simplified code branching in some areas

Author: justinwilaby

README.md

@@ -8,7 +8,7 @@
 The first streamable, low memory XML, HTML, JSX and Angular Template parser for [WebAssembly](https://developer.mozilla.org/en-US/docs/WebAssembly).
 
 Sax Wasm is a sax style parser for XML, HTML, JSX and Angular Templates written in [Rust](https://www.rust-lang.org/en-US/), compiled for
-WebAssembly with the sole motivation to bring **faster than native speeds** to XML and JSX parsing for node and the web.
+WebAssembly with the sole motivation to bring the **fastest possible speeds** to XML and JSX parsing for node and the web.
 Inspired by [sax js](https://github.com/isaacs/sax-js) and rebuilt with Rust for WebAssembly, sax-wasm brings optimizations
 for speed and support for JSX syntax.
 
@@ -20,12 +20,12 @@ All parsers are tested using a large XML document (3 MB) containing a variety of
 
 | Parser with Advanced Features                                                              | time/ms (lower is better)| JS     | Runs in browser |
 |--------------------------------------------------------------------------------------------|-------------------------:|:------:|:---------------:|
-| [sax-wasm](https://github.com/justinwilaby/sax-wasm)                                       |                    0.466 | ☑      | ☑               |
-| [saxes](https://github.com/lddubeau/saxes)                                                 |                    0.868 | ☑      | ☑               |
-| [ltx(using Saxes as the parser)](https://github.com/xmppjs/ltx)                            |                    0.881 | ☑      | ☑               |
-| [node-xml](https://github.com/dylang/node-xml)                                             |                    1.549 | ☑      | ☐               |
-| [node-expat](https://github.com/xmppo/node-expat)                                          |                    1.551 | ☑      | ☐               |
-| [sax-js](https://github.com/isaacs/sax-js)                                                 |                    1.869 | ☑      | ☑*              |
+| [sax-wasm](https://github.com/justinwilaby/sax-wasm)                                       |                    27.91 | ☑      | ☑               |
+| [saxes](https://github.com/lddubeau/saxes)                                                 |                    40.99 | ☑      | ☑               |
+| [ltx(using Saxes as the parser)](https://github.com/xmppjs/ltx)                            |                    41.87 | ☑      | ☑               |
+| [sax-js](https://github.com/isaacs/sax-js)                                                 |                   105.18 | ☑      | ☑*              |
+| [node-xml](https://github.com/dylang/node-xml)                                             |                   115.17 | ☑      | ☐               |
+| [node-expat](https://github.com/xmppo/node-expat)                                          |                  148.885 | ☑      | ☐               |
 <sub>*built for node but *should* run in the browser</sub>
 
 ## Installation

lib/sax-wasm.wasm

@@ -3,7 +3,7 @@ import { URL } from 'url';
 import { resolve } from 'path';
 import { Buffer } from 'buffer';
 
-import { SAXParser } from '../../../lib/cjs/index.js';
+import { SAXParser, SaxEventType } from '../../../lib/cjs/index.js';
 
 import nodeXml from 'node-xml';
 import expat from 'node-expat';
@@ -24,7 +24,7 @@ async function benchmarkSaxWasmParser() {
   let t = process.hrtime();
   let offset = 0;
   while (offset < xml.length) {
-    parser.write(Buffer.from(xml.slice(offset, chunkLen)));
+    parser.write(Buffer.from(xml.slice(offset, chunkLen + offset)));
     offset += chunkLen;
   }
   parser.end();
@@ -38,7 +38,7 @@ async function benchmarkNodeXmlParser() {
   let t = process.hrtime();
   let offset = 0;
   while (offset < xml.length) {
-    parser.parseString(Buffer.from(xml.slice(offset, chunkLen)));
+    parser.parseString(Buffer.from(xml.slice(offset, chunkLen + offset)));
     offset += chunkLen;
   }
   let [s, n] = process.hrtime(t);
@@ -50,7 +50,7 @@ async function benchmarkExpatParser() {
   let t = process.hrtime();
   let offset = 0;
   while (offset < xml.length) {
-    parser.parse(Buffer.from(xml.slice(offset, chunkLen)));
+    parser.parse(Buffer.from(xml.slice(offset, chunkLen + offset)));
     offset += chunkLen;
   }
 
@@ -64,7 +64,7 @@ async function benchmarkSaxesParser() {
 
   let offset = 0;
   while (offset < xml.length) {
-    parser.write(Buffer.from(xml.slice(offset, chunkLen)));
+    parser.write(Buffer.from(xml.slice(offset, chunkLen + offset)));
     offset += chunkLen;
   }
   let [s, n] = process.hrtime(t);
@@ -78,7 +78,7 @@ async function benchmarkSaxParser() {
 
   let offset = 0;
   while (offset < xml.length) {
-    parser.write(Buffer.from(xml.slice(offset, chunkLen)));
+    parser.write(Buffer.from(xml.slice(offset, chunkLen + offset)));
     offset += chunkLen;
   }
   let [s, n] = process.hrtime(t);
@@ -91,7 +91,7 @@ async function benchmarkLtxParser() {
 
   let offset = 0;
   while (offset < xml.length) {
-    parser.write(Buffer.from(xml.slice(offset, chunkLen)));
+    parser.write(Buffer.from(xml.slice(offset, chunkLen + offset)));
     offset += chunkLen;
   }
 
@@ -100,7 +100,7 @@ async function benchmarkLtxParser() {
 }
 
 async function benchmark() {
-  let t = 100;
+  let t = 10;
   let benchmarks = { saxWasm: [], nodeXml: [], saxes: [], sax: [], expat: [], ltx: [] };
   while (t--) {
     benchmarks.saxWasm.push(await benchmarkSaxWasmParser());

src/js/__test__/benchmark.mjs

@@ -110,16 +110,15 @@ impl GraphemeClusters<'_> {
     /// let result = gc.take_until_ascii(&[b' ']);
     /// assert!(result.is_some());
     /// ```
+
     pub fn take_until_ascii(&mut self, chars: &[u8]) -> Option<GraphemeResult<'_>> {
         let mut cursor = self.cursor;
         let start = self.cursor;
         let mut line = self.line;
         let mut character = self.character;
         let mut end = self.cursor;
-        loop {
-            if self.byte_len <= cursor {
-                return None;
-            }
+
+        while cursor < self.byte_len {
             let next_byte = unsafe { *self.bytes.get_unchecked(cursor) };
             if chars.contains(&next_byte) {
                 break;
@@ -135,6 +134,11 @@ impl GraphemeClusters<'_> {
             end += len;
             cursor = end;
         }
+
+        if self.cursor == cursor {
+            return None;
+        }
+
         self.cursor = cursor;
         self.line = line;
         self.character = character;