scripts/docgen.py

"""Shared Word/PDF renderers for AgroSense documents.

A document is (title, subtitle, sections). Each section is (heading, [blocks]);
a block is (kind, content) where kind is:
  "h2" -> subheading str ; "p" -> paragraph str ; "ul"/"ol" -> list[str].

Keep content ASCII so the PDF renders with built-in core fonts (no bundled TTF).
"""
from __future__ import annotations

from pathlib import Path
from typing import List, Tuple

Section = Tuple[str, list]


def _compose_docx(title: str, subtitle: str, sections: List[Section]):
    from docx import Document

    doc = Document()
    doc.add_heading(title, 0)
    if subtitle:
        doc.add_paragraph(subtitle)
    for heading, blocks in sections:
        doc.add_heading(heading, level=1)
        for kind, content in blocks:
            if kind == "h2":
                doc.add_heading(content, level=2)
            elif kind == "p":
                doc.add_paragraph(content)
            elif kind == "ul":
                for item in content:
                    doc.add_paragraph(item, style="List Bullet")
            elif kind == "ol":
                for item in content:
                    doc.add_paragraph(item, style="List Number")
    return doc


def _compose_pdf(title: str, subtitle: str, sections: List[Section]):
    from fpdf import FPDF
    from fpdf.enums import XPos, YPos

    pdf = FPDF()
    pdf.set_auto_page_break(True, margin=16)
    pdf.add_page()

    def cell(s, size=11, style="", lh=5.5, indent=0):
        pdf.set_font("Helvetica", style, size)
        if indent:
            pdf.set_x(pdf.l_margin + indent)
        pdf.multi_cell(0, lh, s, new_x=XPos.LMARGIN, new_y=YPos.NEXT)

    cell(title, size=18, style="B", lh=9)
    if subtitle:
        cell(subtitle, size=11, lh=6)
    pdf.ln(3)
    for heading, blocks in sections:
        pdf.ln(2)
        cell(heading, size=14, style="B", lh=7)
        for kind, content in blocks:
            if kind == "h2":
                pdf.ln(1)
                cell(content, size=12, style="B", lh=6)
            elif kind == "p":
                cell(content)
            elif kind in ("ul", "ol"):
                for i, item in enumerate(content, 1):
                    prefix = f"{i}. " if kind == "ol" else "- "
                    cell(prefix + item, indent=5)
        pdf.ln(1)
    return pdf


def build_docx(title: str, subtitle: str, sections: List[Section], path: Path) -> None:
    _compose_docx(title, subtitle, sections).save(str(path))


def build_pdf(title: str, subtitle: str, sections: List[Section], path: Path) -> None:
    _compose_pdf(title, subtitle, sections).output(str(path))


def build_docx_bytes(title: str, subtitle: str, sections: List[Section]) -> bytes:
    """Render a .docx to bytes (for streaming a generated document)."""
    import io
    buf = io.BytesIO()
    _compose_docx(title, subtitle, sections).save(buf)
    return buf.getvalue()


def build_pdf_bytes(title: str, subtitle: str, sections: List[Section]) -> bytes:
    """Render a .pdf to bytes (for streaming a generated document)."""
    return bytes(_compose_pdf(title, subtitle, sections).output())


def build_both(title: str, subtitle: str, sections: List[Section],
               out_dir: Path, basename: str) -> None:
    out_dir.mkdir(parents=True, exist_ok=True)
    docx_path = out_dir / f"{basename}.docx"
    pdf_path = out_dir / f"{basename}.pdf"
    build_docx(title, subtitle, sections, docx_path)
    build_pdf(title, subtitle, sections, pdf_path)
    print(f"Wrote {docx_path} ({docx_path.stat().st_size} b) and "
          f"{pdf_path} ({pdf_path.stat().st_size} b)")