node-html-parser 5.4.2 → 6.0.0

package/CHANGELOG.md

@@ -2,6 +2,13 @@
 
 All notable changes to this project will be documented in this file. See [standard-version](https://github.com/conventional-changelog/standard-version) for commit guidelines.
 
+## [6.0.0](https://github.com/taoqf/node-fast-html-parser/compare/v5.4.2-0...v6.0.0) (2022-09-08)
+
+
+### Bug Fixes
+
+* Preserve invalid nested A tags in AST (see [#215](https://github.com/taoqf/node-fast-html-parser/issues/215) for detail) ([374188f](https://github.com/taoqf/node-fast-html-parser/commit/374188f1c6d6c6d0567348b8e8d20957f5a93fb8))
+
 ### [5.4.2](https://github.com/taoqf/node-fast-html-parser/compare/v5.4.2-0...v5.4.2) (2022-08-30)
 
 ## [5.1.0](https://github.com/taoqf/node-fast-html-parser/compare/v4.1.5...v5.1.0) (2021-10-28)

package/dist/main.js

@@ -1393,7 +1393,7 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
         var match;
         // https://github.com/taoqf/node-html-parser/issues/38
         data = "<".concat(frameflag, ">").concat(data, "</").concat(frameflag, ">");
-        var lowerCaseTagName = options.lowerCaseTagName;
+        var lowerCaseTagName = options.lowerCaseTagName, fixNestedATags = options.fixNestedATags;
         var dataEndPos = data.length - (frameflag.length + 2);
         var frameFlagOffset = frameflag.length + 2;
         while ((match = kMarkupPattern.exec(data))) {
@@ -1445,7 +1445,7 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
                     }
                 }
                 // Prevent nested A tags by terminating the last A and starting a new one : see issue #144
-                if (tagName === 'a' || tagName === 'A') {
+                if (fixNestedATags && (tagName === 'a' || tagName === 'A')) {
                     if (noNestedTagIndex !== undefined) {
                         stack.splice(noNestedTagIndex);
                         currentParent = (0, back_1.default)(stack);
@@ -1484,7 +1484,7 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
             // Handle closing tags or self-closed elements (ie </tag> or <br>)
             if (leadingSlash || closingSlash || kSelfClosingElements[tagName]) {
                 while (true) {
-                    if (tagName === 'a' || tagName === 'A')
+                    if (noNestedTagIndex != null && (tagName === 'a' || tagName === 'A'))
                         noNestedTagIndex = undefined;
                     if (currentParent.rawTagName === tagName) {
                         // Update range end for closed tag

package/dist/nodes/html.d.ts

@@ -203,6 +203,10 @@ export default class HTMLElement extends Node {
 export interface Options {
     lowerCaseTagName: boolean;
     comment: boolean;
+    /**
+     * @see PR #215 for explanation
+     */
+    fixNestedATags?: boolean;
     parseNoneClosedTags?: boolean;
     blockTextElements: {
         [tag: string]: boolean;

package/dist/nodes/html.js

@@ -1060,7 +1060,7 @@ function base_parse(data, options) {
     var match;
     // https://github.com/taoqf/node-html-parser/issues/38
     data = "<".concat(frameflag, ">").concat(data, "</").concat(frameflag, ">");
-    var lowerCaseTagName = options.lowerCaseTagName;
+    var lowerCaseTagName = options.lowerCaseTagName, fixNestedATags = options.fixNestedATags;
     var dataEndPos = data.length - (frameflag.length + 2);
     var frameFlagOffset = frameflag.length + 2;
     while ((match = kMarkupPattern.exec(data))) {
@@ -1112,7 +1112,7 @@ function base_parse(data, options) {
                 }
             }
             // Prevent nested A tags by terminating the last A and starting a new one : see issue #144
-            if (tagName === 'a' || tagName === 'A') {
+            if (fixNestedATags && (tagName === 'a' || tagName === 'A')) {
                 if (noNestedTagIndex !== undefined) {
                     stack.splice(noNestedTagIndex);
                     currentParent = (0, back_1.default)(stack);
@@ -1151,7 +1151,7 @@ function base_parse(data, options) {
         // Handle closing tags or self-closed elements (ie </tag> or <br>)
         if (leadingSlash || closingSlash || kSelfClosingElements[tagName]) {
             while (true) {
-                if (tagName === 'a' || tagName === 'A')
+                if (noNestedTagIndex != null && (tagName === 'a' || tagName === 'A'))
                     noNestedTagIndex = undefined;
                 if (currentParent.rawTagName === tagName) {
                     // Update range end for closed tag

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "node-html-parser",
-	"version": "5.4.2",
+	"version": "6.0.0",
 	"description": "A very fast HTML parser, generating a simplified DOM, with basic element query support.",
 	"main": "dist/index.js",
 	"types": "dist/index.d.ts",