@ptolemy2002/rgx 6.2.0 → 7.1.0

package/README.md

@@ -86,6 +86,7 @@ type RGXCapture<T = unknown> = {
     value: T;
     start: number;
     end: number;
+    ownerId: string | null;
 };
 
 type RGXPartOptions<R, T=string> = {
@@ -663,8 +664,8 @@ constructor(source: string, tokens: RGXTokenCollectionInput, options?: RGXWalker
 - `tokens` (`RGXTokenCollection`): The internal collection of tokens in 'concat' mode (readonly).
 - `tokenPosition` (`number`): The current index in the token collection. Setting this validates that the value is >= 0 and <= `tokens.length`, throwing `RGXOutOfBoundsError` if not.
 - `reduced` (`R`): A user-defined accumulator value, typically updated by `RGXPart` callbacks during walking.
-- `captures` (`RGXCapture[]`): An array of structured capture results recorded during walking. Each entry has a `raw` string and a `value` (the transform result for Parts, or the raw string for plain tokens).
-- `namedCaptures` (`Record<string, RGXCapture>`): An object mapping capture IDs to their corresponding `RGXCapture` results. Only Parts with non-null IDs are included.
+- `captures` (`RGXCapture[]`): An array of structured capture results recorded during walking. Each entry has a `raw` string, a `value` (the transform result for Parts, or the raw string for plain tokens), `start` and `end` indices in the source string, and an `ownerId` that is the `id` of the Part that produced it (or `null` for captures from plain tokens or parts without ids).
+- `namedCaptures` (`Record<string, RGXCapture[]>`): An object mapping capture IDs to their corresponding `RGXCapture` results. Only Parts with non-null IDs are included. The captures occur in the same order as they appear in the `captures` array.
 - `stopped` (`boolean`, readonly): Whether the walker has been stopped, either by a Part's `beforeCapture` returning `"stop"` or by calling `stop()` in an `afterCapture` callback.
 
 #### Methods
@@ -1353,6 +1354,8 @@ A pre-built `RegExpFlagTransformer` that makes a regex pattern accent-insensitiv
 - `o` / `O`: ó, ò, ö, ô, õ / Ó, Ò, Ö, Ô, Õ
 - `u` / `U`: ú, ù, ü, û / Ú, Ù, Ü, Û
 
+Note that this transformer intentionally excludes replacing characters preceded by an odd number of backslashes, to allow for escaping. For example, in the pattern `\\a`, the `a` is preceded by two backslashes (an even number), so it will be replaced with `(a|á|à|ä|â|ã)`. In the pattern `\a`, the `a` is preceded by one backslash (an odd number), so it will not be replaced.
+
 #### Parameters
   - `exp` (`RegExp`): The regular expression to transform.
 
@@ -1659,6 +1662,11 @@ Since these are defined as native tokens (strings), they are automatically wrapp
 | `"non-word-char"` | `\W` | Any non-word character |
 | `"backspace"` | `[\b]` | Backspace character |
 
+### Complex Constructs
+| Name | Resolves To | Description |
+| --- | --- | --- |
+| `"non-escape-bound"` | `(?<=(?<!\\)(?:\\\\)*)(?=[^\\]\|$)` | Matches a position that is not preceded by an odd number of backslashes, i.e., the next character is not escaped. Note that this doesn't match when the next character is a backslash, since allowing it to do that would cause non-escaped backslashes within a series of backslashes to be treated as escaped. For example, in the string `\\\a`, the first and third backslashes would be treated as escaped. |
+
 ## Peer Dependencies
 - `@ptolemy2002/immutability-utils` ^2.0.0
 - `@ptolemy2002/js-utils` ^3.2.2

package/dist/constants.js

@@ -191,3 +191,12 @@ defineRGXConstant("backspace", {
         return /[\b]/;
     }
 });
+// Complex Constructs
+// Put this before any pattern to ensure that the pattern is not escaped, i.e., not preceded by an odd number of backslashes.
+defineRGXConstant("non-escape-bound", {
+    rgxGroupWrap: false,
+    rgxIsRepeatable: false,
+    toRgx() {
+        return /(?<=(?<!\\)(?:\\\\)*)(?=[^\\]|$)/;
+    }
+});

package/dist/flag-transformer/accentInsensitive.js

@@ -1,6 +1,8 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.accentInsensitiveFlagTransformer = void 0;
+const constants_1 = require("../constants");
+const resolve_1 = require("../resolve");
 const accentPatterns = [
     "(a|á|à|ä|â|ã)", "(A|Á|À|Ä|Â|Ã)",
     "(e|é|è|ë|ê)", "(E|É|È|Ë|Ê)",
@@ -11,9 +13,10 @@ const accentPatterns = [
 const accentInsensitiveFlagTransformer = function (exp) {
     let source = exp.source;
     const flags = exp.flags;
+    const nonEscapeBound = (0, resolve_1.resolveRGXToken)((0, constants_1.rgxConstant)("non-escape-bound"));
     accentPatterns.forEach((pattern) => {
         // Replace any of the characters in the pattern with the pattern itself
-        source = source.replaceAll(new RegExp(pattern, "g"), pattern);
+        source = source.replaceAll(new RegExp(nonEscapeBound + pattern, "g"), pattern);
     });
     return new RegExp(source, flags);
 };

package/dist/walker/base.d.ts

@@ -13,7 +13,7 @@ export declare class RGXWalker<R> implements RGXConvertibleToken {
     _tokenPosition: number;
     reduced: R;
     captures: RGXCapture[];
-    namedCaptures: Record<string, RGXCapture>;
+    namedCaptures: Record<string, RGXCapture[]>;
     private _stopped;
     static check: (value: unknown) => value is RGXWalker<unknown>;
     static assert: (value: unknown) => asserts value is RGXWalker<unknown>;

package/dist/walker/base.js

@@ -98,7 +98,7 @@ class RGXWalker {
         const raw = this.capture(token);
         const end = this.sourcePosition;
         const value = isPart ? token.transform(raw) : raw;
-        const captureResult = { raw, value, start, end };
+        const captureResult = { raw, value, start, end, ownerId: isPart && token.hasId() ? token.id : null };
         // Validate the part. If validation fails, it will throw an error, so nothing below will run.
         if (isPart) {
             token.validate(captureResult, this);
@@ -107,7 +107,9 @@ class RGXWalker {
         if (!silent) {
             this.captures.push(captureResult);
             if (isPart && token.hasId()) {
-                this.namedCaptures[token.id] = captureResult;
+                if (!(token.id in this.namedCaptures))
+                    this.namedCaptures[token.id] = [];
+                this.namedCaptures[token.id].push(captureResult);
             }
         }
         // Notify Part after capture

package/dist/walker/part.d.ts

@@ -7,6 +7,7 @@ export type RGXCapture<T = unknown> = {
     value: T;
     start: number;
     end: number;
+    ownerId: string | null;
 };
 export type RGXPartOptions<R, T = string> = {
     id: string;

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@ptolemy2002/rgx",
-    "version": "6.2.0",
+    "version": "7.1.0",
     "private": false,
     "main": "dist/index.js",
     "types": "dist/index.d.ts",