from __future__ import annotations

import re

_CONTROL_CHAR_PATTERN = re.compile(r"[\x00-\x08\x0b\x0c\x0e-\x1f\x7f]")
_MULTI_NEWLINE_PATTERN = re.compile(r"\n{3,}")


def sanitize_text(text: str) -> str:
    cleaned = _CONTROL_CHAR_PATTERN.sub("", text)
    cleaned = _MULTI_NEWLINE_PATTERN.sub("\n\n", cleaned)
    return cleaned.strip()


def chunk_text(text: str, limit: int = 1000) -> list[str]:
    if len(text) <= limit:
        return [text]

    chunks = []
    remaining = text
    while remaining:
        if len(remaining) <= limit:
            chunks.append(remaining)
            break

        split_at = remaining.rfind("\n", 0, limit)
        if split_at == -1 or split_at < limit // 2:
            split_at = remaining.rfind(" ", 0, limit)
        if split_at == -1:
            split_at = limit

        chunks.append(remaining[:split_at])
        remaining = remaining[split_at:].lstrip()
    return chunks