FAILURE MAP
← Case archive

FA-82458 / Chess move legality / Member archive

Pawn attacks on castling squares looked up backwards · case 03

A black pawn on h2 attacking g1 does not stop white from castling.

Member previewVariant 3 · 3 implementations · 8 checks per implementation

Case contract

Input: {"board", "side": "w" or "b", "wing": "K" or "Q", "rights": subset of "KQkq"}. Standard castling is legal when the right is held, the king is on e1/e8 and the matching own rook on h- or a-file, the squares between them are empty (f,g or b,c,d), and none of the king's squares e, f, g (kingside) or e, d, c (queenside) is attacked by the opponent.

Why this case matters

Move validators gate every move in engines, servers and arbiters; accepting an illegal move or rejecting a legal one corrupts the game record.

One recorded failure

Sample boundary fixture

This sample comes from the broken implementation of a controlled reproducer.

Boundary fixtureActualExpectedOutcome
pawn attacks g1 from h2truefalseFailed

MEMBER ARCHIVE

The complete case is available to members.

This record includes three runnable implementations, regression fixtures, execution results, and source hashes.

Member access is invitation-based. Sign in with your invited account to inspect the sources.

Sign in to the archive ↗