src/summary.py

# src/summary.py
from __future__ import annotations

import re
from typing import Dict, List, Tuple

from schema import ExtractedDocument


def _top_keywords(keywords: Dict[str, int], top_n: int = 10) -> List[Tuple[str, int]]:
    items = sorted(keywords.items(), key=lambda x: x[1], reverse=True)
    return [(k, v) for k, v in items if v > 0][:top_n]


def _guess_address(text: str) -> str:
    # Very naive US-style address guess
    # e.g., "123 Main St, Dallas, TX 75201"
    pattern = r"\b\d{1,6}\s+[A-Za-z0-9.\- ]+\s+(Street|St|Avenue|Ave|Road|Rd|Boulevard|Blvd|Lane|Ln|Drive|Dr)\b.*?\b[A-Z]{2}\s+\d{5}\b"
    m = re.search(pattern, text, flags=re.IGNORECASE)
    return m.group(0).strip() if m else ""


def naive_summary(doc: ExtractedDocument) -> Dict[str, object]:
    words = doc.text.split()
    address_guess = _guess_address(doc.text[:20000])  # only scan early chunk

    return {
        "filename": doc.filename,
        "reader": doc.reader,
        "num_pages": doc.num_pages,
        "word_count": len(words),
        "table_count": len(doc.tables),
        "top_keyword_hits": _top_keywords(doc.keywords, top_n=12),
        "address_guess": address_guess,
    }