Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -1,65 +1,21 @@
|
|
1 |
-
import os
|
2 |
-
import io
|
3 |
-
import re
|
4 |
import streamlit as st
|
5 |
-
|
6 |
-
# Must be the very first Streamlit command.
|
7 |
-
st.set_page_config(layout="wide", initial_sidebar_state="collapsed")
|
8 |
-
|
9 |
-
from PIL import Image
|
10 |
-
import fitz # PyMuPDF
|
11 |
-
|
12 |
from reportlab.lib.pagesizes import A4
|
13 |
from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, Table, TableStyle
|
14 |
from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
|
15 |
from reportlab.lib import colors
|
16 |
-
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
"NotoEmoji Medium": "NotoEmoji-Medium.ttf",
|
26 |
-
"NotoEmoji Regular": "NotoEmoji-Regular.ttf",
|
27 |
-
"NotoEmoji SemiBold": "NotoEmoji-SemiBold.ttf"
|
28 |
-
}
|
29 |
-
|
30 |
-
# Sidebar: Let the user choose the desired NotoEmoji font.
|
31 |
-
selected_font_name = st.sidebar.selectbox(
|
32 |
-
"Select NotoEmoji Font",
|
33 |
-
options=list(available_fonts.keys())
|
34 |
-
)
|
35 |
-
selected_font_path = available_fonts[selected_font_name]
|
36 |
-
|
37 |
-
# Register the chosen emoji font with ReportLab.
|
38 |
-
pdfmetrics.registerFont(TTFont(selected_font_name, selected_font_path))
|
39 |
-
|
40 |
-
# ---------------------------------------------------------------
|
41 |
-
# Helper function to wrap emoji characters with a font tag.
|
42 |
-
def apply_emoji_font(text, emoji_font):
|
43 |
-
# This regex attempts to capture many common emoji ranges.
|
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]+)"
|
56 |
-
)
|
57 |
-
# Wrap found emoji with a font tag using the selected emoji font.
|
58 |
-
return emoji_pattern.sub(r'<font face="{}">\1</font>'.format(emoji_font), text)
|
59 |
|
60 |
-
#
|
61 |
-
|
62 |
-
default_markdown = """# Pillow-PyMuPDF-ReportLab - Markdown to PDF One Pager
|
63 |
|
64 |
## Core ML Techniques
|
65 |
1. π **Mixture of Experts (MoE)**
|
@@ -126,9 +82,12 @@ default_markdown = """# Pillow-PyMuPDF-ReportLab - Markdown to PDF One Pager
|
|
126 |
- Documentation synthesis
|
127 |
"""
|
128 |
|
129 |
-
#
|
130 |
-
# Process markdown into a two-column layout for the PDF.
|
131 |
def markdown_to_pdf_content(markdown_text):
|
|
|
|
|
|
|
|
|
132 |
lines = markdown_text.strip().split('\n')
|
133 |
pdf_content = []
|
134 |
in_list_item = False
|
@@ -141,230 +100,627 @@ def markdown_to_pdf_content(markdown_text):
|
|
141 |
continue
|
142 |
|
143 |
if line.startswith('# '):
|
144 |
-
#
|
145 |
pass
|
146 |
elif line.startswith('## '):
|
|
|
147 |
if current_item and sub_items:
|
|
|
148 |
pdf_content.append([current_item, sub_items])
|
149 |
sub_items = []
|
150 |
current_item = None
|
|
|
151 |
section = line.replace('## ', '').strip()
|
152 |
pdf_content.append(f"<b>{section}</b>")
|
153 |
in_list_item = False
|
154 |
elif re.match(r'^\d+\.', line):
|
|
|
155 |
if current_item and sub_items:
|
|
|
156 |
pdf_content.append([current_item, sub_items])
|
157 |
sub_items = []
|
|
|
158 |
current_item = line.strip()
|
159 |
in_list_item = True
|
160 |
elif line.startswith('- ') and in_list_item:
|
|
|
161 |
sub_items.append(line.strip())
|
162 |
else:
|
|
|
163 |
if not in_list_item:
|
164 |
pdf_content.append(line.strip())
|
165 |
|
|
|
166 |
if current_item and sub_items:
|
167 |
pdf_content.append([current_item, sub_items])
|
168 |
|
|
|
169 |
mid_point = len(pdf_content) // 2
|
170 |
left_column = pdf_content[:mid_point]
|
171 |
right_column = pdf_content[mid_point:]
|
172 |
|
173 |
return left_column, right_column
|
174 |
|
175 |
-
#
|
176 |
-
|
177 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
178 |
buffer = io.BytesIO()
|
179 |
-
|
180 |
-
|
181 |
-
|
182 |
-
|
183 |
-
|
184 |
-
|
185 |
-
|
186 |
-
)
|
187 |
|
188 |
-
|
189 |
-
|
190 |
-
|
191 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
192 |
|
193 |
-
|
194 |
-
|
195 |
-
|
196 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
197 |
main_item, sub_items = item
|
198 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
199 |
else:
|
200 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
201 |
|
202 |
-
|
203 |
-
|
|
|
204 |
|
205 |
-
|
206 |
-
|
207 |
-
section_font_size = base_font_size * 1.2
|
208 |
-
title_font_size = min(16, base_font_size * 1.5)
|
209 |
|
210 |
-
|
211 |
-
|
212 |
-
|
213 |
-
|
214 |
-
|
215 |
-
|
216 |
-
|
217 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
218 |
)
|
219 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
220 |
section_style = ParagraphStyle(
|
221 |
'SectionStyle',
|
222 |
parent=styles['Heading2'],
|
223 |
-
fontName="Helvetica-Bold",
|
224 |
textColor=colors.darkblue,
|
225 |
-
|
226 |
-
leading=section_font_size * 1.2,
|
227 |
-
spaceAfter=2
|
228 |
)
|
229 |
|
230 |
item_style = ParagraphStyle(
|
231 |
'ItemStyle',
|
232 |
parent=styles['Normal'],
|
233 |
-
|
234 |
-
|
235 |
-
|
236 |
-
spaceAfter=1
|
237 |
)
|
238 |
|
239 |
subitem_style = ParagraphStyle(
|
240 |
'SubItemStyle',
|
241 |
parent=styles['Normal'],
|
242 |
-
|
243 |
-
|
244 |
-
|
245 |
-
leftIndent=10,
|
246 |
-
spaceAfter=1
|
247 |
)
|
248 |
|
249 |
-
|
250 |
-
story.append(
|
|
|
251 |
|
|
|
252 |
left_cells = []
|
253 |
for item in left_column:
|
254 |
if isinstance(item, str) and item.startswith('<b>'):
|
255 |
-
#
|
256 |
text = item.replace('<b>', '').replace('</b>', '')
|
257 |
-
left_cells.append(Paragraph(
|
258 |
elif isinstance(item, list):
|
|
|
259 |
main_item, sub_items = item
|
260 |
-
left_cells.append(Paragraph(
|
|
|
|
|
261 |
for sub_item in sub_items:
|
262 |
-
left_cells.append(Paragraph(
|
263 |
else:
|
264 |
-
|
|
|
265 |
|
266 |
right_cells = []
|
267 |
for item in right_column:
|
268 |
if isinstance(item, str) and item.startswith('<b>'):
|
|
|
269 |
text = item.replace('<b>', '').replace('</b>', '')
|
270 |
-
right_cells.append(Paragraph(
|
271 |
elif isinstance(item, list):
|
|
|
272 |
main_item, sub_items = item
|
273 |
-
right_cells.append(Paragraph(
|
|
|
|
|
274 |
for sub_item in sub_items:
|
275 |
-
right_cells.append(Paragraph(
|
276 |
else:
|
277 |
-
|
|
|
278 |
|
|
|
279 |
max_cells = max(len(left_cells), len(right_cells))
|
280 |
-
|
281 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
282 |
|
283 |
-
|
284 |
-
col_width = (A4[1] - 72) / 2.0
|
285 |
-
table = Table(table_data, colWidths=[col_width, col_width], hAlign='CENTER')
|
286 |
table.setStyle(TableStyle([
|
287 |
('VALIGN', (0, 0), (-1, -1), 'TOP'),
|
288 |
-
('ALIGN', (0, 0), (
|
|
|
289 |
('BACKGROUND', (0, 0), (-1, -1), colors.white),
|
290 |
-
('GRID', (0, 0), (-1, -1), 0, colors.white),
|
291 |
-
('LINEAFTER', (0, 0), (0, -1),
|
292 |
-
('LEFTPADDING', (0, 0), (-1, -1), 2),
|
293 |
-
('RIGHTPADDING', (0, 0), (-1, -1), 2),
|
294 |
-
('TOPPADDING', (0, 0), (-1, -1), 1),
|
295 |
-
('BOTTOMPADDING', (0, 0), (-1, -1), 1),
|
296 |
]))
|
297 |
|
298 |
story.append(table)
|
|
|
299 |
doc.build(story)
|
300 |
buffer.seek(0)
|
301 |
return buffer.getvalue()
|
302 |
|
303 |
-
|
304 |
-
|
305 |
-
|
306 |
-
|
307 |
-
|
308 |
-
page = doc[0]
|
309 |
-
pix = page.get_pixmap(matrix=fitz.Matrix(2.0, 2.0))
|
310 |
-
img = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
|
311 |
-
doc.close()
|
312 |
-
return img
|
313 |
-
except Exception as e:
|
314 |
-
st.error(f"Failed to render PDF preview: {e}")
|
315 |
-
return None
|
316 |
-
|
317 |
-
# ---------------------------------------------------------------
|
318 |
-
# Sidebar options for text size.
|
319 |
-
with st.sidebar:
|
320 |
-
auto_size = st.checkbox("Auto-size text", value=True)
|
321 |
-
if not auto_size:
|
322 |
-
base_font_size = st.slider("Base Font Size (points)", min_value=6, max_value=16, value=10, step=1)
|
323 |
-
else:
|
324 |
-
base_font_size = 10
|
325 |
-
st.info("Font size will auto-adjust between 6-12 points based on content length.")
|
326 |
|
327 |
-
#
|
328 |
-
|
329 |
-
st.session_state.markdown_content = default_markdown
|
330 |
|
331 |
-
|
332 |
-
# Generate the PDF.
|
333 |
-
with st.spinner("Generating PDF..."):
|
334 |
-
pdf_bytes = create_main_pdf(st.session_state.markdown_content, base_font_size, auto_size)
|
335 |
|
336 |
-
|
337 |
-
|
338 |
-
|
339 |
-
|
340 |
-
|
341 |
-
|
342 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
343 |
|
344 |
-
|
345 |
-
st.
|
346 |
-
|
347 |
-
|
348 |
-
|
349 |
-
|
350 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
351 |
|
352 |
-
#
|
353 |
-
|
354 |
-
"
|
355 |
-
|
356 |
-
|
357 |
-
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
358 |
|
359 |
-
#
|
360 |
-
|
361 |
-
|
362 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
363 |
|
364 |
-
#
|
365 |
-
st.
|
366 |
-
|
367 |
-
|
368 |
-
|
369 |
-
|
370 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import streamlit as st
|
2 |
+
import base64
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
from reportlab.lib.pagesizes import A4
|
4 |
from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, Table, TableStyle
|
5 |
from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
|
6 |
from reportlab.lib import colors
|
7 |
+
import pikepdf
|
8 |
+
import fpdf
|
9 |
+
import fitz # pymupdf
|
10 |
+
import cv2
|
11 |
+
import numpy as np
|
12 |
+
from PIL import Image
|
13 |
+
import io
|
14 |
+
import os
|
15 |
+
import re
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
|
17 |
+
# Define the ML outline as a markdown string for multilevel content
|
18 |
+
ml_markdown = """# Cutting-Edge ML Outline
|
|
|
19 |
|
20 |
## Core ML Techniques
|
21 |
1. π **Mixture of Experts (MoE)**
|
|
|
82 |
- Documentation synthesis
|
83 |
"""
|
84 |
|
85 |
+
# Process multilevel markdown for PDF output
|
|
|
86 |
def markdown_to_pdf_content(markdown_text):
|
87 |
+
"""Convert markdown text to a format suitable for PDF generation"""
|
88 |
+
import re
|
89 |
+
|
90 |
+
# Convert markdown headers to styled text for PDF
|
91 |
lines = markdown_text.strip().split('\n')
|
92 |
pdf_content = []
|
93 |
in_list_item = False
|
|
|
100 |
continue
|
101 |
|
102 |
if line.startswith('# '):
|
103 |
+
# Main header - will be handled separately in the PDF generation
|
104 |
pass
|
105 |
elif line.startswith('## '):
|
106 |
+
# Section header - add as a bold item
|
107 |
if current_item and sub_items:
|
108 |
+
# Store the previous item with its sub-items
|
109 |
pdf_content.append([current_item, sub_items])
|
110 |
sub_items = []
|
111 |
current_item = None
|
112 |
+
|
113 |
section = line.replace('## ', '').strip()
|
114 |
pdf_content.append(f"<b>{section}</b>")
|
115 |
in_list_item = False
|
116 |
elif re.match(r'^\d+\.', line):
|
117 |
+
# Numbered list item
|
118 |
if current_item and sub_items:
|
119 |
+
# Store the previous item with its sub-items
|
120 |
pdf_content.append([current_item, sub_items])
|
121 |
sub_items = []
|
122 |
+
|
123 |
current_item = line.strip()
|
124 |
in_list_item = True
|
125 |
elif line.startswith('- ') and in_list_item:
|
126 |
+
# Sub-item under a numbered list item
|
127 |
sub_items.append(line.strip())
|
128 |
else:
|
129 |
+
# Regular line
|
130 |
if not in_list_item:
|
131 |
pdf_content.append(line.strip())
|
132 |
|
133 |
+
# Add the last item if there is one
|
134 |
if current_item and sub_items:
|
135 |
pdf_content.append([current_item, sub_items])
|
136 |
|
137 |
+
# Split the content for two columns
|
138 |
mid_point = len(pdf_content) // 2
|
139 |
left_column = pdf_content[:mid_point]
|
140 |
right_column = pdf_content[mid_point:]
|
141 |
|
142 |
return left_column, right_column
|
143 |
|
144 |
+
# Demo functions for PDF libraries
|
145 |
+
def demo_pikepdf():
|
146 |
+
"""Create a two-column PDF with the markdown outline using pikepdf"""
|
147 |
+
# Process markdown content
|
148 |
+
left_column, right_column = markdown_to_pdf_content(ml_markdown)
|
149 |
+
|
150 |
+
# We'll use pymupdf (fitz) to create the content, then save with pikepdf
|
151 |
+
doc = fitz.open()
|
152 |
+
page = doc.new_page(width=842, height=595) # A4 Landscape
|
153 |
+
|
154 |
+
# Set up fonts and colors
|
155 |
+
title_font = "helv-b"
|
156 |
+
section_font = "helv-b"
|
157 |
+
item_font = "helv-b"
|
158 |
+
subitem_font = "helv"
|
159 |
+
blue_color = (0, 0, 0.8)
|
160 |
+
black_color = (0, 0, 0)
|
161 |
+
|
162 |
+
# Add title
|
163 |
+
page.insert_text((50, 40), "Cutting-Edge ML Outline (PikePDF Demo)", fontname=title_font, fontsize=16, color=blue_color)
|
164 |
+
|
165 |
+
# First column
|
166 |
+
x1, y1 = 50, 80
|
167 |
+
current_y = y1
|
168 |
+
|
169 |
+
for item in left_column:
|
170 |
+
if isinstance(item, str) and item.startswith('<b>'):
|
171 |
+
# Section header
|
172 |
+
# Add extra space before sections (except the first one)
|
173 |
+
if current_y > y1:
|
174 |
+
current_y += 10
|
175 |
+
|
176 |
+
text = item.replace('<b>', '').replace('</b>', '')
|
177 |
+
page.insert_text((x1, current_y), text, fontname=section_font, fontsize=14, color=blue_color)
|
178 |
+
current_y += 25
|
179 |
+
elif isinstance(item, list):
|
180 |
+
# Main item with sub-items
|
181 |
+
main_item, sub_items = item
|
182 |
+
page.insert_text((x1, current_y), main_item, fontname=item_font, fontsize=12, color=black_color)
|
183 |
+
current_y += 20
|
184 |
+
|
185 |
+
# Add sub-items
|
186 |
+
for sub_item in sub_items:
|
187 |
+
page.insert_text((x1 + 20, current_y), sub_item, fontname=subitem_font, fontsize=10, color=black_color)
|
188 |
+
current_y += 15
|
189 |
+
|
190 |
+
current_y += 5 # Extra space after a group
|
191 |
+
else:
|
192 |
+
# Regular item
|
193 |
+
page.insert_text((x1, current_y), item, fontname=item_font, fontsize=12, color=black_color)
|
194 |
+
current_y += 20
|
195 |
+
|
196 |
+
# Second column
|
197 |
+
x2, y2 = 450, 80
|
198 |
+
current_y = y2
|
199 |
+
|
200 |
+
for item in right_column:
|
201 |
+
if isinstance(item, str) and item.startswith('<b>'):
|
202 |
+
# Section header
|
203 |
+
# Add extra space before sections (except the first one)
|
204 |
+
if current_y > y2:
|
205 |
+
current_y += 10
|
206 |
+
|
207 |
+
text = item.replace('<b>', '').replace('</b>', '')
|
208 |
+
page.insert_text((x2, current_y), text, fontname=section_font, fontsize=14, color=blue_color)
|
209 |
+
current_y += 25
|
210 |
+
elif isinstance(item, list):
|
211 |
+
# Main item with sub-items
|
212 |
+
main_item, sub_items = item
|
213 |
+
page.insert_text((x2, current_y), main_item, fontname=item_font, fontsize=12, color=black_color)
|
214 |
+
current_y += 20
|
215 |
+
|
216 |
+
# Add sub-items
|
217 |
+
for sub_item in sub_items:
|
218 |
+
page.insert_text((x2 + 20, current_y), sub_item, fontname=subitem_font, fontsize=10, color=black_color)
|
219 |
+
current_y += 15
|
220 |
+
|
221 |
+
current_y += 5 # Extra space after a group
|
222 |
+
else:
|
223 |
+
# Regular item
|
224 |
+
page.insert_text((x2, current_y), item, fontname=item_font, fontsize=12, color=black_color)
|
225 |
+
current_y += 20
|
226 |
+
|
227 |
+
# Draw a dividing line
|
228 |
+
page.draw_line((421, 70), (421, 550))
|
229 |
+
|
230 |
+
# Convert to pikepdf
|
231 |
+
temp_buffer = io.BytesIO()
|
232 |
+
doc.save(temp_buffer)
|
233 |
+
temp_buffer.seek(0)
|
234 |
+
|
235 |
+
pdf = pikepdf.Pdf.open(temp_buffer)
|
236 |
+
|
237 |
+
# Save to buffer
|
238 |
buffer = io.BytesIO()
|
239 |
+
pdf.save(buffer)
|
240 |
+
buffer.seek(0)
|
241 |
+
return buffer.getvalue()
|
242 |
+
|
243 |
+
def demo_fpdf():
|
244 |
+
"""Create a two-column PDF with the markdown outline using FPDF"""
|
245 |
+
# Process markdown content
|
246 |
+
left_column, right_column = markdown_to_pdf_content(ml_markdown)
|
247 |
|
248 |
+
pdf = fpdf.FPDF(orientation='L') # Landscape
|
249 |
+
pdf.add_page()
|
250 |
+
|
251 |
+
# Set title
|
252 |
+
pdf.set_font("Arial", 'B', size=16)
|
253 |
+
pdf.set_text_color(0, 0, 128) # Dark blue
|
254 |
+
pdf.cell(0, 10, txt="Cutting-Edge ML Outline (FPDF Demo)", ln=True, align='C')
|
255 |
+
pdf.ln(10)
|
256 |
+
|
257 |
+
# Define coordinates for columns
|
258 |
+
x_col1 = 20
|
259 |
+
x_col2 = pdf.w / 2 + 10
|
260 |
+
y_start = pdf.get_y()
|
261 |
|
262 |
+
# Function to render a column
|
263 |
+
def render_column(items, x_start, y_start):
|
264 |
+
y_pos = y_start
|
265 |
+
|
266 |
+
for item in items:
|
267 |
+
if isinstance(item, str) and item.startswith('<b>'):
|
268 |
+
# Section header
|
269 |
+
text = item.replace('<b>', '').replace('</b>', '')
|
270 |
+
pdf.set_font("Arial", 'B', size=14)
|
271 |
+
pdf.set_text_color(0, 0, 128) # Dark blue
|
272 |
+
pdf.set_xy(x_start, y_pos)
|
273 |
+
pdf.cell(0, 10, txt=text, ln=True)
|
274 |
+
y_pos += 10
|
275 |
+
elif isinstance(item, list):
|
276 |
+
# Main item with sub-items
|
277 |
main_item, sub_items = item
|
278 |
+
|
279 |
+
# Main item
|
280 |
+
pdf.set_font("Arial", 'B', size=11)
|
281 |
+
pdf.set_text_color(0, 0, 0) # Black
|
282 |
+
pdf.set_xy(x_start, y_pos)
|
283 |
+
pdf.multi_cell(180, 6, txt=main_item, align='L')
|
284 |
+
main_height = pdf.get_y() - y_pos
|
285 |
+
y_pos += main_height + 2
|
286 |
+
|
287 |
+
# Sub-items
|
288 |
+
pdf.set_font("Arial", size=10)
|
289 |
+
for sub_item in sub_items:
|
290 |
+
pdf.set_xy(x_start + 10, y_pos)
|
291 |
+
pdf.multi_cell(170, 5, txt=sub_item, align='L')
|
292 |
+
sub_height = pdf.get_y() - y_pos
|
293 |
+
y_pos += sub_height + 1
|
294 |
+
|
295 |
+
y_pos += 2 # Extra space after a group
|
296 |
else:
|
297 |
+
# Regular item
|
298 |
+
pdf.set_font("Arial", 'B', size=11)
|
299 |
+
pdf.set_text_color(0, 0, 0) # Black
|
300 |
+
pdf.set_xy(x_start, y_pos)
|
301 |
+
pdf.multi_cell(180, 6, txt=item, align='L')
|
302 |
+
item_height = pdf.get_y() - y_pos
|
303 |
+
y_pos += item_height + 4
|
304 |
|
305 |
+
# Render both columns
|
306 |
+
render_column(left_column, x_col1, y_start)
|
307 |
+
render_column(right_column, x_col2, y_start)
|
308 |
|
309 |
+
# Draw a dividing line
|
310 |
+
pdf.line(pdf.w/2, 30, pdf.w/2, 280)
|
|
|
|
|
311 |
|
312 |
+
buffer = io.BytesIO()
|
313 |
+
pdf.output(buffer)
|
314 |
+
buffer.seek(0)
|
315 |
+
return buffer.getvalue()
|
316 |
+
|
317 |
+
def demo_pymupdf():
|
318 |
+
"""Create a two-column PDF with the markdown outline using PyMuPDF"""
|
319 |
+
# Process markdown content
|
320 |
+
left_column, right_column = markdown_to_pdf_content(ml_markdown)
|
321 |
+
|
322 |
+
doc = fitz.open()
|
323 |
+
page = doc.new_page(width=842, height=595) # A4 Landscape
|
324 |
+
|
325 |
+
# Set up fonts and colors
|
326 |
+
title_font = "helv-b"
|
327 |
+
section_font = "helv-b"
|
328 |
+
item_font = "helv-b"
|
329 |
+
subitem_font = "helv"
|
330 |
+
blue_color = (0, 0, 0.8)
|
331 |
+
black_color = (0, 0, 0)
|
332 |
+
|
333 |
+
# Add title
|
334 |
+
page.insert_text((300, 40), "Cutting-Edge ML Outline (PyMuPDF Demo)", fontname=title_font, fontsize=16, color=blue_color)
|
335 |
+
|
336 |
+
# First column
|
337 |
+
x1, y1 = 50, 80
|
338 |
+
current_y = y1
|
339 |
+
|
340 |
+
for item in left_column:
|
341 |
+
if isinstance(item, str) and item.startswith('<b>'):
|
342 |
+
# Section header
|
343 |
+
# Add extra space before sections (except the first one)
|
344 |
+
if current_y > y1:
|
345 |
+
current_y += 10
|
346 |
+
|
347 |
+
text = item.replace('<b>', '').replace('</b>', '')
|
348 |
+
page.insert_text((x1, current_y), text, fontname=section_font, fontsize=14, color=blue_color)
|
349 |
+
current_y += 25
|
350 |
+
elif isinstance(item, list):
|
351 |
+
# Main item with sub-items
|
352 |
+
main_item, sub_items = item
|
353 |
+
page.insert_text((x1, current_y), main_item, fontname=item_font, fontsize=12, color=black_color)
|
354 |
+
current_y += 20
|
355 |
+
|
356 |
+
# Add sub-items
|
357 |
+
for sub_item in sub_items:
|
358 |
+
page.insert_text((x1 + 20, current_y), sub_item, fontname=subitem_font, fontsize=10, color=black_color)
|
359 |
+
current_y += 15
|
360 |
+
|
361 |
+
current_y += 5 # Extra space after a group
|
362 |
+
else:
|
363 |
+
# Regular item
|
364 |
+
page.insert_text((x1, current_y), item, fontname=item_font, fontsize=12, color=black_color)
|
365 |
+
current_y += 20
|
366 |
+
|
367 |
+
# Second column
|
368 |
+
x2, y2 = 450, 80
|
369 |
+
current_y = y2
|
370 |
+
|
371 |
+
for item in right_column:
|
372 |
+
if isinstance(item, str) and item.startswith('<b>'):
|
373 |
+
# Section header
|
374 |
+
# Add extra space before sections (except the first one)
|
375 |
+
if current_y > y2:
|
376 |
+
current_y += 10
|
377 |
+
|
378 |
+
text = item.replace('<b>', '').replace('</b>', '')
|
379 |
+
page.insert_text((x2, current_y), text, fontname=section_font, fontsize=14, color=blue_color)
|
380 |
+
current_y += 25
|
381 |
+
elif isinstance(item, list):
|
382 |
+
# Main item with sub-items
|
383 |
+
main_item, sub_items = item
|
384 |
+
page.insert_text((x2, current_y), main_item, fontname=item_font, fontsize=12, color=black_color)
|
385 |
+
current_y += 20
|
386 |
+
|
387 |
+
# Add sub-items
|
388 |
+
for sub_item in sub_items:
|
389 |
+
page.insert_text((x2 + 20, current_y), sub_item, fontname=subitem_font, fontsize=10, color=black_color)
|
390 |
+
current_y += 15
|
391 |
+
|
392 |
+
current_y += 5 # Extra space after a group
|
393 |
+
else:
|
394 |
+
# Regular item
|
395 |
+
page.insert_text((x2, current_y), item, fontname=item_font, fontsize=12, color=black_color)
|
396 |
+
current_y += 20
|
397 |
+
|
398 |
+
# Draw a dividing line
|
399 |
+
page.draw_line((421, 70), (421, 550))
|
400 |
+
|
401 |
+
buffer = io.BytesIO()
|
402 |
+
doc.save(buffer)
|
403 |
+
buffer.seek(0)
|
404 |
+
return buffer.getvalue()
|
405 |
+
|
406 |
+
# Demo function for image capture
|
407 |
+
def demo_image_capture():
|
408 |
+
"""Generate a demo image (fake capture) since we can't access the camera in this environment"""
|
409 |
+
# Create a simple gradient image using numpy and PIL
|
410 |
+
width, height = 640, 480
|
411 |
+
|
412 |
+
# Create a gradient array
|
413 |
+
x = np.linspace(0, 1, width)
|
414 |
+
y = np.linspace(0, 1, height)
|
415 |
+
xx, yy = np.meshgrid(x, y)
|
416 |
+
gradient = (xx + yy) / 2
|
417 |
+
|
418 |
+
# Convert to RGB image
|
419 |
+
img_array = (gradient * 255).astype(np.uint8)
|
420 |
+
rgb_array = np.stack([img_array, img_array//2, img_array*2], axis=2)
|
421 |
+
|
422 |
+
# Create PIL Image
|
423 |
+
img = Image.fromarray(rgb_array)
|
424 |
+
|
425 |
+
# Add text to the image
|
426 |
+
from PIL import ImageDraw, ImageFont
|
427 |
+
draw = ImageDraw.Draw(img)
|
428 |
+
try:
|
429 |
+
font = ImageFont.truetype("arial.ttf", 30)
|
430 |
+
except:
|
431 |
+
font = ImageFont.load_default()
|
432 |
+
|
433 |
+
draw.text((width//4, height//2), "OpenCV Demo Image", fill=(255, 255, 255), font=font)
|
434 |
+
|
435 |
+
# Save to buffer
|
436 |
+
buffer = io.BytesIO()
|
437 |
+
img.save(buffer, format="JPEG")
|
438 |
+
buffer.seek(0)
|
439 |
+
return buffer.getvalue()
|
440 |
+
|
441 |
+
# Main PDF creation using ReportLab
|
442 |
+
def create_main_pdf(markdown_text):
|
443 |
+
"""Create a single-page landscape PDF with the outline in two columns"""
|
444 |
+
from reportlab.platypus import Table, TableStyle, Paragraph, Spacer
|
445 |
+
from reportlab.lib import pagesizes
|
446 |
+
from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
|
447 |
+
|
448 |
+
# Process markdown content
|
449 |
+
left_column, right_column = markdown_to_pdf_content(markdown_text)
|
450 |
+
|
451 |
+
buffer = io.BytesIO()
|
452 |
+
doc = SimpleDocTemplate(
|
453 |
+
buffer,
|
454 |
+
pagesize=(A4[1], A4[0]), # Landscape
|
455 |
+
leftMargin=50,
|
456 |
+
rightMargin=50,
|
457 |
+
topMargin=50,
|
458 |
+
bottomMargin=50
|
459 |
)
|
460 |
|
461 |
+
styles = getSampleStyleSheet()
|
462 |
+
story = []
|
463 |
+
|
464 |
+
# Create custom styles
|
465 |
+
title_style = styles['Heading1']
|
466 |
+
title_style.textColor = colors.darkblue
|
467 |
+
title_style.alignment = 1 # Center alignment
|
468 |
+
|
469 |
section_style = ParagraphStyle(
|
470 |
'SectionStyle',
|
471 |
parent=styles['Heading2'],
|
|
|
472 |
textColor=colors.darkblue,
|
473 |
+
spaceAfter=6
|
|
|
|
|
474 |
)
|
475 |
|
476 |
item_style = ParagraphStyle(
|
477 |
'ItemStyle',
|
478 |
parent=styles['Normal'],
|
479 |
+
fontSize=11,
|
480 |
+
leading=14,
|
481 |
+
fontName='Helvetica-Bold'
|
|
|
482 |
)
|
483 |
|
484 |
subitem_style = ParagraphStyle(
|
485 |
'SubItemStyle',
|
486 |
parent=styles['Normal'],
|
487 |
+
fontSize=10,
|
488 |
+
leading=12,
|
489 |
+
leftIndent=20
|
|
|
|
|
490 |
)
|
491 |
|
492 |
+
# Add title
|
493 |
+
story.append(Paragraph("Cutting-Edge ML Outline (ReportLab)", title_style))
|
494 |
+
story.append(Spacer(1, 20))
|
495 |
|
496 |
+
# Prepare data for table
|
497 |
left_cells = []
|
498 |
for item in left_column:
|
499 |
if isinstance(item, str) and item.startswith('<b>'):
|
500 |
+
# Section header
|
501 |
text = item.replace('<b>', '').replace('</b>', '')
|
502 |
+
left_cells.append(Paragraph(text, section_style))
|
503 |
elif isinstance(item, list):
|
504 |
+
# Main item with sub-items
|
505 |
main_item, sub_items = item
|
506 |
+
left_cells.append(Paragraph(main_item, item_style))
|
507 |
+
|
508 |
+
# Sub items
|
509 |
for sub_item in sub_items:
|
510 |
+
left_cells.append(Paragraph(sub_item, subitem_style))
|
511 |
else:
|
512 |
+
# Regular item
|
513 |
+
left_cells.append(Paragraph(item, item_style))
|
514 |
|
515 |
right_cells = []
|
516 |
for item in right_column:
|
517 |
if isinstance(item, str) and item.startswith('<b>'):
|
518 |
+
# Section header
|
519 |
text = item.replace('<b>', '').replace('</b>', '')
|
520 |
+
right_cells.append(Paragraph(text, section_style))
|
521 |
elif isinstance(item, list):
|
522 |
+
# Main item with sub-items
|
523 |
main_item, sub_items = item
|
524 |
+
right_cells.append(Paragraph(main_item, item_style))
|
525 |
+
|
526 |
+
# Sub items
|
527 |
for sub_item in sub_items:
|
528 |
+
right_cells.append(Paragraph(sub_item, subitem_style))
|
529 |
else:
|
530 |
+
# Regular item
|
531 |
+
right_cells.append(Paragraph(item, item_style))
|
532 |
|
533 |
+
# Make sure both columns have the same number of rows by adding empty cells
|
534 |
max_cells = max(len(left_cells), len(right_cells))
|
535 |
+
if len(left_cells) < max_cells:
|
536 |
+
for i in range(max_cells - len(left_cells)):
|
537 |
+
left_cells.append("")
|
538 |
+
if len(right_cells) < max_cells:
|
539 |
+
for i in range(max_cells - len(right_cells)):
|
540 |
+
right_cells.append("")
|
541 |
+
|
542 |
+
# Create table data (one row per cell)
|
543 |
+
table_data = []
|
544 |
+
for i in range(max_cells):
|
545 |
+
table_data.append([left_cells[i], right_cells[i]])
|
546 |
+
|
547 |
+
# Calculate column widths
|
548 |
+
col_width = (A4[1] - 120) / 2.0 # Page width minus margins divided by 2
|
549 |
+
|
550 |
+
# Create the table with the data
|
551 |
+
table = Table(table_data, colWidths=[col_width, col_width])
|
552 |
|
553 |
+
# Style the table
|
|
|
|
|
554 |
table.setStyle(TableStyle([
|
555 |
('VALIGN', (0, 0), (-1, -1), 'TOP'),
|
556 |
+
('ALIGN', (0, 0), (0, -1), 'LEFT'),
|
557 |
+
('ALIGN', (1, 0), (1, -1), 'LEFT'),
|
558 |
('BACKGROUND', (0, 0), (-1, -1), colors.white),
|
559 |
+
('GRID', (0, 0), (-1, -1), 0.5, colors.white),
|
560 |
+
('LINEAFTER', (0, 0), (0, -1), 1, colors.grey),
|
|
|
|
|
|
|
|
|
561 |
]))
|
562 |
|
563 |
story.append(table)
|
564 |
+
|
565 |
doc.build(story)
|
566 |
buffer.seek(0)
|
567 |
return buffer.getvalue()
|
568 |
|
569 |
+
def get_binary_file_downloader_html(bin_data, file_label='File'):
|
570 |
+
"""Create a download link for binary data"""
|
571 |
+
bin_str = base64.b64encode(bin_data).decode()
|
572 |
+
href = f'<a href="data:application/octet-stream;base64,{bin_str}" download="{file_label}">Download {file_label}</a>'
|
573 |
+
return href
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
574 |
|
575 |
+
# Streamlit UI
|
576 |
+
st.title("π Cutting-Edge ML Outline Generator")
|
|
|
577 |
|
578 |
+
col1, col2 = st.columns(2)
|
|
|
|
|
|
|
579 |
|
580 |
+
with col1:
|
581 |
+
st.header("π Markdown Outline")
|
582 |
+
|
583 |
+
# Display the markdown content
|
584 |
+
st.markdown(ml_markdown)
|
585 |
+
|
586 |
+
# Create a download button for the markdown file
|
587 |
+
st.download_button(
|
588 |
+
label="Download Markdown",
|
589 |
+
data=ml_markdown,
|
590 |
+
file_name="ml_outline.md",
|
591 |
+
mime="text/markdown"
|
592 |
+
)
|
593 |
+
|
594 |
+
# Show the markdown source code in an expandable section
|
595 |
+
with st.expander("View Markdown Source"):
|
596 |
+
st.code(ml_markdown, language="markdown")
|
597 |
|
598 |
+
with col2:
|
599 |
+
st.header("π PDF Preview & Demos")
|
600 |
+
|
601 |
+
# Library Demos
|
602 |
+
st.subheader("Library Demos")
|
603 |
+
|
604 |
+
# PikePDF demo
|
605 |
+
if st.button("Generate PikePDF Demo"):
|
606 |
+
with st.spinner("Generating PikePDF demo..."):
|
607 |
+
pike_pdf = demo_pikepdf()
|
608 |
+
st.download_button("Download PikePDF Demo", pike_pdf, "pikepdf_demo.pdf")
|
609 |
+
st.success("PikePDF demo created successfully!")
|
610 |
+
st.info("This PDF contains the multilevel markdown outline in a two-column layout.")
|
611 |
+
|
612 |
+
# FPDF demo
|
613 |
+
if st.button("Generate FPDF Demo"):
|
614 |
+
with st.spinner("Generating FPDF demo..."):
|
615 |
+
fpdf_pdf = demo_fpdf()
|
616 |
+
st.download_button("Download FPDF Demo", fpdf_pdf, "fpdf_demo.pdf")
|
617 |
+
st.success("FPDF demo created successfully!")
|
618 |
+
st.info("This PDF contains the multilevel markdown outline in a two-column layout.")
|
619 |
+
|
620 |
+
# PyMuPDF demo
|
621 |
+
if st.button("Generate PyMuPDF Demo"):
|
622 |
+
with st.spinner("Generating PyMuPDF demo..."):
|
623 |
+
pymupdf_pdf = demo_pymupdf()
|
624 |
+
st.download_button("Download PyMuPDF Demo", pymupdf_pdf, "pymupdf_demo.pdf")
|
625 |
+
st.success("PyMuPDF demo created successfully!")
|
626 |
+
st.info("This PDF contains the multilevel markdown outline in a two-column layout.")
|
627 |
+
|
628 |
+
# Image demo
|
629 |
+
if st.button("Generate Demo Image"):
|
630 |
+
with st.spinner("Generating demo image..."):
|
631 |
+
img_data = demo_image_capture()
|
632 |
+
st.image(img_data, caption="Demo Image (Camera simulation)")
|
633 |
+
|
634 |
+
# Add download button for the image
|
635 |
+
st.download_button(
|
636 |
+
label="Download Image",
|
637 |
+
data=img_data,
|
638 |
+
file_name="demo_image.jpg",
|
639 |
+
mime="image/jpeg"
|
640 |
+
)
|
641 |
|
642 |
+
# Main PDF Generation
|
643 |
+
st.subheader("Main Outline PDF")
|
644 |
+
if st.button("Generate Main PDF"):
|
645 |
+
with st.spinner("Generating PDF..."):
|
646 |
+
try:
|
647 |
+
pdf_bytes = create_main_pdf(ml_markdown)
|
648 |
+
|
649 |
+
st.download_button(
|
650 |
+
label="Download Main PDF",
|
651 |
+
data=pdf_bytes,
|
652 |
+
file_name="ml_outline.pdf",
|
653 |
+
mime="application/pdf"
|
654 |
+
)
|
655 |
+
|
656 |
+
# Display the PDF in the app
|
657 |
+
base64_pdf = base64.b64encode(pdf_bytes).decode('utf-8')
|
658 |
+
pdf_display = f'''
|
659 |
+
<embed
|
660 |
+
src="data:application/pdf;base64,{base64_pdf}"
|
661 |
+
width="100%"
|
662 |
+
height="400px"
|
663 |
+
type="application/pdf">
|
664 |
+
'''
|
665 |
+
st.markdown(pdf_display, unsafe_allow_html=True)
|
666 |
+
|
667 |
+
st.success("PDF generated successfully! The PDF displays the multilevel markdown outline in a two-column layout.")
|
668 |
+
except Exception as e:
|
669 |
+
st.error(f"Error generating PDF: {str(e)}")
|
670 |
|
671 |
+
# Show the PDF rendering code in an expandable section
|
672 |
+
with st.expander("View PDF Rendering Code"):
|
673 |
+
st.code("""
|
674 |
+
# Process multilevel markdown for PDF output
|
675 |
+
def markdown_to_pdf_content(markdown_text):
|
676 |
+
# Convert markdown headers to styled text for PDF
|
677 |
+
lines = markdown_text.strip().split('\\n')
|
678 |
+
pdf_content = []
|
679 |
+
|
680 |
+
for line in lines:
|
681 |
+
if line.startswith('# '):
|
682 |
+
# Main header - will be handled separately
|
683 |
+
pass
|
684 |
+
elif line.startswith('## '):
|
685 |
+
# Section header - add as a bold item
|
686 |
+
section = line.replace('## ', '').strip()
|
687 |
+
pdf_content.append(f"<b>{section}</b>")
|
688 |
+
elif re.match(r'^\\d+\\.', line):
|
689 |
+
# Numbered list item
|
690 |
+
item = line.strip()
|
691 |
+
pdf_content.append(item)
|
692 |
+
elif line.startswith('- '):
|
693 |
+
# Sub-item under a numbered list item
|
694 |
+
sub_item = line.strip()
|
695 |
+
pdf_content.append(" " + sub_item)
|
696 |
+
|
697 |
+
# Split the content for two columns
|
698 |
+
mid_point = len(pdf_content) // 2
|
699 |
+
left_column = pdf_content[:mid_point]
|
700 |
+
right_column = pdf_content[mid_point:]
|
701 |
+
|
702 |
+
return left_column, right_column
|
703 |
+
""", language="python")
|
704 |
|
705 |
+
# Add custom CSS for better appearance
|
706 |
+
st.markdown("""
|
707 |
+
<style>
|
708 |
+
.stButton>button {
|
709 |
+
background-color: #4CAF50;
|
710 |
+
color: white;
|
711 |
+
font-weight: bold;
|
712 |
+
}
|
713 |
+
.stTabs [data-baseweb="tab-list"] {
|
714 |
+
gap: 2px;
|
715 |
+
}
|
716 |
+
.stTabs [data-baseweb="tab"] {
|
717 |
+
height: 50px;
|
718 |
+
white-space: pre-wrap;
|
719 |
+
background-color: #f0f2f6;
|
720 |
+
border-radius: 4px 4px 0px 0px;
|
721 |
+
gap: 1px;
|
722 |
+
padding-top: 10px;
|
723 |
+
padding-bottom: 10px;
|
724 |
+
}
|
725 |
+
</style>
|
726 |
+
""", unsafe_allow_html=True)
|