app/services/pdf_processor.py

from __future__ import annotations

import io
import re
from typing import BinaryIO

import pdfplumber

from app.models.schemas import PageContent

_HEADER_FOOTER_PATTERN = re.compile(
    r"^(?:page\s*\d+|halaman\s*\d+|\d+\s*(?:of|dari)\s*\d+).*$",
    re.IGNORECASE | re.MULTILINE,
)


# Complexity: Time O(p) | Space O(p) — p = page count
def extract_pages(file_bytes: bytes, filename: str) -> list[PageContent]:
    stream = io.BytesIO(file_bytes)
    return _extract_from_stream(stream)


def _extract_from_stream(stream: BinaryIO) -> list[PageContent]:
    pages: list[PageContent] = []
    with pdfplumber.open(stream) as pdf:
        for page_num, page in enumerate(pdf.pages, start=1):
            raw_text = page.extract_text() or ""
            cleaned = _clean_page_text(raw_text)
            if cleaned.strip():
                pages.append(PageContent(page_number=page_num, text=cleaned))
    return pages


def _clean_page_text(text: str) -> str:
    text = _HEADER_FOOTER_PATTERN.sub("", text)
    text = re.sub(r"\n{3,}", "\n\n", text)
    return text.strip()