assemblyscript-regex - npm Package Compare versions

Comparing version 1.4.0 to 1.4.1

asconfig.json

@@ -12,7 +12,7 @@ {
       "textFile": "build/optimized.wat",
-      "sourceMap": true,
+      "sourceMap": false,
       "optimizeLevel": 3,
-      "shrinkLevel": 1,
-      "converge": false,
-      "noAssert": false
+      "shrinkLevel": 0,
+      "converge": true,
+      "noAssert": true
     },
@@ -19,0 +19,0 @@ "test": {

assembly/__tests__/character-sets.spec.ts

@@ -46,1 +46,10 @@ import { expectMatch, expectNotMatch } from "./utils";
 });
+
+it("supports case insensitive matching", () => {
+  // simple ranges
+  expectMatch("[a-c]", ["A", "C", "a", "c"], "i");
+  expectNotMatch("[a-c]", ["D", "d"], "i");
+  // complex
+  expectMatch("[W-c]", ["W", "w", "C", "c"], "i");
+  expectNotMatch("[W-c]", ["V", "v", "D", "d"], "i");
+});

assembly/__tests__/utils.ts

 import { RegExp, Match } from "..";
 
-export function expectMatch(regex: string, arr: string[]): void {
-  let regexp = new RegExp(regex);
+export function expectMatch(
+  regex: string,
+  arr: string[],
+  flags: string = ""
+): void {
+  let regexp = new RegExp(regex, flags);
   for (let i = 0; i < arr.length; i++) {
@@ -12,4 +16,8 @@ const value = arr[i];
 
-export function expectNotMatch(regex: string, arr: string[]): void {
-  let regexp = new RegExp(regex);
+export function expectNotMatch(
+  regex: string,
+  arr: string[],
+  flags: string = ""
+): void {
+  let regexp = new RegExp(regex, flags);
   for (let i = 0; i < arr.length; i++) {
@@ -16,0 +24,0 @@ const match = regexp.exec(arr[i]);

assembly/nfa/matcher.ts

@@ -11,2 +11,3 @@ import { isDigit, isAlpha, isWhitespace, Char } from "../char";
 import { Flags } from "../regexp";
+import { Range } from "../util";
 
@@ -40,3 +41,6 @@ const enum MatcherType {
   ): CharacterRangeMatcher {
-    return new CharacterRangeMatcher(node.from, node.to, flags.ignoreCase);
+    return new CharacterRangeMatcher(
+      new Range(node.from, node.to),
+      flags.ignoreCase
+    );
   }
@@ -94,8 +98,22 @@
 
+const LOWERCASE_LETTERS = new Range(Char.a, Char.z);
+const UPPERCASE_LETTERS = new Range(Char.A, Char.Z);
+const UPPER_LOWER_OFFSET = Char.a - Char.A;
+
 export class CharacterRangeMatcher extends Matcher {
-  constructor(private from: u32, private to: u32, private ignoreCase: bool) {
+  private ranges: Range[];
+
+  constructor(private range: Range, ignoreCase: bool) {
     super(MatcherType.CharacterRange);
+    this.ranges = [range];
+
     if (ignoreCase) {
-      this.from |= 0x20;
-      this.to |= 0x20;
+      const lowerIntersect = range.intersection(LOWERCASE_LETTERS);
+      if (lowerIntersect) {
+        this.ranges.push(lowerIntersect.offset(-UPPER_LOWER_OFFSET));
+      }
+      const upperIntersect = range.intersection(UPPERCASE_LETTERS);
+      if (upperIntersect) {
+        this.ranges.push(upperIntersect.offset(UPPER_LOWER_OFFSET));
+      }
     }
@@ -105,6 +123,8 @@ }
   matches(code: u32): bool {
-    if (this.ignoreCase) {
-      code |= 0x20;
+    for (let i = 0, len = this.ranges.length; i < len; i++) {
+      if (code >= u32(this.ranges[i].from) && code <= u32(this.ranges[i].to)) {
+        return true;
+      }
     }
-    return code >= this.from && code <= this.to;
+    return false;
   }
@@ -111,0 +131,0 @@ }

assembly/util.ts

@@ -14,1 +14,15 @@ export function last<T>(arr: T[]): T {
 }
+
+export class Range {
+  constructor(public from: i32, public to: i32) {}
+
+  intersection(other: Range): Range | null {
+    const lower = i32(Math.max(this.from, other.from));
+    const upper = i32(Math.min(this.to, other.to));
+    return lower < upper ? new Range(lower, upper) : null;
+  }
+
+  offset(value: i32): Range {
+    return new Range(this.from + value, this.to + value);
+  }
+}

benchmark/benchmark.js

@@ -8,13 +8,10 @@ global.TextDecoder = require("text-encoding").TextDecoder;
 
-wasmModule = loader.instantiateSync(
-  fs.readFileSync("./build/debug/assemblyscript-regex.wasm"),
-  {
-    env: {
-      log: () => {
-        const { __getString } = wasmModule.exports;
-        console.log(__getString(strPtr));
-      },
+wasmModule = loader.instantiateSync(fs.readFileSync("./build/optimized.wasm"), {
+  env: {
+    log: () => {
+      const { __getString } = wasmModule.exports;
+      console.log(__getString(strPtr));
     },
-  }
-);
+  },
+});
 
@@ -21,0 +18,0 @@ // the executeRegExp exported function is ex

package.json

 {
   "name": "assemblyscript-regex",
-  "version": "1.4.0",
+  "version": "1.4.1",
   "description": "A regex engine built with AssemblyScript",
@@ -14,4 +14,4 @@ "ascMain": "assembly/index.ts",
     "eslint:write": "npm run eslint -- --fix ",
-    "asbuild:untouched": "asb --target debug",
-    "asbuild:optimized": "asb",
+    "asbuild:untouched": "asc assembly/index.ts --target debug",
+    "asbuild:optimized": "asc assembly/index.ts --target release",
     "asbuild": "npm run asbuild:untouched && npm run asbuild:optimized",
@@ -30,3 +30,2 @@ "tsrun": "ts-node ts/index.ts",
     "@typescript-eslint/parser": "^4.14.1",
-    "asbuild": "0.0.10",
     "assemblyscript": "^0.18.0",
@@ -33,0 +32,0 @@ "benchmark": "^2.1.4",

spec/test-generator.js

@@ -24,4 +24,7 @@ const fs = require("fs");
   "test contains an octal escape sequence": [1102],
+  // the test results measure captured groups using character length / locations
+  // see: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/String/length
+  // this is tricky to reproduce
+  "test requires a substring function": [1087],
   "requires triage": [
-    1087,
     1363,
@@ -36,11 +39,6 @@ 1369,
   ],
-  "as-pect test issue": [1145, 1146],
   "test indicates a malformed regex, whereas it appears OK in JS": [1189],
   "test regex contains syntax not supported in JS": [82, 1158, 281],
   "the test behaviour differs between PCRE and JS": [290],
-  "aspect [Actual]: <Match>null vs [Expected]: Not <Match>null issue": [
-    153,
-    203,
-    204,
-  ],
+  "test appears to be incorrect?": [203, 204],
 };
@@ -47,0 +45,0 @@

ts/index.ts

@@ -8,8 +8,4 @@ import "assemblyscript/std/portable/index";
 
-const regexObj = new RegExp("^(a){1,3}");
-const match = regexObj.exec("abc");
+const regexObj = new RegExp("[a-c]", "i");
+const match = regexObj.exec("A");
 console.log(JSON.stringify(match, null, 2));
-
-const regexObj2 = new RegExp("(a|b)c|a(b|c)");
-const match2 = regexObj2.exec("ab");
-console.log(JSON.stringify(match2, null, 2));

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

254397

increased by0.68%

Dev dependency count

14

decreased by-6.67%

Lines of code

4506

increased by1.01%

No dependency changes