Benchmark: text encoders - MeasureThat.net

Tests:

str2ab
str2ab('Hello world!'); function str2ab(str) { var buf = new ArrayBuffer(str.length*2); var bufView = new Uint16Array(buf); for (var i=0, strLen=str.length; i < strLen; i++) { bufView[i] = str.charCodeAt(i); } return buf; }
textencoder api
str2ab('Hello world!'); function str2ab(str) { return new TextEncoder('utf-8').encode(str) }

Rendered benchmark preparation results:

Suite status: <idle, ready to run>

Previous results

Test case name	Result
str2ab
textencoder api

Fastest: N/A

Slowest: N/A

Latest run results:

Run details: (Test run date: one month ago)

User agent: Mozilla/5.0 (Android 15; Mobile; rv:148.0) Gecko/148.0 Firefox/148.0

Browser/OS: Firefox Mobile 148 on Android

View result in a separate tab

Test name	Executions per second
str2ab	1336300.1 Ops/sec
textencoder api	2783721.0 Ops/sec

Autogenerated LLM Summary (model llama3.2:3b, generated one year ago):

LLMs can make mistakes. Check important info.

Let's dive into the world of JavaScript microbenchmarks!

**Benchmark Definition**

The benchmark is titled "text encoders" and has three empty fields for Script Preparation Code, Html Preparation Code, and Description.

**Individual Test Cases**

There are two test cases:

1. **str2ab**: This test case measures the performance of a custom function `str2ab` that converts a string to an array buffer using UTF-16 encoding.
	* The script preparation code is an empty string, which means the JavaScript engine should generate it automatically when running the benchmark.
	* In this case, the script preparation code generates an ArrayBuffer and a Uint16Array view to perform the conversion.
2. **textencoder api**: This test case measures the performance of using the Web API `TextEncoder` to convert a string to an array buffer with UTF-8 encoding.
	* The script preparation code is also empty, as it relies on the engine's generation.
	* In this case, the engine should create a new TextEncoder instance and call its encode method.

**Comparison of Approaches**

The two test cases compare the performance of:

1. Custom implementation (`str2ab`) vs. Web API (`TextEncoder api`):
	* Pros of custom implementation:
		+ Can be optimized for specific use cases or platforms.
		+ May provide more control over the conversion process.
	* Cons of custom implementation:
		+ Requires manual memory management and handling of errors.
		+ May not be as efficient as a native Web API implementation.
	* Pros of Web API (`TextEncoder`):
		+ Provided by the browser engine, ensuring consistency across platforms.
		+ Generally faster and more efficient than custom implementations.
	* Cons of Web API:
		+ Less control over the conversion process.
		+ May not be available in older browsers or environments.

**Library Used**

In the `str2ab` test case, a library is implicitly used: the browser engine's internal functions for working with arrays and buffers. However, if we were to implement this function ourselves without using these built-in functions, it would require manual memory management and handling of errors.

**Special JS Feature/Syntax**

Neither test case uses any special JavaScript features or syntax. They only rely on standard JavaScript constructs and the Web API `TextEncoder`.

**Other Alternatives**

If you wanted to create similar benchmarks for other text encoding algorithms, you could consider using different libraries or implementations, such as:

* Unicode normalization
* UTF-32 or UTF-16 with surrogate pairs
* Other encoding schemes (e.g., Base64)

Keep in mind that each alternative would require a new benchmark definition and test cases.

By running these benchmarks, developers can compare the performance of custom implementations versus Web API solutions, helping to inform decisions about which approach to use in their projects.

Related benchmarks:

text encoders (version: 0)

Comparing performance of: str2ab vs textencoder api

Created: 9 years ago by: Guest

Jump to the latest result

str2ab

textencoder api

Suite status: <idle, ready to run>

Fastest: N/A

Slowest: N/A

Autogenerated LLM Summary (model llama3.2:3b, generated one year ago):