Fixtures

Shared test data — hex, @file, data/encoding, format/selector

Fixtures define shared test data passed to every language implementation. They ensure all languages operate on identical input, so your benchmarks compare apples to apples.

Type Mapping

Each language receives fixture data in its native type. The table below shows the mapping:

Language	Fixture type
Go	`[]byte`
TypeScript	`Uint8Array`
Rust	`&[u8]`
Python	`bytes`
C	`uint8_t*` + length
C#	`byte[]`
Zig	`[]const u8`

Each language receives fixture data in its native type so benchmarks compare apples to apples.

Inline Hex

You can provide binary data as a hex-encoded string directly in the fixture. This works well for small, fixed test vectors.

Inline hex works well for small, fixed test vectors; use @file for larger payloads.

1fixture shortData {
2  hex: "68656c6c6f20776f726c64"   # "hello world" in UTF-8
3}
4
5fixture address {
6  hex: "d8da6bf26964af9d7eed9e03e53415d37aa96045"   # 20-byte Ethereum address
7}
8
9bench hashShort {
10  go: keccak256(shortData)
11  ts: keccak256(shortData)
12  rust: keccak256(&shortData)
13}

File-referenced Hex

For larger payloads, load hex from an external file using the @file directive. The path is relative to the .bench file.

1fixture largePayload {
2  hex: @file("fixtures/sort/sort_1000.hex")
3}
4
5fixture abiData {
6  hex: @file("fixtures/abi/erc20_calldata.hex")
7}
8
9bench sortLarge {
10  go: bubbleSort(largePayload)
11  ts: bubbleSort(largePayload)
12  rust: bubble_sort(&largePayload)
13}

Keep fixture files in a fixtures/ subdirectory. Organize by benchmark type (e.g. fixtures/sort/, fixtures/keccak/).

Generic data with encoding

When your data is not hex-encoded, use the data field with an encoding. The encoding controls how the bytes are decoded.

encoding	Use
`raw`	Raw binary bytes
`utf8`	UTF-8 string
`base64`	Base64-encoded string

Use encoding: raw for binary, utf8 for strings, base64 for base64-encoded data.

1fixture packetRaw {
2  data: @file("fixtures/net/packet.bin")
3  encoding: raw
4}
5
6fixture payloadUtf8 {
7  data: "hello world"
8  encoding: utf8
9}
10
11fixture payloadB64 {
12  data: "aGVsbG8gd29ybGQ="
13  encoding: base64
14}

Structured Extraction

When your data is in JSON or CSV, use format and selector to extract a specific value. The selector syntax depends on the format.

JSON uses JSONPath ($.path); CSV uses row,col (0-indexed).

1fixture requestId {
2  data: @file("fixtures/requests.json")
3  format: json
4  selector: "$.items[0].id"
5}
6
7fixture csvCell {
8  data: @file("fixtures/table.csv")
9  format: csv
10  selector: "1,2"    # row,col (0-indexed)
11}

Optional Fields

Fixtures support optional metadata fields:

Field	Purpose
`description`	Human-readable description
`shape`	JSON-like descriptor for metadata (e.g. `{ rows: 64, cols: 64 }`)

1fixture keccakInput {
2  description: "32-byte input for keccak256 hashing"
3  hex: "68656c6c6f20776f726c6468656c6c6f20776f726c6468656c6c6f20776f72"
4}
5
6fixture matrixData {
7  hex: @file("fixtures/matmul/mat_64.hex")
8  shape: { rows: 64, cols: 64, dtype: "float64" }
9}

Summary

The table below summarizes the fixture types and their syntax:

Fixture type	Syntax
Inline hex	`hex: "68656c6c6f"`
File hex	`hex: @file("path/to/file.hex")`
Raw / UTF-8 / Base64	`data: ...` + `encoding: raw \| utf8 \| base64`
JSON extraction	`data: @file(...)`, `format: json`, `selector: "$.path"`
CSV extraction	`data: @file(...)`, `format: csv`, `selector: "row,col"`

See the Fixtures Guide for practical patterns and scaling strategies.