ahvn.klengine.scan_engine module

class ahvn.klengine.scan_engine.ScanKLEngine(storage, name=None, condition=None, *args, **kwargs)

Bases: BaseKLEngine

A brute-force scan KLEngine implementation with zero storage overhead.

This engine performs search by scanning through the entire attached KLStore and using eval_filter on each KL to find matches. It is always inplace and supports all kinds of KLStores.

This is the simplest possible search engine - no indexing, no optimization, just linear scan. Useful for small datasets or as a fallback.

Search Methods:

_search(topk, offset, include, **kwargs): Perform brute-force scan using eval_filter.

Abstract Methods (inherited from BaseKLEngine):

_upsert(kl): No-op (always inplace). _remove(key): No-op (always inplace). _clear(): No-op (always inplace).

Parameters:

• storage (BaseKLStore)

• name (str | None)

• condition (Callable | None)

inplace: bool = True

recoverable: bool = True

__init__(storage, name=None, condition=None, *args, **kwargs)

Initialize the ScanKLEngine.

Parameters:

• storage (BaseKLStore) – attach ScanKLEngine to a BaseKLStore (required).

• name (Optional[str]) – Name of the KLEngine instance. If None, defaults to “{storage.name}_scan_idx”.

• condition (Optional[Callable]) – Optional upsert/insert condition to apply to the KLEngine. KLs that do not satisfy the condition will be ignored. If None, all KLs are accepted.

• *args – Additional positional arguments passed to ScanKLEngine.

• **kwargs – Additional keyword arguments passed to ScanKLEngine.

__len__()

Return the number of KLs in the storage.

Returns:

The number of KLs in the storage.

Return type:

int

__iter__()

Iterate over all KLs in the storage.