("๐๏ธ Clear Code"):
|
589 |
+
st.session_state.code = ''
|
590 |
+
st.rerun()
|