Skip to content

Support batch page adding #3

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
promentol wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Support batch page adding #3

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
142 changes: 103 additions & 39 deletions nano_pdf/main.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -164,17 +164,16 @@ def process_single_page(page_num: int, prompt_text: str):
@app.command()
def add(
pdf_path: str = typer.Argument(..., help="Path to the PDF file"),
after_page: int = typer.Argument(..., help="Insert after this page number (0 for beginning)"),
prompt: str = typer.Argument(..., help="Description of the new slide to create"),
adds: List[str] = typer.Argument(..., help="Pairs of 'AfterPage Prompt' (e.g. 0 'Title slide' 2 'Summary slide')"),
style_refs: Optional[str] = typer.Option(None, help="Comma-separated list of reference page numbers for style (e.g. '1,2'). Defaults to first page."),
use_context: bool = typer.Option(True, help="Include full PDF text as context (enabled by default for better slide generation)"),
output: Optional[str] = typer.Option(None, help="Output path for the PDF. Defaults to 'edited_<filename>'"),
resolution: str = typer.Option("4K", help="Image resolution: '4K', '2K', '1K' (higher = better quality but slower)"),
disable_google_search: bool = typer.Option(False, help="Disable Google Search (enabled by default)")
):
"""
Add a new slide to a PDF using AI generation.
Usage: nano-pdf add deck.pdf 0 "Title slide with 'Welcome to Q3 Review'"
Add new slide(s) to a PDF using AI generation.
Usage: nano-pdf add deck.pdf 0 "Title slide" 2 "Summary slide"
"""
# Check system dependencies first
try:
Expand All @@ -191,13 +190,37 @@ def add(
if not output:
output = f"edited_{input_path.name}"

# Validate after_page
total_pages = pdf_utils.get_page_count(str(input_path))
if after_page < 0 or after_page > total_pages:
typer.echo(f"Error: after_page must be between 0 and {total_pages}. Use 0 to insert at the beginning.")
# Parse Adds
if len(adds) % 2 != 0:
typer.echo("Error: Adds must be pairs of 'AfterPage Prompt'.")
raise typer.Exit(code=1)

typer.echo(f"Adding new slide to {pdf_path} after page {after_page}...")
parsed_adds = []
for i in range(0, len(adds), 2):
try:
after_page = int(adds[i])
prompt = adds[i+1]
parsed_adds.append((after_page, prompt))
except ValueError:
typer.echo(f"Error: Invalid page number '{adds[i]}'")
raise typer.Exit(code=1)

# Validate after_page values
# Sort by after_page to validate sequentially (pages added earlier increase the valid range for later ones)
total_pages = pdf_utils.get_page_count(str(input_path))
sorted_adds = sorted(parsed_adds, key=lambda x: x[0])

for idx, (after_page, _) in enumerate(sorted_adds):
# Each previously added page increases the max valid position by 1
max_valid_position = total_pages + idx
if after_page < 0 or after_page > max_valid_position:
typer.echo(f"Error: Invalid after_page value {after_page}. Must be between 0 and {max_valid_position} (considering {idx} page(s) added before it).")
raise typer.Exit(code=1)

# Use sorted order for processing
parsed_adds = sorted_adds

typer.echo(f"Adding {len(parsed_adds)} new slide(s) to {pdf_path}...")

# Extract text context
full_text = ""
Expand Down Expand Up @@ -231,45 +254,86 @@ def add(
except Exception as e:
typer.echo(f"Warning: Could not render Page 1: {e}")

# Generate the new slide
typer.echo("Generating new slide with AI...")
try:
generated_image, response_text = ai_utils.generate_new_slide(
style_reference_images=style_images,
user_prompt=prompt,
full_text_context=full_text,
resolution=resolution,
enable_search=not disable_google_search
)
except Exception as e:
typer.echo(f"Error generating slide: {e}")
raise typer.Exit(code=1)
# Generate new slides (Parallel)
typer.echo(f"Generating {len(parsed_adds)} slide(s) with AI in parallel...")
generated_slides = {} # after_page -> temp_pdf_path
temp_files = []

def process_single_slide(after_page: int, prompt_text: str):
typer.echo(f"Starting slide for insertion after page {after_page}...")
try:
generated_image, response_text = ai_utils.generate_new_slide(
style_reference_images=style_images,
user_prompt=prompt_text,
full_text_context=full_text,
resolution=resolution,
enable_search=not disable_google_search
)

# Print model's text response if any
if response_text:
typer.echo(f"Model response for slide after page {after_page}: {response_text}")

# Print model's text response if any
if response_text:
typer.echo(f"Model response: {response_text}")
# Re-hydrate to PDF
temp_pdf_file = tempfile.NamedTemporaryFile(mode='wb', suffix='.pdf', delete=False)
temp_pdf = temp_pdf_file.name
temp_pdf_file.close()
pdf_utils.rehydrate_image_to_pdf(generated_image, temp_pdf)

# Re-hydrate to PDF
typer.echo("Converting to PDF with text layer...")
temp_pdf_file = tempfile.NamedTemporaryFile(mode='wb', suffix='.pdf', delete=False)
temp_pdf = temp_pdf_file.name
temp_pdf_file.close()
typer.echo(f"Finished slide for insertion after page {after_page}")
return (after_page, temp_pdf)
except Exception as e:
typer.echo(f"Error generating slide for insertion after page {after_page}: {e}")
return None

try:
pdf_utils.rehydrate_image_to_pdf(generated_image, temp_pdf)
completed_count = 0
with concurrent.futures.ThreadPoolExecutor(max_workers=1) as executor:
futures = [executor.submit(process_single_slide, after_page, prompt) for after_page, prompt in parsed_adds]

# Insert into the PDF
typer.echo("Inserting slide into PDF...")
pdf_utils.insert_page(str(input_path), temp_pdf, after_page, output)
for future in concurrent.futures.as_completed(futures):
result = future.result()
if result:
after_page, temp_pdf = result
generated_slides[after_page] = temp_pdf
temp_files.append(temp_pdf)
completed_count += 1
typer.echo(f"Progress: {completed_count}/{len(parsed_adds)} slides completed")

if not generated_slides:
typer.echo("No slides were successfully generated.")
raise typer.Exit(code=1)

# Insert all slides into the PDF
typer.echo(f"\nInserting {len(generated_slides)} slide(s) into PDF...")
try:
# Sort by after_page in ASCENDING order for sequential additions
# Each insertion increases the page count, allowing the next sequential position to exist
sorted_adds = sorted(generated_slides.items(), key=lambda x: x[0], reverse=False)

current_pdf = str(input_path)
for i, (after_page, temp_pdf) in enumerate(sorted_adds):
if i == len(sorted_adds) - 1:
# Last insertion, write to final output
pdf_utils.insert_page(current_pdf, temp_pdf, after_page, output)
else:
# Intermediate insertion, write to temp file
temp_intermediate = tempfile.NamedTemporaryFile(mode='wb', suffix='.pdf', delete=False)
intermediate_path = temp_intermediate.name
temp_intermediate.close()
temp_files.append(intermediate_path)

pdf_utils.insert_page(current_pdf, temp_pdf, after_page, intermediate_path)
current_pdf = intermediate_path
except Exception as e:
typer.echo(f"Error creating PDF: {e}")
typer.echo(f"Error inserting slides: {e}")
raise typer.Exit(code=1)
finally:
# Cleanup
if Path(temp_pdf).exists():
Path(temp_pdf).unlink()
for f in temp_files:
if Path(f).exists():
Path(f).unlink()

typer.echo(f"Done! New slide added after page {after_page}. Saved to {output}")
typer.echo(f"Done! Added {len(generated_slides)} new slide(s). Saved to {output}")

@app.command()
def version():
Expand Down