For which kind of attack does an attacker use known information in encrypted files to break the encryption scheme for the rest of

In a known-plaintext attack, the attacker possesses one or more samples of both original plaintext and the corresponding encrypted ciphertext. By analyzing these pairs, the attacker can reverse-engineer the encryption key or exploit algorithmic patterns to decrypt other ciphertext produced by the same key. This technique was famously used against the Enigma cipher during World War II.

A ciphertext-only attack means the attacker has only the encrypted output and no known plaintext samples, making it a strictly harder problem than a known-plaintext attack.

Unknown key is not a recognized cryptographic attack classification - the encryption key is unknown to the attacker by definition in virtually every attack scenario, so this term has no specific meaning.

A man-in-the-middle attack involves intercepting or relaying communications between two parties in transit, which is unrelated to using plaintext-ciphertext pairs to break a cipher.