Benchmark: TextDecoder - MeasureThat.net

Script Preparation code:

var ToStringFromUTF16_1 = (function() {
            var UTF16 = new TextDecoder("utf-16");
            return function(bytes, startIndex, length) {
                startIndex = startIndex | 0;
                if (typeof length !== "undefined") length += startIndex;
                return UTF16.decode(bytes.subarray(startIndex, length));
            };
    })();

var ToStringFromUTF16_2 = (function() {
            var fromCharCode = String.fromCharCode;
            return function (bytes, startIndex, length) {
                startIndex = startIndex | 0;
                if (typeof length === "undefined")
                    length = bytes.length - startIndex;
                var s = "",
                    endIndex = startIndex + length,
                    i = startIndex;
                while (i < endIndex) {
                    var octet = bytes[i],
                        bytesNeeded = 0,
                        codePoint = 0;
                    if (octet <= 0x7F) {
                        bytesNeeded = 0;
                        codePoint = octet & 0xFF;
                    } else if (octet <= 0xDF) {
                        bytesNeeded = 1;
                        codePoint = octet & 0x1F;
                    } else if (octet <= 0xEF) {
                        bytesNeeded = 2;
                        codePoint = octet & 0x0F;
                    } else if (octet <= 0xF4) {
                        bytesNeeded = 3;
                        codePoint = octet & 0x07;
                    }
                    if (endIndex - i - bytesNeeded > 0) {
                        for (var j = 0; j < bytesNeeded; j++) {
                          octet = bytes[i + j + 1];
                          codePoint = (codePoint << 6) | (octet & 0x3F);
                        }
                    } else {
                        codePoint = 0xFFFD; // REPLACEMENT CHARACTER
                        bytesNeeded = endIndex - i;
                    }
                    //s += String.fromCodePoint(codePoint);                   
                    //if (!(codePoint < 0x10FFFF && (codePoint>>>0) === codePoint))
                    //    throw RangeError("Invalid code point: " + codePoint);
                    if (codePoint <= 0xFFFF) { // BMP code point
                        s += fromCharCode(codePoint);
                    } else {
                        codePoint -= 0x10000;
                        s += fromCharCode((codePoint >> 10) + 0xD800,    // high surrogate
                                          (codePoint % 0x400) + 0xDC00); // low surrogate
                    }
                    i += bytesNeeded + 1;
                }
                return s;
            };
    })();

var bytes = new Uint8Array([32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0])

Tests:

TextDecoder
ToStringFromUTF16_1(bytes);
Concat
ToStringFromUTF16_2(bytes);

Rendered benchmark preparation results:

Suite status: <idle, ready to run>

Previous results

Test case name	Result
TextDecoder
Concat

Fastest: N/A

Slowest: N/A

Latest run results:

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

User agent: Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/138.0.0.0 Mobile Safari/537.36

Browser/OS: Chrome Mobile 138 on Android

View result in a separate tab

Concat

5.8M/s

TextDecoder

3.1M/s

View exact numbers

Test name	Executions per second
✓ Concat	5,828,570 Ops/sec
TextDecoder	3,136,647 Ops/sec

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

LLMs can make mistakes. Check important info.

**Benchmark Overview**

The provided JSON represents a JavaScript microbenchmark for the `TextDecoder` API, which is used to decode bytes from different encoding schemes into a string.

**Benchmark Definition**

The benchmark definition consists of two scripts: `ToStringFromUTF16_1` and `ToStringFromUTF16_2`. These scripts test two different approaches to decoding UTF-16 encoded strings:

1. **`ToStringFromUTF16_1`**: This script uses the `TextDecoder` API with the `utf-16` encoding scheme. It takes a byte array, a start index, and a length as input and returns the decoded string.
2. **`ToStringFromUTF16_2`**: This script uses a manual decoding approach to convert UTF-16 encoded bytes into a string. It also takes a byte array, a start index, and a length as input.

**Options Compared**

The benchmark compares two options:

1. Using the `TextDecoder` API (`StringFromUTF16_1`)
2. Manual decoding using the `fromCharCode` method (`StringFromUTF16_2`)

**Pros and Cons of Each Approach**

**Using the `TextDecoder` API (`StringFromUTF16_1`)**

Pros:

* Faster execution times due to optimized native implementation
* More accurate results as it handles all encoding schemes and edge cases correctly
* Easier to maintain and update, as changes are handled by the browser's engine

Cons:

* May have dependencies on specific browser versions or platforms
* Can be slower in older browsers or with limited resources

**Manual Decoding (`StringFromUTF16_2`)**

Pros:

* No dependencies on specific browser versions or platforms
* Can be faster for certain input data due to optimized algorithmic performance

Cons:

* More prone to errors and edge cases, as it relies on manual implementation
* May require more maintenance and updates to ensure accuracy and compatibility

**Other Considerations**

1. **Browser Support**: The `TextDecoder` API is supported by most modern browsers, but older versions or platforms may not support it.
2. **Encoding Schemes**: The benchmark tests decoding UTF-16 encoded strings, but other encoding schemes (e.g., UTF-8) may require different approaches.
3. **Optimization Techniques**: Browser engines and JavaScript interpreters often apply optimization techniques, such as inlining, caching, or code reordering, which can affect the benchmark results.

**Alternatives**

Other alternatives for decoding UTF-16 encoded strings include:

1. Using the `Intl` API's `TextDecoder` interface
2. Utilizing third-party libraries or tools for Unicode normalization and encoding conversion

These alternatives may offer different trade-offs in terms of performance, accuracy, and maintainability compared to using the `TextDecoder` API or manual decoding approaches.

**Special JS Feature**

The benchmark uses a special JavaScript feature: **bitwise operations** (e.g., `startIndex | 0;`, `bytesNeeded = endIndex - i;`). These operations are used to perform bitwise arithmetic on integers, which can be beneficial for performance-critical code paths.

TextDecoder (version: 0)

Comparing performance of: `TextDecoder vs Concat`

Created: 5 years ago by: Guest

Go to the latest result

TextDecoder

Concat

Suite status: <idle, ready to run>

Fastest: N/A

Slowest: N/A

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

TextDecoder (version: 0)

Comparing performance of: TextDecoder vs Concat

Created: 5 years ago by: Guest

Go to the latest result

TextDecoder

Concat

Suite status: <idle, ready to run>

Fastest: N/A

Slowest: N/A

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

Related benchmarks

Comparing performance of: `TextDecoder vs Concat`