/**
 * Bytes as hex, in prose.
 *
 * Layer 0. Imports nothing.
 *
 * Two modules had a private `hexBytes` under the same name and they disagreed about the answer.
 * `header/fields.ts` wrote `0x4b 0x61 0x9f`; `header/variant.ts` wrote `44 6f 77`, in a sentence
 * that goes on to say "nor BDF's 0xFF followed by BIOSEMI" — both spellings of a byte, three
 * words apart, in one message. Without the prefix a short run reads as a decimal number until the
 * eye reaches a digit above 9, which on the `NOT_AN_EDF_FILE` diagnostic is the fifth byte.
 *
 * `0x` per byte, because that message is the argument: the author reached for it the moment a
 * single byte had to be named, and a list is the same claim repeated.
 *
 * This is NOT the renderer for the `rawBytes` block under a diagnostic. That one is a dump with an
 * ASCII column beside it — `30 20 20 20  |0   |` — where the columns are the point and a prefix on
 * every byte would push the ASCII off the line. It stays in `diagnostics/format.ts`, bare, and the
 * difference is that one is prose and the other is a hex dump.
 */
/** `'0x9f'`. */
export declare function hexByte(byte: number): string;
/**
 * `'0x4b 0x61 0x9f'`, from `from`, eliding with `...` on whichever side was cut.
 *
 * The ellipses are what stop a quoted run being read as the whole field: a caller pointing at the
 * first bad byte 40 bytes into an 80-byte identification field is showing a window, and a window
 * that does not say so is a claim about the field.
 */
export declare function hexBytes(bytes: Uint8Array, from?: number, max?: number): string;
//# sourceMappingURL=hex.d.ts.map