Spaces:
Build error
Build error
Update app.py
Browse files
app.py
CHANGED
@@ -5,475 +5,281 @@ import re
|
|
5 |
import os
|
6 |
import shutil
|
7 |
import fitz # PyMuPDF
|
8 |
-
from PIL import Image
|
9 |
-
from collections import defaultdict
|
10 |
import io
|
11 |
-
from pypdf import PdfWriter
|
12 |
-
import random
|
13 |
|
14 |
-
|
15 |
-
from lxml import etree
|
16 |
-
from svglib.svglib import svg2rlg
|
17 |
-
|
18 |
-
from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, Table, TableStyle, PageBreak, BaseDocTemplate, Frame, PageTemplate, Image as ReportLabImage, Flowable
|
19 |
from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
|
20 |
-
from reportlab.lib.pagesizes import letter
|
21 |
from reportlab.lib.units import inch
|
22 |
from reportlab.lib import colors
|
23 |
from reportlab.pdfbase import pdfmetrics
|
24 |
from reportlab.pdfbase.ttfonts import TTFont
|
25 |
-
from functools import partial
|
26 |
|
27 |
-
# --- Configuration
|
28 |
CWD = Path.cwd()
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
"Letter Landscape": {"size": landscape(letter)},
|
35 |
-
"Legal Portrait": {"size": legal},
|
36 |
-
"Legal Landscape": {"size": landscape(legal)},
|
37 |
-
}
|
38 |
-
OUTPUT_DIR = CWD / "generated_pdfs"
|
39 |
-
PREVIEW_DIR = CWD / "previews"
|
40 |
-
FONT_DIR = CWD # Assumes fonts are in the same directory as the script
|
41 |
-
|
42 |
-
# Create necessary directories
|
43 |
-
OUTPUT_DIR.mkdir(exist_ok=True)
|
44 |
-
PREVIEW_DIR.mkdir(exist_ok=True)
|
45 |
-
TEMP_DIR.mkdir(exist_ok=True)
|
46 |
|
|
|
|
|
|
|
47 |
|
48 |
# --- Font & Emoji Handling ---
|
49 |
-
|
50 |
-
|
51 |
|
52 |
-
def
|
53 |
-
"""Finds
|
54 |
-
global EMOJI_FONT_PATH
|
55 |
-
print("---
|
56 |
-
text_font_names = []
|
57 |
|
|
|
58 |
noto_emoji_path = FONT_DIR / "NotoColorEmoji-Regular.ttf"
|
59 |
if noto_emoji_path.exists():
|
60 |
EMOJI_FONT_PATH = str(noto_emoji_path)
|
|
|
61 |
else:
|
62 |
-
print("CRITICAL: 'NotoColorEmoji-Regular.ttf' not found
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
-
try:
|
69 |
-
font_name = font_path.stem
|
70 |
-
if not UI_FONT_PATH: UI_FONT_PATH = str(font_path) # Grab first available font for UI icons
|
71 |
-
|
72 |
-
pdfmetrics.registerFont(TTFont(font_name, str(font_path)))
|
73 |
-
pdfmetrics.registerFontFamily(font_name, normal=font_name, bold=font_name, italic=font_name, boldItalic=font_name)
|
74 |
-
|
75 |
-
if "notocoloremoji-regular" not in font_name.lower() and "notoemoji" not in font_name.lower():
|
76 |
-
text_font_names.append(font_name)
|
77 |
-
except Exception as e:
|
78 |
-
print(f"Could not register font {font_path.name}: {e}")
|
79 |
-
|
80 |
-
if not text_font_names:
|
81 |
-
text_font_names.append('Helvetica')
|
82 |
-
print(f"Successfully registered user-selectable fonts: {text_font_names}")
|
83 |
-
return sorted(text_font_names)
|
84 |
-
|
85 |
-
def render_emoji_as_image(emoji_char, size_pt):
|
86 |
-
if not EMOJI_FONT_PATH: return None
|
87 |
-
if (emoji_char, size_pt) in EMOJI_IMAGE_CACHE: return EMOJI_IMAGE_CACHE[(emoji_char, size_pt)]
|
88 |
-
try:
|
89 |
-
rect, doc = fitz.Rect(0, 0, size_pt*1.5, size_pt*1.5), fitz.open()
|
90 |
-
page = doc.new_page(width=rect.width, height=rect.height)
|
91 |
-
page.insert_font(fontname="emoji", fontfile=EMOJI_FONT_PATH)
|
92 |
-
page.insert_text(fitz.Point(0, size_pt * 1.1), emoji_char, fontname="emoji", fontsize=size_pt)
|
93 |
-
pix, img_buffer = page.get_pixmap(alpha=True, dpi=300), io.BytesIO(pix.tobytes("png"))
|
94 |
-
doc.close(); img_buffer.seek(0)
|
95 |
-
EMOJI_IMAGE_CACHE[(emoji_char, size_pt)] = img_buffer
|
96 |
-
return img_buffer
|
97 |
-
except Exception as e: print(f"Could not render emoji {emoji_char} as image: {e}"); return None
|
98 |
-
|
99 |
-
def render_emoji_as_svg(emoji_char, size_pt):
|
100 |
-
if not EMOJI_FONT_PATH: return None
|
101 |
-
if (emoji_char, size_pt) in EMOJI_SVG_CACHE: return EMOJI_SVG_CACHE[(emoji_char, size_pt)]
|
102 |
try:
|
103 |
-
|
104 |
-
|
105 |
-
|
106 |
-
|
107 |
-
drawing = svg2rlg(io.BytesIO(svg_image.encode('utf-8')))
|
108 |
-
scale_factor = (size_pt * 1.2) / drawing.height
|
109 |
-
drawing.width, drawing.height = drawing.width*scale_factor, drawing.height*scale_factor
|
110 |
-
drawing.scale(scale_factor, scale_factor)
|
111 |
-
EMOJI_SVG_CACHE[(emoji_char, size_pt)] = drawing
|
112 |
-
return drawing
|
113 |
-
except Exception as e: print(f"Could not render emoji {emoji_char} as SVG: {e}"); return None
|
114 |
|
115 |
-
|
116 |
-
def generate_ai_content_api(prompt):
|
117 |
"""
|
118 |
-
|
|
|
119 |
"""
|
120 |
-
if not
|
121 |
-
|
122 |
-
|
123 |
-
|
124 |
-
|
125 |
-
|
126 |
-
|
127 |
-
# The Quest for the Sunstone โ๏ธ
|
128 |
-
|
129 |
-
Long ago, in the mystical land of Aerthos, a creeping shadow began to fall across the valleys. The once-vibrant flowers ๐ธ drooped, and the rivers ran slow. The elders knew the cause: the light of the Sunstone, hidden deep within Mount Cinder, was fading.
|
130 |
-
|
131 |
-
## The Prophecy ๐
|
132 |
-
|
133 |
-
An ancient prophecy spoke of a hero who would rekindle the stone. It read:
|
134 |
-
> When darkness drapes the verdant ground,
|
135 |
-
> A soul of courage shall be found.
|
136 |
-
> Through trials of fire ๐ฅ, wit, and might,
|
137 |
-
> They'll bring once more the sacred light.
|
138 |
-
|
139 |
-
## The Chosen One ๐ฆธโโ๏ธ
|
140 |
-
|
141 |
-
A young woman named Elara, known for her kindness and bravery, was chosen. She accepted the quest, her only companions a loyal wolf ๐บ and a map gifted by the village shaman.
|
142 |
-
"""
|
143 |
-
return f"# Golem's Vision for: '{prompt}'\n\n{sample_story}"
|
144 |
-
|
145 |
-
|
146 |
-
# --- UI & File Handling ---
|
147 |
-
FILE_ICONS = defaultdict(lambda: 'โ', {
|
148 |
-
'.md': '๐', '.txt': '๐', '.py': '๐', '.js': 'โก', '.html': '๐', '.css': '๐จ', '.json': '๏ฟฝ',
|
149 |
-
'.png': '๐ผ๏ธ', '.jpg': '๐ผ๏ธ', '.jpeg': '๐ผ๏ธ', '.gif': '๐ผ๏ธ', '.bmp': '๐ผ๏ธ', '.tiff': '๐ผ๏ธ',
|
150 |
-
'.pdf': '๐'
|
151 |
-
})
|
152 |
|
153 |
-
def create_file_icon_image(file_path: Path):
|
154 |
-
"""Creates a preview image for a file with an icon and its name."""
|
155 |
-
icon = FILE_ICONS[file_path.suffix.lower()]
|
156 |
-
img = Image.new('RGB', (150, 100), color='#1f2937')
|
157 |
-
draw = ImageDraw.Draw(img)
|
158 |
try:
|
159 |
-
|
160 |
-
|
161 |
-
|
162 |
-
|
163 |
|
164 |
-
|
165 |
-
|
166 |
-
|
167 |
-
|
168 |
-
|
169 |
-
return out_path
|
170 |
-
|
171 |
-
def update_staging_and_manuscript(files, current_manuscript):
|
172 |
-
"""Processes uploaded files, updates staging galleries, and consolidates text content."""
|
173 |
-
shutil.rmtree(TEMP_DIR, ignore_errors=True); TEMP_DIR.mkdir(exist_ok=True)
|
174 |
-
|
175 |
-
text_gallery, image_gallery, pdf_gallery = [], [], []
|
176 |
-
text_content_to_add = []
|
177 |
-
has_text_files = False
|
178 |
-
|
179 |
-
if not files:
|
180 |
-
return text_gallery, image_gallery, pdf_gallery, current_manuscript
|
181 |
-
|
182 |
-
sorted_files = sorted(files, key=lambda x: Path(x.name).suffix)
|
183 |
-
|
184 |
-
for file_obj in sorted_files:
|
185 |
-
file_path = Path(file_obj.name)
|
186 |
-
ext = file_path.suffix.lower()
|
187 |
|
188 |
-
|
189 |
-
|
190 |
-
|
191 |
-
pdf_gallery.append((create_file_icon_image(file_path), file_path.name))
|
192 |
-
else: # Treat everything else as a text file
|
193 |
-
has_text_files = True
|
194 |
-
text_gallery.append((create_file_icon_image(file_path), file_path.name))
|
195 |
-
try:
|
196 |
-
with open(file_path, 'r', encoding='utf-8') as f:
|
197 |
-
text_content_to_add.append(f.read())
|
198 |
-
except Exception as e:
|
199 |
-
print(f"Could not read {file_path.name}: {e}")
|
200 |
|
201 |
-
|
202 |
-
|
203 |
-
|
204 |
-
updated_manuscript = "\n\n---\n\n".join(text_content_to_add)
|
205 |
-
else:
|
206 |
-
updated_manuscript = current_manuscript
|
207 |
|
208 |
-
|
|
|
|
|
|
|
|
|
|
|
209 |
|
210 |
-
# ---
|
211 |
-
def
|
212 |
-
"""
|
213 |
-
|
214 |
-
|
|
|
|
|
215 |
|
216 |
-
|
217 |
-
final_header_text = header_text.replace("[Page #]", str(page_num)).replace("[Title]", title)
|
218 |
-
|
219 |
-
if final_header_text:
|
220 |
-
canvas.setFont('Helvetica', 9)
|
221 |
-
canvas.setFillColor(colors.grey)
|
222 |
-
canvas.drawRightString(doc.width + doc.leftMargin, doc.height + doc.topMargin + 0.25*inch, final_header_text)
|
223 |
-
|
224 |
-
if final_footer_text:
|
225 |
-
canvas.setFont('Helvetica', 9)
|
226 |
-
canvas.setFillColor(colors.grey)
|
227 |
-
canvas.drawString(doc.leftMargin, doc.bottomMargin - 0.25*inch, final_footer_text)
|
228 |
-
|
229 |
-
canvas.restoreState()
|
230 |
-
|
231 |
-
def markdown_to_story(markdown_text, font_name, font_sizes, use_svg_engine):
|
232 |
styles = getSampleStyleSheet()
|
233 |
-
|
234 |
-
|
235 |
-
|
236 |
-
|
237 |
-
|
238 |
-
|
239 |
-
|
240 |
-
|
241 |
emoji_pattern = re.compile(f"([{re.escape(''.join(map(chr, range(0x1f600, 0x1f650))))}"
|
242 |
f"{re.escape(''.join(map(chr, range(0x1f300, 0x1f5ff))))}"
|
243 |
f"{re.escape(''.join(map(chr, range(0x1f900, 0x1f9ff))))}"
|
244 |
-
f"{re.escape(''.join(map(chr, range(0x2600, 0x26ff))))}"
|
245 |
-
|
246 |
-
|
247 |
-
|
248 |
-
|
|
|
|
|
|
|
|
|
249 |
for part in parts:
|
250 |
if not part: continue
|
|
|
251 |
if emoji_pattern.match(part):
|
|
|
252 |
for emoji_char in part:
|
253 |
-
|
254 |
-
if
|
|
|
|
|
|
|
255 |
else:
|
256 |
-
|
257 |
-
flowables.append(Paragraph(
|
258 |
-
|
259 |
-
|
260 |
-
|
261 |
-
|
262 |
-
|
263 |
-
story, lines = [], markdown_text.split('\n')
|
264 |
-
in_code_block, in_table, first_heading, document_title = False, False, True, "Untitled Document"
|
265 |
-
code_block_text, table_data = "", []
|
266 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
267 |
for line in lines:
|
268 |
stripped_line = line.strip()
|
269 |
-
if stripped_line.startswith("```"):
|
270 |
-
if in_code_block:
|
271 |
-
escaped_code = code_block_text.replace('&', '&').replace('<', '<').replace('>', '>')
|
272 |
-
story.extend([Paragraph(escaped_code.replace('\n', '<br/>'), style_code), Spacer(1, 0.1*inch)])
|
273 |
-
in_code_block, code_block_text = False, ""
|
274 |
-
else: in_code_block = True
|
275 |
-
continue
|
276 |
-
if in_code_block: code_block_text += line + '\n'; continue
|
277 |
-
if stripped_line.startswith('|'):
|
278 |
-
if not in_table: in_table = True
|
279 |
-
if all(c in '-|: ' for c in stripped_line): continue
|
280 |
-
table_data.append([cell.strip() for cell in stripped_line.strip('|').split('|')])
|
281 |
-
continue
|
282 |
-
if in_table:
|
283 |
-
in_table = False
|
284 |
-
if table_data:
|
285 |
-
processed_table_data = [[create_flowables_for_line(cell, style_normal)[0] for cell in row] for row in table_data[1:]]
|
286 |
-
header_row = [create_flowables_for_line(f"<b>{cell}</b>", style_table_header)[0] for cell in table_data[0]]
|
287 |
-
table = Table([header_row] + processed_table_data, hAlign='LEFT', repeatRows=1)
|
288 |
-
table.setStyle(TableStyle([('BACKGROUND', (0,0), (-1,0), colors.HexColor("#4a044e")), ('GRID', (0,0), (-1,-1), 1, colors.HexColor("#6b21a8")), ('VALIGN', (0,0), (-1,-1), 'MIDDLE'), ('TOPPADDING', (0,0), (-1,-1), 6), ('BOTTOMPADDING', (0,0), (-1,-1), 6)]))
|
289 |
-
story.extend([table, Spacer(1, 0.2*inch)])
|
290 |
-
table_data = []
|
291 |
-
continue
|
292 |
-
|
293 |
-
if not stripped_line: continue
|
294 |
-
content, style, bullet_text = stripped_line, style_normal, None
|
295 |
-
if stripped_line.startswith("# "):
|
296 |
-
if not first_heading: story.append(PageBreak())
|
297 |
-
content, style, first_heading = stripped_line.lstrip('# '), style_h1, False
|
298 |
-
if document_title == "Untitled Document": document_title = content
|
299 |
-
elif stripped_line.startswith("## "): content, style = stripped_line.lstrip('## '), style_h2
|
300 |
-
elif stripped_line.startswith("### "): content, style = stripped_line.lstrip('### '), style_h3
|
301 |
-
elif stripped_line.startswith(("- ", "* ")): content, bullet_text = stripped_line[2:], 'โข '
|
302 |
|
303 |
-
|
304 |
-
if
|
305 |
-
|
306 |
-
|
307 |
-
|
308 |
-
|
309 |
-
|
310 |
-
|
311 |
-
|
312 |
-
|
313 |
-
|
314 |
-
|
315 |
-
if not EMOJI_FONT_PATH: raise gr.Error("CRITICAL: Cannot generate PDFs. 'NotoColorEmoji-Regular.ttf' not found.")
|
316 |
|
317 |
-
|
318 |
-
|
319 |
-
|
320 |
-
|
321 |
-
if files:
|
322 |
-
for f in files:
|
323 |
-
p = Path(f.name); ext = p.suffix.lower()
|
324 |
-
if ext in ['.png','.jpg','.jpeg','.gif','.bmp','.tiff']: image_files.append(p)
|
325 |
-
elif ext == '.pdf': pdf_files.append(p)
|
326 |
-
else: txt_files.append(p)
|
327 |
|
328 |
-
|
329 |
-
|
330 |
-
|
331 |
-
|
332 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
333 |
|
334 |
-
|
335 |
-
EMOJI_IMAGE_CACHE.clear()
|
336 |
-
|
337 |
-
for layout in progress.tqdm(layouts, desc="brewing potions..."):
|
338 |
-
for font in progress.tqdm(fonts, desc=f"enchanting scrolls with {layout}..."):
|
339 |
-
merger = PdfWriter()
|
340 |
-
if md_content:
|
341 |
-
buffer, (story, title) = io.BytesIO(), markdown_to_story(md_content, font, font_sizes, use_svg_engine)
|
342 |
-
doc = BaseDocTemplate(buffer, pagesize=LAYOUTS[layout]["size"], leftMargin=margins[2]*inch, rightMargin=margins[3]*inch, topMargin=margins[0]*inch, bottomMargin=margins[1]*inch)
|
343 |
-
frame_w = (doc.width / num_columns) - (num_columns - 1)*0.1*inch
|
344 |
-
frames = [Frame(doc.leftMargin + i*(frame_w+0.2*inch), doc.bottomMargin, frame_w, doc.height, id=f'col_{i}') for i in range(num_columns)]
|
345 |
-
doc.addPageTemplates([PageTemplate(id='main', frames=frames, onPage=partial(_draw_header_footer, header_text=header_text, footer_text=footer_text, title=title))])
|
346 |
-
doc.build(story); buffer.seek(0); merger.append(fileobj=buffer)
|
347 |
-
|
348 |
-
for p in image_files:
|
349 |
-
try:
|
350 |
-
with Image.open(p) as img: w, h = img.size
|
351 |
-
w_pt, h_pt = w * (inch/72), h * (inch/72)
|
352 |
-
buffer = io.BytesIO()
|
353 |
-
SimpleDocTemplate(buffer, pagesize=(w_pt, h_pt), leftMargin=0,rightMargin=0,topMargin=0,bottomMargin=0).build([ReportLabImage(p, width=w_pt, height=h_pt)])
|
354 |
-
buffer.seek(0); merger.append(fileobj=buffer)
|
355 |
-
except Exception as e: log += f"โ ๏ธ Failed to process image {p.name}: {e}\n"
|
356 |
-
for p in pdf_files:
|
357 |
-
try: merger.append(str(p))
|
358 |
-
except Exception as e: log += f"โ ๏ธ Failed to merge PDF {p.name}: {e}\n"
|
359 |
-
|
360 |
-
if len(merger.pages) > 0:
|
361 |
-
filename = f"Scroll_{layout.replace(' ','')}_{font}_x{num_columns}_{datetime.datetime.now().strftime('%H%M%S')}.pdf"
|
362 |
-
out_path = OUTPUT_DIR / filename
|
363 |
-
with open(out_path, "wb") as f: merger.write(f)
|
364 |
-
generated_pdfs.append(out_path); log += f"โ
Alchemized: {filename}\n"
|
365 |
|
366 |
-
|
367 |
-
|
368 |
-
|
369 |
-
# --- Gradio UI Definition ---
|
370 |
-
AVAILABLE_FONTS = register_local_fonts()
|
371 |
-
|
372 |
-
def get_theme():
|
373 |
-
"""Dynamically selects a font for the theme to avoid warnings for missing fonts."""
|
374 |
-
desired = "MedievalSharp"
|
375 |
-
fallback = ("ui-sans-serif", "system-ui", "sans-serif")
|
376 |
-
font_family = fallback
|
377 |
-
|
378 |
-
# Check if the desired font is in the list of registered fonts
|
379 |
-
if any(desired in s for s in AVAILABLE_FONTS):
|
380 |
-
font_family = (gr.themes.GoogleFont(desired),) + fallback
|
381 |
-
elif AVAILABLE_FONTS:
|
382 |
-
# If desired font is not found, use the first available font as a better fallback
|
383 |
-
first_font = AVAILABLE_FONTS[0]
|
384 |
-
print(f"WARNING: '{desired}' font not found. Using '{first_font}' for UI theme instead.")
|
385 |
-
font_family = (gr.themes.GoogleFont(first_font),) + fallback
|
386 |
-
else:
|
387 |
-
# Absolute fallback if no custom fonts are found
|
388 |
-
print(f"WARNING: No custom fonts found. Using system default for UI.")
|
389 |
-
|
390 |
-
return gr.themes.Base(
|
391 |
-
primary_hue=gr.themes.colors.purple,
|
392 |
-
secondary_hue=gr.themes.colors.indigo,
|
393 |
-
neutral_hue=gr.themes.colors.gray,
|
394 |
-
font=font_family
|
395 |
-
).set(
|
396 |
-
body_background_fill="#111827",
|
397 |
-
body_text_color="#d1d5db",
|
398 |
-
button_primary_background_fill="#a855f7",
|
399 |
-
button_primary_text_color="#ffffff",
|
400 |
-
button_secondary_background_fill="#6366f1",
|
401 |
-
button_secondary_text_color="#ffffff",
|
402 |
-
block_background_fill="#1f2937",
|
403 |
-
block_label_background_fill="#1f2937",
|
404 |
-
block_title_text_color="#a855f7",
|
405 |
-
input_background_fill="#374151"
|
406 |
-
)
|
407 |
-
|
408 |
-
with gr.Blocks(theme=get_theme(), title="The PDF Alchemist") as demo:
|
409 |
-
gr.Markdown("# โจ The PDF Alchemist โจ")
|
410 |
-
gr.Markdown("A single-page grimoire to turn your ideas into beautifully crafted PDF scrolls. Use the power of AI or upload your own treasures.")
|
411 |
-
|
412 |
-
with gr.Row(equal_height=False):
|
413 |
-
# --- LEFT COLUMN: INPUTS & CONTROLS ---
|
414 |
-
with gr.Column(scale=2):
|
415 |
-
with gr.Group():
|
416 |
-
with gr.Accordion("๐ Content Crucible (Your Ingredients)", open=True):
|
417 |
-
gr.Markdown("### ๐ค Command Your Idea Golem")
|
418 |
-
ai_prompt = gr.Textbox(label="Incantation (Prompt)", placeholder="e.g., 'A recipe for a dragon's breath chili...'")
|
419 |
-
generate_ai_btn = gr.Button("๐ง Animate Golem!")
|
420 |
-
ai_content_output = gr.Textbox(label="Golem's Manuscript (Editable)", lines=10, interactive=True, value="# The Golem awaits your command!\n\n")
|
421 |
-
gr.Markdown("<hr style='border-color: #374151; margin-top: 20px; margin-bottom: 20px;'>")
|
422 |
-
gr.Markdown("### ๐ค Add Your Physical Treasures")
|
423 |
-
uploaded_files = gr.File(label="Upload Files (TXT, MD, Code, Images, PDFs)", file_count="multiple", file_types=['text', '.md', 'image', '.pdf', '.py', '.js', '.html', '.css', '.json'])
|
424 |
-
|
425 |
-
with gr.Accordion("๐ Arcane Blueprints (Layout & Structure)", open=True):
|
426 |
-
selected_layouts = gr.CheckboxGroup(choices=list(LAYOUTS.keys()), label="Page Layouts", value=["A4 Portrait"])
|
427 |
-
num_columns_slider = gr.Slider(label="Number of Text Columns", minimum=1, maximum=4, step=1, value=1)
|
428 |
-
header_input = gr.Textbox(label="Header Inscription", value="[Title]", placeholder="e.g., Arcane Folio - [Page #]")
|
429 |
-
footer_input = gr.Textbox(label="Footer Inscription", value="Page [Page #] of [Total Pages]", placeholder="e.g., Top Secret - Page [Page #]")
|
430 |
|
431 |
-
|
432 |
-
|
433 |
-
|
434 |
-
|
435 |
-
|
436 |
-
|
437 |
-
|
438 |
-
|
439 |
-
|
440 |
-
|
441 |
-
|
442 |
-
|
443 |
-
with gr.Row():
|
444 |
-
margin_left_slider = gr.Slider(label="Margin Left (in)", minimum=0.25, maximum=1.5, step=0.05, value=0.75)
|
445 |
-
margin_right_slider = gr.Slider(label="Margin Right (in)", minimum=0.25, maximum=1.5, step=0.05, value=0.75)
|
446 |
-
|
447 |
-
generate_pdfs_btn = gr.Button("๐ฎ Alchemize PDF!", variant="primary", size="lg")
|
448 |
-
|
449 |
-
# --- RIGHT COLUMN: STAGING & OUTPUTS ---
|
450 |
-
with gr.Column(scale=3):
|
451 |
-
with gr.Group():
|
452 |
-
with gr.Accordion("๐ Staging Area (Your Uploaded Treasures)", open=True):
|
453 |
-
text_gallery = gr.Gallery(label="๐ Manuscripts & Code", show_label=True, columns=4, height=120, object_fit="contain")
|
454 |
-
image_gallery = gr.Gallery(label="๐ผ๏ธ Images & Glyphs", show_label=True, columns=4, height=120, object_fit="contain")
|
455 |
-
pdf_gallery = gr.Gallery(label="๐ Imported Scrolls", show_label=True, columns=4, height=120, object_fit="contain")
|
456 |
-
|
457 |
-
with gr.Group():
|
458 |
-
gr.Markdown("### โ๏ธ Transmuted Scrolls (Final PDFs)")
|
459 |
-
final_gallery_output = gr.Gallery(label="PDF Previews", show_label=False, elem_id="gallery", columns=2, height=400, object_fit="contain")
|
460 |
-
log_output = gr.Markdown(label="Alchemist's Log", value="Your log of successful transmutations will appear here...")
|
461 |
-
downloadable_files_output = gr.Files(label="Collect Your Scrolls")
|
462 |
|
463 |
-
#
|
464 |
-
|
465 |
-
margin_inputs = [margin_top_slider, margin_bottom_slider, margin_left_slider, margin_right_slider]
|
466 |
-
|
467 |
-
uploaded_files.upload(update_staging_and_manuscript, inputs=[uploaded_files, ai_content_output], outputs=[text_gallery, image_gallery, pdf_gallery, ai_content_output])
|
468 |
-
generate_ai_btn.click(generate_ai_content_api, inputs=[ai_prompt], outputs=[ai_content_output])
|
469 |
|
470 |
-
|
471 |
-
inputs=[uploaded_files, ai_content_output, selected_layouts, selected_fonts, num_columns_slider, header_input, footer_input] + font_size_inputs + margin_inputs + [use_svg_engine_toggle],
|
472 |
-
outputs=[final_gallery_output, log_output, downloadable_files_output])
|
473 |
|
474 |
-
|
|
|
|
|
475 |
|
|
|
476 |
if __name__ == "__main__":
|
477 |
-
|
478 |
-
|
479 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
5 |
import os
|
6 |
import shutil
|
7 |
import fitz # PyMuPDF
|
8 |
+
from PIL import Image
|
|
|
9 |
import io
|
|
|
|
|
10 |
|
11 |
+
from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, Table, TableStyle, PageBreak, BaseDocTemplate, Frame, PageTemplate, Image as ReportLabImage
|
|
|
|
|
|
|
|
|
12 |
from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
|
13 |
+
from reportlab.lib.pagesizes import letter
|
14 |
from reportlab.lib.units import inch
|
15 |
from reportlab.lib import colors
|
16 |
from reportlab.pdfbase import pdfmetrics
|
17 |
from reportlab.pdfbase.ttfonts import TTFont
|
|
|
18 |
|
19 |
+
# --- Configuration ---
|
20 |
CWD = Path.cwd()
|
21 |
+
# Create dedicated directories for clarity
|
22 |
+
INPUT_DIR = CWD / "uploaded_files"
|
23 |
+
OUTPUT_DIR = CWD / "output_pdfs"
|
24 |
+
TEMP_DIR = CWD / "temp_emoji_images"
|
25 |
+
FONT_DIR = CWD # Assumes fonts are in the same directory as the script
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
|
27 |
+
# Ensure all directories exist
|
28 |
+
for d in [INPUT_DIR, OUTPUT_DIR, TEMP_DIR]:
|
29 |
+
d.mkdir(exist_ok=True)
|
30 |
|
31 |
# --- Font & Emoji Handling ---
|
32 |
+
EMOJI_FONT_PATH = None
|
33 |
+
EMOJI_IMAGE_CACHE = {}
|
34 |
|
35 |
+
def setup_fonts():
|
36 |
+
"""Finds the NotoColorEmoji font, which is critical for this process."""
|
37 |
+
global EMOJI_FONT_PATH
|
38 |
+
print("--- Setting up fonts ---")
|
|
|
39 |
|
40 |
+
# Locate the essential NotoColorEmoji font
|
41 |
noto_emoji_path = FONT_DIR / "NotoColorEmoji-Regular.ttf"
|
42 |
if noto_emoji_path.exists():
|
43 |
EMOJI_FONT_PATH = str(noto_emoji_path)
|
44 |
+
print(f"Found emoji font: {EMOJI_FONT_PATH}")
|
45 |
else:
|
46 |
+
print("CRITICAL ERROR: 'NotoColorEmoji-Regular.ttf' not found in the application directory.")
|
47 |
+
print("This file is required to render emojis as images. Please add it to the directory.")
|
48 |
+
# Raise an error to stop the app from running in a broken state
|
49 |
+
raise FileNotFoundError("Could not find NotoColorEmoji-Regular.ttf. The application cannot proceed.")
|
50 |
+
|
51 |
+
# Register a basic font for ReportLab
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
52 |
try:
|
53 |
+
pdfmetrics.registerFont(TTFont('DejaVuSans', 'DejaVuSans.ttf'))
|
54 |
+
pdfmetrics.registerFontFamily('DejaVuSans', normal='DejaVuSans', bold='DejaVuSans-Bold', italic='DejaVuSans-Oblique', boldItalic='DejaVuSans-BoldOblique')
|
55 |
+
except:
|
56 |
+
print("Warning: DejaVuSans font not found. Falling back to Helvetica. Please add DejaVuSans.ttf for better Unicode support.")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
57 |
|
58 |
+
def render_emoji_as_image(emoji_char, size_pt):
|
|
|
59 |
"""
|
60 |
+
Takes a single emoji character and renders it as a transparent PNG image in memory.
|
61 |
+
This is the core of the solution to ensure emojis appear in color in any PDF viewer.
|
62 |
"""
|
63 |
+
if not EMOJI_FONT_PATH:
|
64 |
+
print("Cannot render emoji: Emoji font path not set.")
|
65 |
+
return None
|
66 |
+
|
67 |
+
# Use a cache to avoid re-rendering the same emoji multiple times
|
68 |
+
if (emoji_char, size_pt) in EMOJI_IMAGE_CACHE:
|
69 |
+
return EMOJI_IMAGE_CACHE[(emoji_char, size_pt)]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
70 |
|
|
|
|
|
|
|
|
|
|
|
71 |
try:
|
72 |
+
# Use PyMuPDF (fitz) to draw the emoji onto a temporary, transparent canvas
|
73 |
+
rect = fitz.Rect(0, 0, size_pt * 1.5, size_pt * 1.5)
|
74 |
+
doc = fitz.open()
|
75 |
+
page = doc.new_page(width=rect.width, height=rect.height)
|
76 |
|
77 |
+
# Load the color emoji font
|
78 |
+
page.insert_font(fontname="emoji", fontfile=EMOJI_FONT_PATH)
|
79 |
+
|
80 |
+
# Insert the emoji character. The vertical alignment may need tweaking.
|
81 |
+
page.insert_text(fitz.Point(0, size_pt * 1.1), emoji_char, fontname="emoji", fontsize=size_pt)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
82 |
|
83 |
+
# Get a high-resolution PNG of the emoji with a transparent background
|
84 |
+
pix = page.get_pixmap(alpha=True, dpi=300)
|
85 |
+
doc.close()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
86 |
|
87 |
+
# Save the PNG to an in-memory buffer
|
88 |
+
img_buffer = io.BytesIO(pix.tobytes("png"))
|
89 |
+
img_buffer.seek(0)
|
|
|
|
|
|
|
90 |
|
91 |
+
# Add the buffer to the cache and return it
|
92 |
+
EMOJI_IMAGE_CACHE[(emoji_char, size_pt)] = img_buffer
|
93 |
+
return img_buffer
|
94 |
+
except Exception as e:
|
95 |
+
print(f"ERROR: Could not render emoji '{emoji_char}': {e}")
|
96 |
+
return None
|
97 |
|
98 |
+
# --- PDF Generation ---
|
99 |
+
def create_pdf_from_markdown(md_filepath: Path):
|
100 |
+
"""
|
101 |
+
The main function to convert a single Markdown file into a PDF.
|
102 |
+
It reads the text, processes it line by line, and replaces emojis with images.
|
103 |
+
"""
|
104 |
+
print(f"--- Starting PDF conversion for: {md_filepath.name} ---")
|
105 |
|
106 |
+
# Define styles for the PDF document
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
107 |
styles = getSampleStyleSheet()
|
108 |
+
# Use a font that supports a wide range of characters, if available
|
109 |
+
body_font = 'DejaVuSans' if 'DejaVuSans' in pdfmetrics.getRegisteredFontNames() else 'Helvetica'
|
110 |
+
style_body = ParagraphStyle('Body', fontName=body_font, fontSize=11, leading=14)
|
111 |
+
style_h1 = ParagraphStyle('H1', fontName=body_font, fontSize=24, leading=28, spaceAfter=12, textColor=colors.darkblue)
|
112 |
+
style_h2 = ParagraphStyle('H2', fontName=body_font, fontSize=18, leading=22, spaceAfter=10)
|
113 |
+
|
114 |
+
# Regex to find all emojis in a string
|
|
|
115 |
emoji_pattern = re.compile(f"([{re.escape(''.join(map(chr, range(0x1f600, 0x1f650))))}"
|
116 |
f"{re.escape(''.join(map(chr, range(0x1f300, 0x1f5ff))))}"
|
117 |
f"{re.escape(''.join(map(chr, range(0x1f900, 0x1f9ff))))}"
|
118 |
+
f"{re.escape(''.join(map(chr, range(0x2600, 0x26ff))))}]+)")
|
119 |
+
|
120 |
+
def text_to_flowables(text, style):
|
121 |
+
"""
|
122 |
+
Splits a line of text into a list of Paragraphs and Images.
|
123 |
+
This allows text and emoji-images to flow together on the same line.
|
124 |
+
"""
|
125 |
+
parts = emoji_pattern.split(text)
|
126 |
+
flowables = []
|
127 |
for part in parts:
|
128 |
if not part: continue
|
129 |
+
|
130 |
if emoji_pattern.match(part):
|
131 |
+
# This part is an emoji or a sequence of them
|
132 |
for emoji_char in part:
|
133 |
+
img_buffer = render_emoji_as_image(emoji_char, style.fontSize)
|
134 |
+
if img_buffer:
|
135 |
+
# Create an Image object, slightly larger than the text for better spacing
|
136 |
+
img = ReportLabImage(img_buffer, height=style.fontSize * 1.2, width=style.fontSize * 1.2)
|
137 |
+
flowables.append(img)
|
138 |
else:
|
139 |
+
# This part is plain text
|
140 |
+
flowables.append(Paragraph(part.replace('&', '&').replace('<', '<').replace('>', '>'), style))
|
141 |
+
|
142 |
+
# Use a Table to keep all parts on the same line. This is a common ReportLab technique.
|
143 |
+
if flowables:
|
144 |
+
return Table([flowables], colWidths=[None] * len(flowables), style=[('VALIGN', (0,0), (-1,-1), 'MIDDLE')])
|
145 |
+
return None
|
|
|
|
|
|
|
146 |
|
147 |
+
# Read the markdown file
|
148 |
+
try:
|
149 |
+
with open(md_filepath, 'r', encoding='utf-8') as f:
|
150 |
+
lines = f.readlines()
|
151 |
+
except Exception as e:
|
152 |
+
print(f"ERROR: Could not read file {md_filepath.name}: {e}")
|
153 |
+
return None
|
154 |
+
|
155 |
+
# The "story" is ReportLab's list of things to draw in the PDF
|
156 |
+
story = []
|
157 |
for line in lines:
|
158 |
stripped_line = line.strip()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
159 |
|
160 |
+
# Simple Markdown parsing
|
161 |
+
if stripped_line.startswith('# '):
|
162 |
+
flowable = text_to_flowables(stripped_line[2:], style_h1)
|
163 |
+
elif stripped_line.startswith('## '):
|
164 |
+
flowable = text_to_flowables(stripped_line[2:], style_h2)
|
165 |
+
elif stripped_line:
|
166 |
+
flowable = text_to_flowables(stripped_line, style_body)
|
167 |
+
else:
|
168 |
+
flowable = Spacer(1, 0.2 * inch)
|
169 |
+
|
170 |
+
if flowable:
|
171 |
+
story.append(flowable)
|
|
|
172 |
|
173 |
+
# Generate a unique filename and path for the output PDF
|
174 |
+
timestamp = datetime.datetime.now().strftime("%Y-%m-%d_%H%M")
|
175 |
+
output_filename = f"{md_filepath.stem}_{timestamp}.pdf"
|
176 |
+
output_filepath = OUTPUT_DIR / output_filename
|
|
|
|
|
|
|
|
|
|
|
|
|
177 |
|
178 |
+
# Build the PDF document
|
179 |
+
doc = SimpleDocTemplate(str(output_filepath), pagesize=letter)
|
180 |
+
try:
|
181 |
+
doc.build(story)
|
182 |
+
print(f"SUCCESS: Successfully created PDF: {output_filename}")
|
183 |
+
return output_filepath
|
184 |
+
except Exception as e:
|
185 |
+
print(f"ERROR: Failed to build PDF for {md_filepath.name}. Reason: {e}")
|
186 |
+
return None
|
187 |
+
|
188 |
+
# --- Gradio UI and Logic ---
|
189 |
+
def process_uploads(files):
|
190 |
+
"""
|
191 |
+
Takes uploaded files, processes each one into a PDF, and returns a list of generated filepaths.
|
192 |
+
"""
|
193 |
+
if not files:
|
194 |
+
raise gr.Error("Please upload at least one Markdown (.md) file.")
|
195 |
|
196 |
+
# Clear caches and temp directories for a clean run
|
197 |
+
EMOJI_IMAGE_CACHE.clear()
|
198 |
+
shutil.rmtree(TEMP_DIR, ignore_errors=True); TEMP_DIR.mkdir(exist_ok=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
199 |
|
200 |
+
log_messages = []
|
201 |
+
generated_pdf_paths = []
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
202 |
|
203 |
+
for file_obj in files:
|
204 |
+
input_path = Path(file_obj.name)
|
205 |
+
log_messages.append(f"Processing '{input_path.name}'...")
|
206 |
+
|
207 |
+
# Core PDF creation step
|
208 |
+
output_path = create_pdf_from_markdown(input_path)
|
209 |
+
|
210 |
+
if output_path:
|
211 |
+
generated_pdf_paths.append(str(output_path))
|
212 |
+
log_messages.append(f"โ
Success! PDF saved to '{output_path.name}'.")
|
213 |
+
else:
|
214 |
+
log_messages.append(f"โ Failed to process '{input_path.name}'. Check console for errors.")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
215 |
|
216 |
+
# After processing, get the full list of all PDFs in the output directory for the gallery
|
217 |
+
all_pdfs_in_gallery = sorted([str(p) for p in OUTPUT_DIR.glob("*.pdf")], reverse=True)
|
|
|
|
|
|
|
|
|
218 |
|
219 |
+
return "\n".join(log_messages), all_pdfs_in_gallery
|
|
|
|
|
220 |
|
221 |
+
def refresh_gallery():
|
222 |
+
"""Scans the output directory and returns a list of all PDFs found."""
|
223 |
+
return sorted([str(p) for p in OUTPUT_DIR.glob("*.pdf")], reverse=True)
|
224 |
|
225 |
+
# Main execution block
|
226 |
if __name__ == "__main__":
|
227 |
+
# This must run once at startup to check for the required font
|
228 |
+
try:
|
229 |
+
setup_fonts()
|
230 |
+
except FileNotFoundError as e:
|
231 |
+
# If the font is missing, we stop the app from launching.
|
232 |
+
print("\n" + "="*60)
|
233 |
+
print(e)
|
234 |
+
print("The application cannot start without this font file.")
|
235 |
+
print("Please add 'NotoColorEmoji-Regular.ttf' and 'DejaVuSans.ttf' to your project directory.")
|
236 |
+
print("="*60)
|
237 |
+
exit() # Stop the script
|
238 |
+
|
239 |
+
# Define the Gradio Interface
|
240 |
+
with gr.Blocks(theme=gr.themes.Soft(primary_hue="blue"), title="Markdown-to-PDF Alchemist") as demo:
|
241 |
+
gr.Markdown("# ๐ Markdown-to-PDF Alchemist")
|
242 |
+
gr.Markdown("Upload one or more `.md` files. This tool will convert them into PDFs, preserving emojis by rendering them as high-quality images. All generated PDFs will appear in the library below.")
|
243 |
+
|
244 |
+
with gr.Row():
|
245 |
+
with gr.Column(scale=1):
|
246 |
+
upload_button = gr.File(
|
247 |
+
label="Upload Markdown Files (.md)",
|
248 |
+
file_count="multiple",
|
249 |
+
file_types=[".md"],
|
250 |
+
)
|
251 |
+
generate_button = gr.Button("๐ฎ Alchemize to PDF", variant="primary")
|
252 |
+
log_output = gr.Textbox(label="Alchemist's Log", lines=8, interactive=False)
|
253 |
+
|
254 |
+
with gr.Column(scale=2):
|
255 |
+
gr.Markdown("### ๐ Generated PDF Library")
|
256 |
+
# The gallery will show the first page of the PDF as a preview
|
257 |
+
pdf_gallery = gr.Gallery(
|
258 |
+
label="Generated PDFs",
|
259 |
+
show_label=False,
|
260 |
+
elem_id="gallery",
|
261 |
+
columns=3,
|
262 |
+
object_fit="contain",
|
263 |
+
height="auto"
|
264 |
+
)
|
265 |
+
# This button allows manual refreshing of the gallery
|
266 |
+
refresh_button = gr.Button("๐ Refresh Library")
|
267 |
+
|
268 |
+
# Define the actions when buttons are clicked
|
269 |
+
generate_button.click(
|
270 |
+
fn=process_uploads,
|
271 |
+
inputs=[upload_button],
|
272 |
+
outputs=[log_output, pdf_gallery]
|
273 |
+
)
|
274 |
+
|
275 |
+
refresh_button.click(
|
276 |
+
fn=refresh_gallery,
|
277 |
+
inputs=None,
|
278 |
+
outputs=[pdf_gallery]
|
279 |
+
)
|
280 |
+
|
281 |
+
# Load the gallery with existing PDFs when the app starts
|
282 |
+
demo.load(refresh_gallery, None, pdf_gallery)
|
283 |
+
|
284 |
+
# Launch the application
|
285 |
+
demo.launch(debug=True)
|