#!/usr/bin/env python3 """ PDF Translation Script Translates a Spanish PDF document to English and creates a new PDF file. """ import os import sys import glob from pathlib import Path try: import PyPDF2 except ImportError: print("Installing PyPDF2...") os.system("pip3 install PyPDF2") import PyPDF2 try: from fpdf import FPDF except ImportError: print("Installing fpdf2...") os.system("pip3 install fpdf2") from fpdf import FPDF try: from deep_translator import GoogleTranslator except ImportError: print("Installing deep-translator...") os.system("pip3 install deep-translator") from deep_translator import GoogleTranslator import re def extract_text_from_pdf(pdf_path): """Extract text from a PDF file, preserving page structure.""" pages_content = [] try: with open(pdf_path, 'rb') as file: pdf_reader = PyPDF2.PdfReader(file) num_pages = len(pdf_reader.pages) print(f"Found {num_pages} pages in {pdf_path}") for page_num in range(num_pages): page = pdf_reader.pages[page_num] text = page.extract_text() if text.strip(): pages_content.append({ 'page_num': page_num + 1, 'text': text }) print(f"Extracted {len(text)} characters from page {page_num + 1}") return pages_content except Exception as e: print(f"Error reading PDF: {e}") return None def detect_spanish(text): """Simple detection of Spanish text based on common Spanish words.""" spanish_indicators = [ 'el', 'la', 'los', 'las', 'de', 'que', 'y', 'a', 'en', 'un', 'una', 'es', 'se', 'no', 'te', 'lo', 'le', 'da', 'su', 'por', 'son', 'con', 'para', 'del', 'una', 'las', 'más', 'sus', 'ser', 'tiene', 'haber', 'este', 'desde', 'está', 'mismo', 'bajo', 'según', 'mejor', 'después', 'donde', 'mientras', 'estado', 'parte', 'estos', 'también', 'entre', 'tanto', 'durante', 'están', 'siempre', 'quien', 'todas', 'puede', 'hacer', 'hemos', 'hasta', 'año', 'años', 'día', 'días' ] text_lower = text.lower() spanish_word_count = sum(1 for word in spanish_indicators if word in text_lower) # Check for Spanish-specific characters spanish_chars = ['ñ', 'á', 'é', 'í', 'ó', 'ú', 'ü', '¿', '¡'] has_spanish_chars = any(char in text for char in spanish_chars) # If we find many Spanish words or Spanish-specific characters, it's likely Spanish if spanish_word_count > 10 or has_spanish_chars: return True return False def translate_text_pages(pages_content, source_lang='es', target_lang='en'): """Translate text pages from Spanish to English, preserving structure.""" if not pages_content: return [] print(f"\nTranslating {len(pages_content)} pages...") translated_pages = [] translator = GoogleTranslator(source=source_lang, target=target_lang) for page_info in pages_content: page_num = page_info['page_num'] text = page_info['text'] print(f"Translating page {page_num} ({len(text)} characters)...") # Split text into chunks if needed chunk_size = 4500 if len(text) <= chunk_size: # Single chunk try: translated = translator.translate(text) translated_pages.append({ 'page_num': page_num, 'text': translated, 'original_lines': text.split('\n') }) except Exception as e: print(f"Error translating page {page_num}: {e}") translated_pages.append({ 'page_num': page_num, 'text': text, 'original_lines': text.split('\n') }) else: # Multiple chunks lines = text.split('\n') chunks = [] current_chunk = "" for line in lines: if len(current_chunk) + len(line) + 1 < chunk_size: if current_chunk: current_chunk += '\n' + line else: current_chunk = line else: if current_chunk: chunks.append(current_chunk) current_chunk = line if current_chunk: chunks.append(current_chunk) translated_chunks = [] for i, chunk in enumerate(chunks): try: translated_chunk = translator.translate(chunk) translated_chunks.append(translated_chunk) except Exception as e: print(f"Error translating chunk {i+1} of page {page_num}: {e}") translated_chunks.append(chunk) translated_pages.append({ 'page_num': page_num, 'text': '\n'.join(translated_chunks), 'original_lines': text.split('\n') }) return translated_pages def detect_line_type(line): """Detect if a line is a header, label, value, or regular text.""" line_stripped = line.strip() if not line_stripped: return 'empty' # Check for headers (ALL CAPS, short lines) if line_stripped.isupper() and len(line_stripped) < 80: return 'header' # Check for label-value pairs (contains colon followed by text) # This is the most common pattern in forms if ':' in line_stripped: parts = line_stripped.split(':', 1) label_part = parts[0].strip() value_part = parts[1].strip() if len(parts) > 1 else '' # If label is short and there's a value, it's a label-value pair if len(label_part) < 60 and len(label_part) > 0: return 'label_value' # Check for section headers (standalone short lines that aren't headers) if len(line_stripped) < 50 and not line_stripped.endswith('.') and not ':' in line_stripped: # Common section header patterns if any(word in line_stripped.lower() for word in ['model', 'registration', 'presenter', 'request']): return 'section_header' return 'regular' def safe_encode(text): """Safely encode text for FPDF.""" try: return text.encode('latin-1', 'replace').decode('latin-1') except: return text.encode('utf-8', 'ignore').decode('utf-8', 'ignore') def create_english_pdf(translated_pages, output_path, title="Translated Document"): """Create a PDF file with the translated English text, preserving formatting.""" print(f"\nCreating English PDF: {output_path}") # Create a PDF document with proper margins pdf = FPDF() pdf.set_auto_page_break(auto=True, margin=20) pdf.set_margins(left=20, top=20, right=20) first_page = True for page_info in translated_pages: if not first_page: pdf.add_page() else: pdf.add_page() first_page = False text = page_info['text'] lines = text.split('\n') for line in lines: line_stripped = line.strip() if not line_stripped: pdf.ln(5) continue line_type = detect_line_type(line_stripped) # Handle different line types with appropriate formatting if line_type == 'header': # Headers: Bold, larger font pdf.ln(4) pdf.set_font("Helvetica", "B", 14) safe_text = safe_encode(line_stripped) pdf.cell(0, 10, safe_text, 0, 1, 'L') pdf.ln(3) elif line_type == 'section_header': # Section headers: Bold, medium font pdf.ln(3) pdf.set_font("Helvetica", "B", 11) safe_text = safe_encode(line_stripped) pdf.cell(0, 8, safe_text, 0, 1, 'L') pdf.ln(2) elif line_type == 'label_value': # Label-value pairs: Label in bold, value in regular if ':' in line_stripped: parts = line_stripped.split(':', 1) label = parts[0].strip() + ':' value = parts[1].strip() if len(parts) > 1 else '' # Set font to bold for label and get width pdf.set_font("Helvetica", "B", 10) safe_label = safe_encode(label) label_width = pdf.get_string_width(safe_label) + 3 label_width = min(label_width, 85) # Cap at 85mm # Print label in bold (cell with no line break) pdf.cell(label_width, 6, safe_label, 0, 0, 'L') # Print value in regular font (multi_cell with line break) pdf.set_font("Helvetica", "", 10) if value: safe_value = safe_encode(value) pdf.multi_cell(0, 6, safe_value, 0, 'L') else: pdf.ln(6) else: pdf.set_font("Helvetica", "", 10) safe_text = safe_encode(line_stripped) pdf.multi_cell(0, 6, safe_text, 0, 'L') pdf.ln(2) else: # regular text pdf.set_font("Helvetica", "", 10) safe_text = safe_encode(line_stripped) pdf.multi_cell(0, 5, safe_text, 0, 'L') pdf.ln(2) # Save PDF pdf.output(output_path) print(f"English PDF created successfully: {output_path}") def find_spanish_pdfs(directory="."): """Find PDF files that might be in Spanish.""" pdf_files = glob.glob(os.path.join(directory, "*.pdf")) pdf_files.extend(glob.glob(os.path.join(directory, "**/*.pdf"), recursive=True)) # Remove duplicates pdf_files = list(set(pdf_files)) print(f"Found {len(pdf_files)} PDF files") spanish_pdfs = [] for pdf_path in pdf_files: print(f"\nChecking: {pdf_path}") pages_content = extract_text_from_pdf(pdf_path) if pages_content: # Combine all pages text for detection all_text = '\n\n'.join([page['text'] for page in pages_content]) # Check if it's likely Spanish if detect_spanish(all_text): print(f"✓ This appears to be a Spanish document: {pdf_path}") spanish_pdfs.append((pdf_path, pages_content)) else: print(f" (Does not appear to be Spanish)") else: print(f" (Could not extract text)") return spanish_pdfs def main(): """Main function to translate PDF.""" print("=" * 60) print("PDF Translation Tool - Spanish to English") print("=" * 60) # Check for command line argument if len(sys.argv) > 1: pdf_path = sys.argv[1] if not os.path.exists(pdf_path): print(f"Error: File not found: {pdf_path}") sys.exit(1) print(f"\nProcessing specified file: {pdf_path}") pages_content = extract_text_from_pdf(pdf_path) if not pages_content: print("Error: Could not extract text from PDF") sys.exit(1) # Translate translated_pages = translate_text_pages(pages_content) # Create output filename base_name = os.path.splitext(os.path.basename(pdf_path))[0] output_path = f"{base_name}_English.pdf" # Create English PDF create_english_pdf(translated_pages, output_path, title="Translated Document") print(f"\n✓ Translation complete!") print(f" Original: {pdf_path}") print(f" English version: {output_path}") else: # Find Spanish PDFs automatically print("\nSearching for Spanish PDF files...") spanish_pdfs = find_spanish_pdfs() if not spanish_pdfs: print("\nNo Spanish PDF files found. Please specify a PDF file:") print("Usage: python3 translate_spanish_pdf.py ") sys.exit(1) # If multiple found, use the first one or let user choose if len(spanish_pdfs) == 1: pdf_path, pages_content = spanish_pdfs[0] else: print(f"\nFound {len(spanish_pdfs)} Spanish PDF files:") for i, (path, _) in enumerate(spanish_pdfs, 1): print(f" {i}. {path}") # Use the first one (or the one from 2024 if we can identify it) pdf_path, pages_content = spanish_pdfs[0] print(f"\nProcessing: {pdf_path}") # Translate translated_pages = translate_text_pages(pages_content) # Create output filename base_name = os.path.splitext(os.path.basename(pdf_path))[0] output_path = f"{base_name}_English.pdf" # Create English PDF create_english_pdf(translated_pages, output_path, title="Translated Document") print(f"\n✓ Translation complete!") print(f" Original: {pdf_path}") print(f" English version: {output_path}") if __name__ == "__main__": main()