node-html-parser 5.3.1 → 5.4.0

package/README.md

@@ -19,18 +19,20 @@ npm install --save node-html-parser
 
 ## Performance
 
+-- 2022-08-10
+
 ```shell
-cheerio         :12.0726 ms/file ± 7.31605
-parse5          :8.18615 ms/file ± 6.15337
-node-html-parser (last release):2.16533 ms/file ± 1.56924
-htmlparser      :17.0658 ms/file ± 120.901
-htmlparser2     :2.62695 ms/file ± 4.17579
-node-html-parser:2.14907 ms/file ± 1.66632
-html-parser     :24.6505 ms/file ± 18.9996
-htmljs-parser   :5.81797 ms/file ± 6.55537
-html-dom-parser :2.52265 ms/file ± 3.54858
-html5parser     :2.01144 ms/file ± 2.53570
-high5           :3.91342 ms/file ± 2.65563
+html-parser     :24.1595 ms/file ± 18.7667
+htmljs-parser   :4.72064 ms/file ± 5.67689
+html-dom-parser :2.18055 ms/file ± 2.96136
+html5parser     :1.69639 ms/file ± 2.17111
+cheerio         :12.2122 ms/file ± 8.10916
+parse5          :6.50626 ms/file ± 4.02352
+htmlparser2     :2.38179 ms/file ± 3.42389
+htmlparser      :17.4820 ms/file ± 128.041
+high5           :3.95188 ms/file ± 2.52313
+node-html-parser:2.04288 ms/file ± 1.25203
+node-html-parser (last release):2.00527 ms/file ± 1.21317
 ```
 
 Tested with [htmlparser-benchmark](https://github.com/AndreasMadsen/htmlparser-benchmark).
@@ -82,6 +84,10 @@ Parse the data provided, and return the root of the generated DOM.
   {
     lowerCaseTagName: false,  // convert tag name to lower case (hurts performance heavily)
     comment: false,            // retrieve comments (hurts performance slightly)
+    voidTag:{
+      tags: ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr'],	// optional and case insensitive, default value is ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr']
+      addClosingSlash: true     // optional, default false. void tag serialisation, add a final slash <br/>
+    },
     blockTextElements: {
       script: true,	// keep text content when parsing
       noscript: true,	// keep text content when parsing

package/dist/main.js

@@ -73,6 +73,20 @@ define("nodes/node", ["require", "exports", "he"], function (require, exports, h
                 value: range !== null && range !== void 0 ? range : [-1, -1]
             });
         }
+        /**
+         * Remove current node
+         */
+        Node.prototype.remove = function () {
+            var _this = this;
+            if (this.parentNode) {
+                var children = this.parentNode.childNodes;
+                this.parentNode.childNodes = children.filter(function (child) {
+                    return _this !== child;
+                });
+                this.parentNode = null;
+            }
+            return this;
+        };
         Object.defineProperty(Node.prototype, "innerText", {
             get: function () {
                 return this.rawText;
@@ -199,6 +213,37 @@ define("matcher", ["require", "exports", "nodes/type"], function (require, expor
         findAll: findAll
     };
 });
+define("void-tag", ["require", "exports"], function (require, exports) {
+    "use strict";
+    Object.defineProperty(exports, "__esModule", { value: true });
+    var VoidTag = /** @class */ (function () {
+        function VoidTag(addClosingSlash, tags) {
+            if (addClosingSlash === void 0) { addClosingSlash = false; }
+            this.addClosingSlash = addClosingSlash;
+            if (Array.isArray(tags)) {
+                this.voidTags = tags.reduce(function (set, tag) {
+                    return set.add(tag.toLowerCase());
+                }, new Set());
+            }
+            else {
+                this.voidTags = ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr'].reduce(function (set, tag) {
+                    return set.add(tag);
+                }, new Set());
+            }
+        }
+        VoidTag.prototype.formatNode = function (tag, attrs, innerHTML) {
+            var addClosingSlash = this.addClosingSlash;
+            var closingSpace = (addClosingSlash && attrs && !attrs.endsWith(' ')) ? ' ' : '';
+            var closingSlash = addClosingSlash ? "".concat(closingSpace, "/") : '';
+            return this.isVoidElement(tag.toLowerCase()) ? "<".concat(tag).concat(attrs).concat(closingSlash, ">") : "<".concat(tag).concat(attrs, ">").concat(innerHTML, "</").concat(tag, ">");
+        };
+        VoidTag.prototype.isVoidElement = function (tag) {
+            return this.voidTags.has(tag);
+        };
+        return VoidTag;
+    }());
+    exports.default = VoidTag;
+});
 define("nodes/text", ["require", "exports", "he", "nodes/node", "nodes/type"], function (require, exports, he_2, node_1, type_2) {
     "use strict";
     Object.defineProperty(exports, "__esModule", { value: true });
@@ -324,18 +369,18 @@ define("nodes/text", ["require", "exports", "he", "nodes/node", "nodes/type"], f
         return (hasLeadingSpace ? ' ' : '') + text.slice(startPos, endPos + 1) + (hasTrailingSpace ? ' ' : '');
     }
 });
-define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher", "nodes/comment", "nodes/node", "nodes/text", "nodes/type"], function (require, exports, css_select_1, he_3, back_1, matcher_1, comment_1, node_2, text_1, type_3) {
+define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher", "void-tag", "nodes/comment", "nodes/node", "nodes/text", "nodes/type"], function (require, exports, css_select_1, he_3, back_1, matcher_1, void_tag_1, comment_1, node_2, text_1, type_3) {
     "use strict";
     Object.defineProperty(exports, "__esModule", { value: true });
     exports.parse = exports.base_parse = void 0;
     he_3 = __importDefault(he_3);
     back_1 = __importDefault(back_1);
     matcher_1 = __importDefault(matcher_1);
+    void_tag_1 = __importDefault(void_tag_1);
     comment_1 = __importDefault(comment_1);
     node_2 = __importDefault(node_2);
     text_1 = __importDefault(text_1);
     type_3 = __importDefault(type_3);
-    var voidTags = new Set(['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr']);
     function decode(val) {
         // clone string
         return JSON.parse(JSON.stringify(he_3.default.decode(val)));
@@ -442,10 +487,12 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
          *
          * @memberof HTMLElement
          */
-        function HTMLElement(tagName, keyAttrs, rawAttrs, parentNode, range) {
+        function HTMLElement(tagName, keyAttrs, rawAttrs, parentNode, range, voidTag) {
             if (rawAttrs === void 0) { rawAttrs = ''; }
+            if (voidTag === void 0) { voidTag = new void_tag_1.default(); }
             var _this = _super.call(this, parentNode, range) || this;
             _this.rawAttrs = rawAttrs;
+            _this.voidTag = voidTag;
             /**
              * Node Type declaration.
              */
@@ -485,20 +532,6 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
             }
             return JSON.stringify(attr.replace(/"/g, '&quot;'));
         };
-        /**
-         * Remove current element
-         */
-        HTMLElement.prototype.remove = function () {
-            var _this = this;
-            if (this.parentNode) {
-                var children = this.parentNode.childNodes;
-                this.parentNode.childNodes = children.filter(function (child) {
-                    return _this !== child;
-                });
-                this.parentNode = null;
-            }
-            return this;
-        };
         /**
          * Remove Child element from childNodes array
          * @param {HTMLElement} node     node to remove
@@ -543,7 +576,7 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
         });
         Object.defineProperty(HTMLElement.prototype, "isVoidElement", {
             get: function () {
-                return voidTags.has(this.localName);
+                return this.voidTag.isVoidElement(this.localName);
             },
             enumerable: false,
             configurable: true
@@ -636,7 +669,7 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
             var tag = this.rawTagName;
             if (tag) {
                 var attrs = this.rawAttrs ? " ".concat(this.rawAttrs) : '';
-                return this.isVoidElement ? "<".concat(tag).concat(attrs, ">") : "<".concat(tag).concat(attrs, ">").concat(this.innerHTML, "</").concat(tag, ">");
+                return this.voidTag.formatNode(tag, attrs, this.innerHTML);
             }
             return this.innerHTML;
         };
@@ -936,7 +969,8 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
          * @return {Node}      node appended
          */
         HTMLElement.prototype.appendChild = function (node) {
-            // node.parentNode = this;
+            // remove the node from it's parent
+            node.remove();
             this.childNodes.push(node);
             node.parentNode = this;
             return node;
@@ -1332,7 +1366,9 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
      * @return {HTMLElement}      root element
      */
     function base_parse(data, options) {
+        var _a, _b;
         if (options === void 0) { options = { lowerCaseTagName: false, comment: false }; }
+        var voidTag = new void_tag_1.default((_a = options === null || options === void 0 ? void 0 : options.voidTag) === null || _a === void 0 ? void 0 : _a.closingSlash, (_b = options === null || options === void 0 ? void 0 : options.voidTag) === null || _b === void 0 ? void 0 : _b.tags);
         var elements = options.blockTextElements || {
             script: true,
             noscript: true,
@@ -1349,7 +1385,7 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
             return kBlockTextElements.some(function (it) { return it.test(tag); });
         }
         var createRange = function (startPos, endPos) { return [startPos - frameFlagOffset, endPos - frameFlagOffset]; };
-        var root = new HTMLElement(null, {}, '', null, [0, data.length]);
+        var root = new HTMLElement(null, {}, '', null, [0, data.length], voidTag);
         var currentParent = root;
         var stack = [root];
         var lastTextPos = -1;
@@ -1420,7 +1456,7 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
                 var tagStartPos_1 = tagEndPos_1 - matchLength;
                 currentParent = currentParent.appendChild(
                 // Initialize range (end position updated later for closed tags)
-                new HTMLElement(tagName, attrs, attributes.slice(1), null, createRange(tagStartPos_1, tagEndPos_1)));
+                new HTMLElement(tagName, attrs, attributes.slice(1), null, createRange(tagStartPos_1, tagEndPos_1), voidTag));
                 stack.push(currentParent);
                 if (is_block_text_element(tagName)) {
                     // Find closing tag

package/dist/nodes/html.d.ts

@@ -1,3 +1,4 @@
+import VoidTag from '../void-tag';
 import Node from './node';
 import NodeType from './type';
 export interface KeyAttributes {
@@ -37,6 +38,7 @@ declare class DOMTokenList {
  */
 export default class HTMLElement extends Node {
     rawAttrs: string;
+    private voidTag;
     private _attrs;
     private _rawAttrs;
     rawTagName: string;
@@ -59,11 +61,7 @@ export default class HTMLElement extends Node {
      *
      * @memberof HTMLElement
      */
-    constructor(tagName: string, keyAttrs: KeyAttributes, rawAttrs: string, parentNode: HTMLElement | null, range?: [number, number]);
-    /**
-     * Remove current element
-     */
-    remove(): this;
+    constructor(tagName: string, keyAttrs: KeyAttributes, rawAttrs: string, parentNode: HTMLElement | null, range: [number, number], voidTag?: VoidTag);
     /**
      * Remove Child element from childNodes array
      * @param {HTMLElement} node     node to remove
@@ -99,7 +97,7 @@ export default class HTMLElement extends Node {
     toString(): string;
     get innerHTML(): string;
     set innerHTML(content: string);
-    set_content(content: string | Node | Node[], options?: Options): this;
+    set_content(content: string | Node | Node[], options?: Partial<Options>): this;
     replaceWith(...nodes: (string | Node)[]): void;
     get outerHTML(): string;
     /**
@@ -209,6 +207,16 @@ export interface Options {
     blockTextElements: {
         [tag: string]: boolean;
     };
+    voidTag?: {
+        /**
+         * options, default value is ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr']
+         */
+        tags?: string[];
+        /**
+         * void tag serialisation, add a final slash <br/>
+         */
+        closingSlash?: boolean;
+    };
 }
 /**
  * Parses HTML and returns a root element

package/dist/nodes/html.js

@@ -43,11 +43,11 @@ var css_select_1 = require("css-select");
 var he_1 = __importDefault(require("he"));
 var back_1 = __importDefault(require("../back"));
 var matcher_1 = __importDefault(require("../matcher"));
+var void_tag_1 = __importDefault(require("../void-tag"));
 var comment_1 = __importDefault(require("./comment"));
 var node_1 = __importDefault(require("./node"));
 var text_1 = __importDefault(require("./text"));
 var type_1 = __importDefault(require("./type"));
-var voidTags = new Set(['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr']);
 function decode(val) {
     // clone string
     return JSON.parse(JSON.stringify(he_1.default.decode(val)));
@@ -154,10 +154,12 @@ var HTMLElement = /** @class */ (function (_super) {
      *
      * @memberof HTMLElement
      */
-    function HTMLElement(tagName, keyAttrs, rawAttrs, parentNode, range) {
+    function HTMLElement(tagName, keyAttrs, rawAttrs, parentNode, range, voidTag) {
         if (rawAttrs === void 0) { rawAttrs = ''; }
+        if (voidTag === void 0) { voidTag = new void_tag_1.default(); }
         var _this = _super.call(this, parentNode, range) || this;
         _this.rawAttrs = rawAttrs;
+        _this.voidTag = voidTag;
         /**
          * Node Type declaration.
          */
@@ -197,20 +199,6 @@ var HTMLElement = /** @class */ (function (_super) {
         }
         return JSON.stringify(attr.replace(/"/g, '"'));
     };
-    /**
-     * Remove current element
-     */
-    HTMLElement.prototype.remove = function () {
-        var _this = this;
-        if (this.parentNode) {
-            var children = this.parentNode.childNodes;
-            this.parentNode.childNodes = children.filter(function (child) {
-                return _this !== child;
-            });
-            this.parentNode = null;
-        }
-        return this;
-    };
     /**
      * Remove Child element from childNodes array
      * @param {HTMLElement} node     node to remove
@@ -255,7 +243,7 @@ var HTMLElement = /** @class */ (function (_super) {
     });
     Object.defineProperty(HTMLElement.prototype, "isVoidElement", {
         get: function () {
-            return voidTags.has(this.localName);
+            return this.voidTag.isVoidElement(this.localName);
         },
         enumerable: false,
         configurable: true
@@ -348,7 +336,7 @@ var HTMLElement = /** @class */ (function (_super) {
         var tag = this.rawTagName;
         if (tag) {
             var attrs = this.rawAttrs ? " ".concat(this.rawAttrs) : '';
-            return this.isVoidElement ? "<".concat(tag).concat(attrs, ">") : "<".concat(tag).concat(attrs, ">").concat(this.innerHTML, "</").concat(tag, ">");
+            return this.voidTag.formatNode(tag, attrs, this.innerHTML);
         }
         return this.innerHTML;
     };
@@ -648,7 +636,8 @@ var HTMLElement = /** @class */ (function (_super) {
      * @return {Node}      node appended
      */
     HTMLElement.prototype.appendChild = function (node) {
-        // node.parentNode = this;
+        // remove the node from it's parent
+        node.remove();
         this.childNodes.push(node);
         node.parentNode = this;
         return node;
@@ -1044,7 +1033,9 @@ var frameflag = 'documentfragmentcontainer';
  * @return {HTMLElement}      root element
  */
 function base_parse(data, options) {
+    var _a, _b;
     if (options === void 0) { options = { lowerCaseTagName: false, comment: false }; }
+    var voidTag = new void_tag_1.default((_a = options === null || options === void 0 ? void 0 : options.voidTag) === null || _a === void 0 ? void 0 : _a.closingSlash, (_b = options === null || options === void 0 ? void 0 : options.voidTag) === null || _b === void 0 ? void 0 : _b.tags);
     var elements = options.blockTextElements || {
         script: true,
         noscript: true,
@@ -1061,7 +1052,7 @@ function base_parse(data, options) {
         return kBlockTextElements.some(function (it) { return it.test(tag); });
     }
     var createRange = function (startPos, endPos) { return [startPos - frameFlagOffset, endPos - frameFlagOffset]; };
-    var root = new HTMLElement(null, {}, '', null, [0, data.length]);
+    var root = new HTMLElement(null, {}, '', null, [0, data.length], voidTag);
     var currentParent = root;
     var stack = [root];
     var lastTextPos = -1;
@@ -1132,7 +1123,7 @@ function base_parse(data, options) {
             var tagStartPos_1 = tagEndPos_1 - matchLength;
             currentParent = currentParent.appendChild(
             // Initialize range (end position updated later for closed tags)
-            new HTMLElement(tagName, attrs, attributes.slice(1), null, createRange(tagStartPos_1, tagEndPos_1)));
+            new HTMLElement(tagName, attrs, attributes.slice(1), null, createRange(tagStartPos_1, tagEndPos_1), voidTag));
             stack.push(currentParent);
             if (is_block_text_element(tagName)) {
                 // Find closing tag

package/dist/nodes/node.d.ts

@@ -13,6 +13,10 @@ export default abstract class Node {
     abstract toString(): string;
     abstract clone(): Node;
     constructor(parentNode?: HTMLElement, range?: [number, number]);
+    /**
+     * Remove current node
+     */
+    remove(): this;
     get innerText(): string;
     get textContent(): string;
     set textContent(val: string);

package/dist/nodes/node.js

@@ -16,6 +16,20 @@ var Node = /** @class */ (function () {
             value: range !== null && range !== void 0 ? range : [-1, -1]
         });
     }
+    /**
+     * Remove current node
+     */
+    Node.prototype.remove = function () {
+        var _this = this;
+        if (this.parentNode) {
+            var children = this.parentNode.childNodes;
+            this.parentNode.childNodes = children.filter(function (child) {
+                return _this !== child;
+            });
+            this.parentNode = null;
+        }
+        return this;
+    };
     Object.defineProperty(Node.prototype, "innerText", {
         get: function () {
             return this.rawText;

package/dist/void-tag.d.ts

@@ -0,0 +1,7 @@
+export default class VoidTag {
+    addClosingSlash: boolean;
+    private voidTags;
+    constructor(addClosingSlash?: boolean, tags?: string[]);
+    formatNode(tag: string, attrs: string, innerHTML: string): string;
+    isVoidElement(tag: string): boolean;
+}

package/dist/void-tag.js

@@ -0,0 +1,29 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+var VoidTag = /** @class */ (function () {
+    function VoidTag(addClosingSlash, tags) {
+        if (addClosingSlash === void 0) { addClosingSlash = false; }
+        this.addClosingSlash = addClosingSlash;
+        if (Array.isArray(tags)) {
+            this.voidTags = tags.reduce(function (set, tag) {
+                return set.add(tag.toLowerCase());
+            }, new Set());
+        }
+        else {
+            this.voidTags = ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr'].reduce(function (set, tag) {
+                return set.add(tag);
+            }, new Set());
+        }
+    }
+    VoidTag.prototype.formatNode = function (tag, attrs, innerHTML) {
+        var addClosingSlash = this.addClosingSlash;
+        var closingSpace = (addClosingSlash && attrs && !attrs.endsWith(' ')) ? ' ' : '';
+        var closingSlash = addClosingSlash ? "".concat(closingSpace, "/") : '';
+        return this.isVoidElement(tag.toLowerCase()) ? "<".concat(tag).concat(attrs).concat(closingSlash, ">") : "<".concat(tag).concat(attrs, ">").concat(innerHTML, "</").concat(tag, ">");
+    };
+    VoidTag.prototype.isVoidElement = function (tag) {
+        return this.voidTags.has(tag);
+    };
+    return VoidTag;
+}());
+exports.default = VoidTag;

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "node-html-parser",
-	"version": "5.3.1",
+	"version": "5.4.0",
 	"description": "A very fast HTML parser, generating a simplified DOM, with basic element query support.",
 	"main": "dist/index.js",
 	"types": "dist/index.d.ts",
@@ -31,8 +31,8 @@
 		"prepare": "cd test && yarn install"
 	},
 	"keywords": [
-		"parser",
 		"html",
+		"parser",
 		"nodejs",
 		"typescript"
 	],