scripts/build_paper_pdf.py

import os
import shutil
import subprocess
import sys

def build_pdf():
    # Define directories
    # Script is in e:\health-tech\papers\vsa\scripts\
    # Papers are in e:\health-tech\papers\vsa\papers\
    # Assets are in e:\health-tech\papers\vsa\data\paper_assets\

    script_dir = os.path.dirname(os.path.abspath(__file__))
    project_root = os.path.dirname(script_dir) # e:\health-tech\papers\vsa\
    papers_dir = os.path.join(project_root, "papers")
    assets_dir = os.path.join(project_root, "data", "paper_assets")
    tex_filename = "The Atomic VSA.tex"
    tex_path = os.path.join(papers_dir, tex_filename)

    # Required assets
    assets = []
    
    # scan .tex file for assets
    print(f"Scanning {tex_filename} for assets...")
    try:
        with open(tex_path, 'r', encoding='utf-8') as f:
            content = f.read()
            import re
            # Look for \IfFileExists{filename} OR \includegraphics{filename}
            # Matches: fig1.png, image.jpg, etc.
            # We explicitly look for the file usage in the template
            matches = re.findall(r'\\IfFileExists\{([^}]+)\}', content)
            matches += re.findall(r'\\includegraphics(?:\[.*?\])?\{([^}]+)\}', content)
            
            # Deduplicate and clean
            assets = sorted(list(set(matches)))
            # Filter out LaTeX macro arguments (start with #)
            assets = [a for a in assets if not a.startswith('#')]
            
            print(f"Found {len(assets)} required assets: {assets}")
            
    except Exception as e:
        print(f"Error reading .tex file: {e}")
        return

    print(f"Building PDF for {tex_filename}...")
    
    # 1. Copy assets to papers directory (where latex expects them)
    print("Copying assets...")
    # Source directories to search in
    search_dirs = [assets_dir, project_root]
    
    for asset in assets:
        # Skip if asset is just a base filename without extension (latex sometimes omits .png)
        # But our current usage includes extensions.
        if not asset: continue
        
        found = False
        for src_dir in search_dirs:
            src = os.path.join(src_dir, asset)
            if os.path.exists(src):
                dst = os.path.join(papers_dir, asset)
                shutil.copy2(src, dst)
                print(f"  Copied {asset} from {src_dir}")
                found = True
                break
        
        if not found:
             # Try recursive search in data/paper_assets if not immediately found
            for root, dirs, files in os.walk(assets_dir):
                if asset in files:
                    src = os.path.join(root, asset)
                    dst = os.path.join(papers_dir, asset)
                    shutil.copy2(src, dst)
                    print(f"  Copied {asset} from {root} (recursive search)")
                    found = True
                    break
        
        if not found:
            print(f"  WARNING: Asset not found in search paths: {asset}")

    # 2. Check for pdflatex
    if shutil.which("pdflatex") is None:
        print("\nERROR: 'pdflatex' executable not found.")
        print("  1. 'pip install pdflatex' is NOT sufficient (it is just a wrapper).")
        print("  2. You must install a LaTeX distribution.")
        print("  -> WINDOWS: Run 'winget install MiKTeX' or download from miktex.org")
        print("  -> LINUX: 'sudo apt install texlive-full'")
        print("  -> MAC: 'brew install mactex'")
        return

    # 3. Run pdflatex
    print("Running pdflatex...")
    
    # Needs to be run twice for cross-references/labels to resolve correctly
    cmd = ["pdflatex", "-interaction=nonstopmode", tex_filename]
    
    try:
        # Run 1
        subprocess.run(cmd, cwd=papers_dir, check=True)
        print("  Pass 1 complete.")
        
        # Run 2
        subprocess.run(cmd, cwd=papers_dir, check=True)
        print("  Pass 2 complete.")
        
        pdf_path = os.path.join(papers_dir, "The Atomic VSA.pdf")
        if os.path.exists(pdf_path):
            print(f"PDF Build Successful!")
            print(f"Output: {pdf_path}")
        else:
            print("ERROR: PDF file not found after build.")
    
    except subprocess.CalledProcessError as e:
        print(f"Error during pdflatex execution: {e}")
        # Could print log file here if needed
    except Exception as e:
        print(f"Unexpected error: {e}")

if __name__ == "__main__":
    build_pdf()