feat: implemented non-capturing groups

README.md

@@ -77,7 +77,7 @@ Based on the classfication within the [MDN cheatsheet](https://developer.mozilla
 - [x] (x) capturing group
 - [ ] \n back reference
 - [ ] (?<Name>x) named capturing group
-- [ ] (?:x) Non-capturing group
+- [x] (?:x) Non-capturing group
 
 **Quantifiers**
 

assembly/__spec_tests__/generated.spec.ts

@@ -1076,7 +1076,22 @@ it("line: 207 - matches ^(a(b(c)))(d(e(f)))(h(i(j)))(k(l(m)))$ against 'abcdefhi
   expect(match.matches[11]).toBe("abcdefhijklm".substring(10, 12));
   expect(match.matches[12]).toBe("abcdefhijklm".substring(11, 12));
 });
-xit("line: 208 - non capturing groups not supported", () => {});
+it("line: 208 - matches ^(?:a(b(c)))(?:d(e(f)))(?:h(i(j)))(?:k(l(m)))$ against 'abcdefhijklm'", () => {
+  const match = exec(
+    "^(?:a(b(c)))(?:d(e(f)))(?:h(i(j)))(?:k(l(m)))$",
+    "abcdefhijklm",
+    "ms"
+  );
+  expect(match.matches[0]).toBe("abcdefhijklm".substring(0, 12));
+  expect(match.matches[1]).toBe("abcdefhijklm".substring(1, 3));
+  expect(match.matches[2]).toBe("abcdefhijklm".substring(2, 3));
+  expect(match.matches[3]).toBe("abcdefhijklm".substring(4, 6));
+  expect(match.matches[4]).toBe("abcdefhijklm".substring(5, 6));
+  expect(match.matches[5]).toBe("abcdefhijklm".substring(7, 9));
+  expect(match.matches[6]).toBe("abcdefhijklm".substring(8, 9));
+  expect(match.matches[7]).toBe("abcdefhijklm".substring(10, 12));
+  expect(match.matches[8]).toBe("abcdefhijklm".substring(11, 12));
+});
 xit("line: 209 - back references are not supported", () => {});
 it("line: 210 - matches ^[.^$|()*+?{,}]+ against '.^$(*+)|{?,?}'", () => {
   const match = exec("^[.^$|()*+?{,}]+", ".^$(*+)|{?,?}", "ms");
@@ -1305,10 +1320,10 @@ it("line: 266 - matches ^12.34 against '12\r34'", () => {
 });
 xit("line: 267 - lookaheads not supported", () => {});
 xit("line: 268 - lookaheads not supported", () => {});
-xit("line: 269 - non capturing groups not supported", () => {});
-xit("line: 270 - non capturing groups not supported", () => {});
-xit("line: 271 - non capturing groups not supported", () => {});
-xit("line: 272 - non capturing groups not supported", () => {});
+xit("line: 269 - lookaheads not supported", () => {});
+xit("line: 270 - lookaheads not supported", () => {});
+xit("line: 271 - lookaheads not supported", () => {});
+xit("line: 272 - lookaheads not supported", () => {});
 xit("line: 273 - lookaheads not supported", () => {});
 xit("line: 274 - lookaheads not supported", () => {});
 xit("line: 281 - test regex contains syntax not supported in JS", () => {});
@@ -1564,8 +1579,14 @@ it("line: 1162 - matches \\Aabc\\Z against 'qqq\nabc\nzzz'", () => {
 });
 xit("line: 1163 - JS does not support the A Z syntax for start and end of string", () => {});
 xit("line: 1164 - JS does not support the A Z syntax for start and end of string", () => {});
-xit("line: 1165 - non capturing groups not supported", () => {});
-xit("line: 1166 - non capturing groups not supported", () => {});
+it("line: 1165 - matches (?:b)|(?::+) against 'b::c'", () => {
+  const match = exec("(?:b)|(?::+)", "b::c", "ms");
+  expect(match.matches[0]).toBe("b::c".substring(0, 1));
+});
+it("line: 1166 - matches (?:b)|(?::+) against 'c::b'", () => {
+  const match = exec("(?:b)|(?::+)", "c::b", "ms");
+  expect(match.matches[0]).toBe("c::b".substring(1, 3));
+});
 it("line: 1167 - matches [-az]+ against 'az-'", () => {
   const match = exec("[-az]+", "az-", "ms");
   expect(match.matches[0]).toBe("az-".substring(0, 3));
@@ -1954,9 +1975,21 @@ it("line: 1311 - matches \\d\\d\\/\\d\\d\\/\\d\\d\\d\\d against '01/01/2000'", (
   const match = exec("\\d\\d\\/\\d\\d\\/\\d\\d\\d\\d", "01/01/2000", "ms");
   expect(match.matches[0]).toBe("01/01/2000".substring(0, 10));
 });
-xit("line: 1312 - non capturing groups not supported", () => {});
-xit("line: 1313 - non capturing groups not supported", () => {});
-xit("line: 1314 - non capturing groups not supported", () => {});
+it("line: 1312 - matches word (?:[a-zA-Z0-9]+ ){0,10}otherword against 'word cat dog elephant mussel cow horse canary baboon snake shark otherword'", () => {
+  const match = exec(
+    "word (?:[a-zA-Z0-9]+ ){0,10}otherword",
+    "word cat dog elephant mussel cow horse canary baboon snake shark otherword",
+    "ms"
+  );
+  expect(match.matches[0]).toBe(
+    "word cat dog elephant mussel cow horse canary baboon snake shark otherword".substring(
+      0,
+      74
+    )
+  );
+});
+xit("line: 1313 - peformance issue", () => {});
+xit("line: 1314 - peformance issue", () => {});
 it("line: 1315 - matches ^(a){0,0} against 'bcd'", () => {
   const match = exec("^(a){0,0}", "bcd", "ms");
   expect(match.matches[0]).toBe("bcd".substring(0, 0));

assembly/__tests__/capture-group.spec.ts

@@ -45,3 +45,9 @@ it("range repitition capture groups should return the last match", () => {
   expect(match.matches[0]).toBe("ac");
   expect(match.matches[1]).toBe("c");
 });
+
+it("non-capturing groups should not capture", () => {
+  const match = exec("(?:foo)bar(baz)", "foobarbaz");
+  expect(match.matches[0]).toBe("foobarbaz");
+  expect(match.matches[1]).toBe("baz");
+});

assembly/char.ts

@@ -16,6 +16,7 @@ export const enum Char {
   Dot = 0x2e, // "."
   Zero = 0x30,
   Nine = 0x39,
+  Colon = 0x3a,
   Question = 0x3f, // "?"
   A = 0x41,
   D = 0x44,

assembly/nfa/nfa.ts

@@ -42,7 +42,7 @@ export class GroupStartMarkerState extends State {
   // captures from the path through the NFA that reaches the end are flagged
   flagged: bool = false;
 
-  constructor(next: State, public groupId: i32) {
+  constructor(next: State, public capturing: bool, public groupId: i32) {
     super();
     this.transitions.push(next);
   }
@@ -60,10 +60,12 @@ export class GroupEndMarkerState extends State {
   }
 
   matches(input: string, position: u32): MatchResult {
-    this.startMarker.capture = input.substring(
-      this.startMarker.location,
-      position
-    );
+    if (this.startMarker.capturing) {
+      this.startMarker.capture = input.substring(
+        this.startMarker.location,
+        position
+      );
+    }
     return MatchResult.Ignore;
   }
 }
@@ -164,10 +166,10 @@ function oneOrMore(nfa: Automata, greedy: bool): Automata {
   return new Automata(start, end);
 }
 
-function group(nfa: Automata, id: i32): Automata {
+function group(nfa: Automata, capturing: bool, id: i32): Automata {
   // groups are implemented by wrapping the automata with
   // a pair of markers that record matches
-  const startMarker = new GroupStartMarkerState(nfa.start, id);
+  const startMarker = new GroupStartMarkerState(nfa.start, capturing, id);
   const end = new State();
   const endMarker = new GroupEndMarkerState(end, startMarker);
   nfa.end.transitions.push(endMarker);
@@ -238,7 +240,11 @@ class AutomataFactor {
         );
       case NodeType.Group: {
         const node = expression as GroupNode;
-        return group(this.automataForNode(node.expression), node.id);
+        return group(
+          this.automataForNode(node.expression),
+          node.capturing,
+          node.id
+        );
       }
       case NodeType.Assertion:
         return Automata.fromEpsilon();

assembly/parser/node.ts

@@ -209,7 +209,11 @@ export class AlternationNode extends Node {
 let _id = 0;
 
 export class GroupNode extends Node {
-  constructor(public expression: Node, public id: i32 = -1) {
+  constructor(
+    public expression: Node,
+    public capturing: bool,
+    public id: i32 = -1
+  ) {
     super(NodeType.Group);
     if (id == -1) {
       this.id = _id++;
@@ -221,7 +225,7 @@ export class GroupNode extends Node {
   }
 
   clone(): Node {
-    return new GroupNode(this.expression.clone(), this.id);
+    return new GroupNode(this.expression.clone(), this.capturing, this.id);
   }
 
   replace(node: Node, replacement: Node): void {

assembly/parser/parser.ts

@@ -204,6 +204,18 @@ export class Parser {
     return true;
   }
 
+  private isCapturing(): bool {
+    if (
+      this.iterator.current == Char.Question &&
+      this.iterator.lookahead(1) == Char.Colon
+    ) {
+      this.eatToken(Char.Question);
+      this.eatToken(Char.Colon);
+      return false;
+    }
+    return true;
+  }
+
   // parses a sequence of chars
   private parseSequence(): Node {
     let nodes = new Array<Node>();
@@ -218,7 +230,8 @@ export class Parser {
         // @ts-ignore
       } else if (token == Char.LeftParenthesis) {
         this.eatToken(Char.LeftParenthesis);
-        nodes.push(new GroupNode(this.parseSequence()));
+        const capturing = this.isCapturing();
+        nodes.push(new GroupNode(this.parseSequence(), capturing));
         this.eatToken(Char.RightParenthesis);
         // @ts-ignore
       } else if (token == Char.LeftCurlyBrace) {

assembly/regexp.ts

@@ -89,9 +89,9 @@ export class Flags {
 
 // capture groups are implemented as GroupStart / GroupEnd states that record (capture)
 // the value of the current state of the string being matched.
-// Repeated capture groups, via rage repetitions (e.g. {2,3}) share the same 'id'. The
+// Repeated capture groups, via range repetitions (e.g. {2,3}) share the same 'id'. The
 // returned regex should only return the value of the final repetition.
-function filterCaptures(groupMarkers: GroupStartMarkerState[]): string[] {
+function lastCapturesForGroup(groupMarkers: GroupStartMarkerState[]): string[] {
   if (!groupMarkers.length) {
     return [];
   }
@@ -139,7 +139,10 @@ export class RegExp {
     gm = new Array<GroupStartMarkerState>();
     nfaWalker(this.nfa.start, (state) => {
       if (state instanceof GroupStartMarkerState) {
-        gm.push(state as GroupStartMarkerState);
+        const startMarker = state as GroupStartMarkerState;
+        if (startMarker.capturing) {
+          gm.push(state as GroupStartMarkerState);
+        }
       }
     });
     this.groupMarkers = gm;
@@ -181,7 +184,7 @@ export class RegExp {
         });
 
         const match = new Match(
-          [matchStr!].concat(filterCaptures(groupMarkers)),
+          [matchStr!].concat(lastCapturesForGroup(groupMarkers)),
           matchIndex,
           str
         );

spec/test-generator.js

@@ -22,6 +22,7 @@ const knownIssues = {
     ...range(141, 143),
     1288,
   ],
+  "peformance issue": [1313, 1314],
 
   /* -------- issues with the tests ------------  */
   "test appears to be incorrect?": [203, 204],
@@ -108,11 +109,6 @@ lines.forEach((line, index) => {
       return;
     }
 
-    if (["(?:"].some((f) => regex.includes(f))) {
-      testCase += `xit("line: ${index} - non capturing groups not supported", () => {});`;
-      return;
-    }
-
     if (["(?!", "(?="].some((f) => regex.includes(f))) {
       testCase += `xit("line: ${index} - lookaheads not supported", () => {});`;
       return;

ts/index.ts

@@ -5,13 +5,8 @@ globalAny.log = console.log;
 
 import { RegExp } from "../assembly/regexp";
 
-const regexObj = new RegExp("abc$", "m");
-let match = regexObj.exec("abc\n");
+const regexObj = new RegExp("word (?:[a-zA-Z0-9]+ ){0,300}otherword", "");
+let match = regexObj.exec(
+  "word cat dog elephant mussel cow horse canary baboon snake shark the quick brown fox and the lazy dog and several other words getting close to thirty by now I hope"
+);
 console.log(JSON.stringify(match, null, 2));
-// match = regexObj.exec("f1\nbar\nbaz\nf2");
-// console.log(JSON.stringify(match, null, 2));
-
-// const regex = new RegExp("^f\\d{1}$", "gm");
-
-//     let match = regex.exec("f1\nbar\nbaz\nf2");
-//     expect(match!.matches[0]).toBe("f1");