Overview

Implement the following features with a focus on extreme performance and efficiency, as these operations will be invoked up to 600 times per second. All code must minimize memory allocations and avoid unnecessary overhead.

1. High-Performance Frame Processing

• Implement code to take a sessionResponseData and userBonesResponse (both as []byte), parse them into gameapi.SessionResponse{} and gameapi.UserBonesResponse{}.
• Build a new LobbySessionStateFrame using these parsed objects, determining all appropriate events.
• The implementation must be able to reference the previous frame to accurately detect and emit events.
• The parsing and event detection must be extremely efficient and optimized for high-frequency invocation.

2. Streaming Codecs

• Zstd Codec: Implement a streaming codec that writes the HEADER and one or more LobbySessionStateFrame objects to a Zstd-compressed .nevrcap file on disk.
• WebSocket Codec: Implement a streaming codec that connects to a WebSocket server and streams LobbySessionStateFrame objects. Add toggles to choose whether to send session data, user bones as raw bytes, or to unmarshal and detect events (generating and sending LobbySessionStateFrame).
• EchoReplay Codec: Implement a streaming codec that writes a .echoreplay file (zip format) to disk.

3. File Conversion Utilities

• Write a function to process a .echoreplay file into a .nevrcap file.
• Write a function to process a .nevrcap file into a .echoreplay file.

4. Acceptance and Benchmark Tests

• Provide acceptance tests for the .echoreplay file processors (external format compatibility).
• Provide a benchmark for uncompressed .echoreplay ➔ uncompressed .nevrcap processing performance.
• Provide a benchmark comparing .echoreplay vs .nevrcap file size.

5. BENCHMARKS.md and Automation

• Create a BENCHMARKS.md to capture the latest benchmark values.
• Implement a script to run all benchmarks and automatically update BENCHMARKS.md.

Acceptance Criteria:

• All codecs and utilities are implemented and tested for performance.
• Benchmarks and acceptance tests are present and automated.
• BENCHMARKS.md is created and kept up to date via script.
• Code is suitable for high-frequency, low-latency processing scenarios.