#!/usr/bin/env python3 """Convert compliance guide markdown files to branded PDFs and upload to MinIO.""" import os import re import subprocess import sys sys.path.insert(0, "/app") GUIDES = [ "scripts/document_gen/templates/guides/dno_list_enforcement.md", "scripts/document_gen/templates/guides/kyc_procedures.md", "scripts/document_gen/templates/guides/material_change_procedures.md", "scripts/document_gen/templates/guides/traceback_response.md", ] FOOTER = ( '

' "Performance West Inc. · performancewest.net · 1-888-411-0383
" "525 Randall Ave Ste 100-1195, Cheyenne, WY 82001

" ) CSS = """ body { font-family: Arial, sans-serif; font-size: 11pt; line-height: 1.6; max-width: 700px; margin: 40px auto; color: #333; } h1 { color: #1a2744; border-bottom: 3px solid #059669; padding-bottom: 8px; } h2 { color: #1a2744; margin-top: 24px; } h3 { color: #1e3a5f; } table { border-collapse: collapse; width: 100%; margin: 12px 0; } th { background: #1a2744; color: white; padding: 8px 12px; text-align: left; font-size: 10pt; } td { padding: 8px 12px; border-bottom: 1px solid #e2e8f0; font-size: 10pt; } code { background: #f1f5f9; padding: 2px 6px; border-radius: 3px; font-size: 10pt; } pre { background: #f1f5f9; padding: 12px; border-radius: 6px; overflow-x: auto; font-size: 9pt; } ul, ol { padding-left: 20px; } li { margin-bottom: 4px; } strong { color: #111; } """ def md_to_html(md_text: str) -> str: html_parts = [f''] in_code = False in_table = False in_list = False for line in md_text.split("\n"): if line.startswith("```"): if in_code: html_parts.append("") in_code = False else: html_parts.append("

")
                in_code = True
            continue
        if in_code:
            html_parts.append(line)
            continue

        # Close table if we left it
        if in_table and not line.startswith("|"):
            html_parts.append("")
            in_table = False

        # Table separator
        if line.startswith("|") and "---" in line:
            continue

        if line.startswith("| "):
            cells = [c.strip() for c in line.split("|")[1:-1]]
            if not in_table:
                html_parts.append("" + "".join(f"" for c in cells) + "")
                in_table = True
            else:
                html_parts.append("" + "".join(f"" for c in cells) + "")
            continue

        # Headings
        if line.startswith("#### "):
            html_parts.append(f"{line[5:]}")
            continue
        if line.startswith("### "):
            html_parts.append(f"{line[4:]}")
            continue
        if line.startswith("## "):
            html_parts.append(f"{line[3:]}")
            continue
        if line.startswith("# "):
            html_parts.append(f"{line[2:]}")
            continue

        # Inline formatting
        line = re.sub(r"\*\*(.+?)\*\*", r"\1", line)
        line = re.sub(r"`(.+?)`", r"\1", line)
        line = re.sub(r"\[(.+?)\]\((.+?)\)", r'\1', line)

        if line.startswith("- "):
            if not in_list:
                html_parts.append("")
                in_list = True
            html_parts.append(f"{line[2:]}")
        elif re.match(r"^\d+\. ", line):
            if not in_list:
                html_parts.append("")
                in_list = True
            html_parts.append(f"{re.sub(r'^[0-9]+. ', '', line)}")
        else:
            if in_list:
                html_parts.append("" if "- " in html_parts[-5:] else "")
                in_list = False
            if line.strip() == "---":
                html_parts.append("")
            elif line.strip():
                html_parts.append(f"{line}")

    if in_table:
        html_parts.append("{c}
{c}")
    if in_list:
        html_parts.append("")

    html_parts.append(FOOTER)
    html_parts.append("")
    return "\n".join(html_parts)


def main():
    for md_path in GUIDES:
        if not os.path.exists(md_path):
            print(f"SKIP: {md_path} not found")
            continue

        with open(md_path) as f:
            md_text = f.read()

        title = md_text.split("\n")[0].replace("# ", "").strip()
        html = md_to_html(md_text)

        html_path = md_path.replace(".md", ".html")
        with open(html_path, "w") as f:
            f.write(html)

        # Convert to PDF
        pdf_dir = os.path.dirname(html_path)
        r = subprocess.run(
            ["libreoffice", "--headless", "--convert-to", "pdf", "--outdir", pdf_dir, html_path],
            capture_output=True, timeout=30,
        )
        pdf_path = html_path.replace(".html", ".pdf")

        if os.path.exists(pdf_path):
            try:
                from scripts.document_gen.minio_client import MinioStorage
                storage = MinioStorage()
                remote = f"guides/{os.path.basename(pdf_path)}"
                storage.upload(pdf_path, remote)
                print(f"OK: {title} -> {remote}")
            except Exception as e:
                print(f"OK (local only): {title} -> {pdf_path} (MinIO: {e})")
        else:
            print(f"FAIL: {title} — PDF not generated. stderr: {r.stderr[:200]}")


if __name__ == "__main__":
    main()