#!/usr/bin/env python3
"""
Parse Lyngdorf PDF manuals and extract structured information.
Requires: pip install pymupdf (PyMuPDF)
"""
import fitz # PyMuPDF
import json
import re
import sys
from pathlib import Path
def extract_text_from_pdf(pdf_path):
"""Extract all text from PDF."""
doc = fitz.open(pdf_path)
text = ""
for page in doc:
text += page.get_text()
doc.close()
return text
def parse_commands(text):
"""Extract command information from manual text."""
commands = {}
# Pattern for commands: !COMMAND or !COMMAND(params)
command_pattern = r'(![\w]+(?:\([^\)]*\))?)'
# Split into lines and look for command documentation
lines = text.split('\n')
for i, line in enumerate(lines):
# Look for lines that contain command patterns
matches = re.findall(command_pattern, line)
for cmd in matches:
# Extract command name
cmd_name = cmd.split('(')[0]
# Get context (next few lines)
context_lines = []
for j in range(i, min(i + 3, len(lines))):
if lines[j].strip():
context_lines.append(lines[j].strip())
context = ' '.join(context_lines)
# Store command with context
if cmd_name not in commands:
commands[cmd_name] = {
'command': cmd,
'description': context[:200] # First 200 chars
}
return commands
def extract_sections(text):
"""Extract major sections from the manual."""
sections = {
'introduction': '',
'connection': '',
'commands': '',
'troubleshooting': '',
'examples': ''
}
# Simple section extraction based on common headings
current_section = 'introduction'
section_text = []
for line in text.split('\n'):
line_lower = line.lower()
# Detect section headers
if any(word in line_lower for word in ['connection', 'network', 'tcp']):
if len(section_text) > 10:
sections[current_section] = '\n'.join(section_text)
current_section = 'connection'
section_text = []
elif any(word in line_lower for word in ['command', 'control']):
if len(section_text) > 10:
sections[current_section] = '\n'.join(section_text)
current_section = 'commands'
section_text = []
elif any(word in line_lower for word in ['troubleshoot', 'problem', 'issue']):
if len(section_text) > 10:
sections[current_section] = '\n'.join(section_text)
current_section = 'troubleshooting'
section_text = []
elif any(word in line_lower for word in ['example', 'sample']):
if len(section_text) > 10:
sections[current_section] = '\n'.join(section_text)
current_section = 'examples'
section_text = []
section_text.append(line)
# Save last section
if section_text:
sections[current_section] = '\n'.join(section_text)
return sections
def main():
if len(sys.argv) < 2:
print("Usage: python parse_manual.py <pdf_file> [output_dir]")
sys.exit(1)
pdf_path = Path(sys.argv[1])
output_dir = Path(sys.argv[2]) if len(sys.argv) > 2 else Path('docs/kb')
output_dir.mkdir(parents=True, exist_ok=True)
print(f"Parsing {pdf_path}...")
# Extract text
text = extract_text_from_pdf(pdf_path)
# Parse commands
commands = parse_commands(text)
print(f"Found {len(commands)} commands")
# Extract sections
sections = extract_sections(text)
# Determine model from filename
model = 'Unknown'
if 'TDAI-1120' in pdf_path.name or '1120' in pdf_path.name:
model = 'TDAI-1120'
elif 'TDAI-2170' in pdf_path.name or '2170' in pdf_path.name:
model = 'TDAI-2170'
elif 'TDAI-3400' in pdf_path.name or '3400' in pdf_path.name:
model = 'TDAI-3400'
# Save parsed data
output = {
'model': model,
'commands': commands,
'sections': sections,
'full_text': text
}
output_file = output_dir / f'{model}.json'
with open(output_file, 'w') as f:
json.dump(output, f, indent=2)
print(f"Saved parsed data to {output_file}")
# Also save individual sections as markdown
for section_name, section_content in sections.items():
if section_content.strip():
section_file = output_dir / f'{model}-{section_name}.md'
with open(section_file, 'w') as f:
f.write(f"# {model} - {section_name.title()}\n\n")
f.write(section_content)
print(f"Saved {section_name} to {section_file}")
if __name__ == '__main__':
main()
