rohitg00/ai-engineering-from-scratch · error · ValueError

size must be positive

Error message

size must be positive

What it means

A generic argument guard at the top of the fixed-window chunker: fixed_window refuses to build chunks when the caller supplies a window size of zero or a negative number, because no positive-length slice of text could ever be produced. It fires before any slicing happens, so it always indicates a bad caller-supplied size (e.g. a config typo or an uninitialized variable), not a problem with the input text itself.

Solutions

1. Use a positive window size (the default is 400 characters)
2. Load window size from config with a bounds check before chunking
3. For token-based chunking, validate the converted token count is positive

When it happens

Trigger: Thrown at phases/19-capstone-projects/64-chunking-strategies-advanced/code/main.py:40 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of rohitg00/ai-engineering-from-scratch@39ea8a1c6d (2026-08-26). Data as JSON: /api/errors/68f4461ed739431c. Report an issue: GitHub.

Appendix: source

Thrown at phases/19-capstone-projects/64-chunking-strategies-advanced/code/main.py:40

@dataclass
class Chunk:
    doc_id: str
    strategy: str
    start: int
    end: int
    text: str

    def overlaps(self, gold_start: int, gold_end: int) -> bool:
        return not (self.end <= gold_start or self.start >= gold_end)


# ---------------------------------------------------------------------------
# strategy 1  --  fixed window
# ---------------------------------------------------------------------------

def fixed_window(doc_id: str, text: str, size: int = 400, overlap: int = 80) -> list[Chunk]:
    if size <= 0:
        raise ValueError("size must be positive")
    if overlap < 0 or overlap >= size:
        raise ValueError("overlap must be non-negative and smaller than size")
    out: list[Chunk] = []
    step = size - overlap
    i = 0
    n = len(text)
    while i < n:
        end = min(i + size, n)
        out.append(Chunk(doc_id, "fixed", i, end, text[i:end]))
        if end == n:
            break
        i += step
    return out


# ---------------------------------------------------------------------------
# strategy 2  --  sentence packer
# ---------------------------------------------------------------------------

View on GitHub (pinned to 39ea8a1c6d)