export_pdf.py

# export_pdf.py
import os
from dotenv import load_dotenv
from jinja2 import Environment, FileSystemLoader, select_autoescape
import markdown as md
import pdfkit
import json

load_dotenv()
WKHTMLTOPDF_PATH = os.getenv("WKHTMLTOPDF_PATH", "")

def _pdfkit_config():
    if WKHTMLTOPDF_PATH and os.path.exists(WKHTMLTOPDF_PATH):
        return pdfkit.configuration(wkhtmltopdf=WKHTMLTOPDF_PATH)
    # Nếu bạn đã set PATH hệ thống cho wkhtmltopdf, có thể trả về None
    return None

# Mode A: Markdown -> HTML -> PDF (dễ áp dụng ngay cho tailored_resume.md)
def export_markdown_to_pdf(md_path: str, pdf_path: str, title: str = "Curriculum Vitae"):
    if not os.path.exists(md_path):
        raise FileNotFoundError(f"Không tìm thấy file Markdown: {md_path}")

    # Đọc markdown & convert sang HTML
    with open(md_path, "r", encoding="utf-8") as f:
        md_text = f.read()
    html_content = md.markdown(md_text, extensions=["extra", "sane_lists", "tables"])

    # Load template wrapper
    env = Environment(
        loader=FileSystemLoader("templates"),
        autoescape=select_autoescape(["html", "xml"])
    )
    template = env.get_template("cv_wrapper.html")
    final_html = template.render(title=title, content=html_content)

    # Đảm bảo thư mục output tồn tại
    os.makedirs(os.path.dirname(pdf_path), exist_ok=True)

    # Xuất PDF
    config = _pdfkit_config()
    pdfkit.from_string(final_html, pdf_path, configuration=config)
    print(f"✅ Đã xuất PDF: {pdf_path}")

# Mode B: JSON -> Template (cv_basic.html) -> PDF (nâng cao)
def export_json_to_pdf(json_path: str, pdf_path: str, template_name: str = "cv_basic.html"):
    if not os.path.exists(json_path):
        raise FileNotFoundError(f"Không tìm thấy file JSON: {json_path}")

    with open(json_path, "r", encoding="utf-8") as f:
        data = json.load(f)

    env = Environment(
        loader=FileSystemLoader("templates"),
        autoescape=select_autoescape(["html", "xml"])
    )
    template = env.get_template(template_name)
    final_html = template.render(data=data)

    os.makedirs(os.path.dirname(pdf_path), exist_ok=True)

    config = _pdfkit_config()
    pdfkit.from_string(final_html, pdf_path, configuration=config)
    print(f"✅ Đã xuất PDF: {pdf_path}")

if __name__ == "__main__":
    # Ví dụ chạy nhanh (Mode A) với file outputs/tailored_resume.md:
    export_markdown_to_pdf(
        md_path="outputs/tailored_resume.md",
        pdf_path="outputs/CV_ATS.pdf",
        title="CV - ATS"
    )