node-html-parser 4.1.5

package/dist/matcher.d.ts

@@ -0,0 +1,6 @@
+import { Adapter } from 'css-select/lib/types';
+import HTMLElement from './nodes/html';
+import Node from './nodes/node';
+export declare type Predicate = (node: Node) => node is HTMLElement;
+declare const _default: Adapter<Node, HTMLElement>;
+export default _default;

package/dist/matcher.js

@@ -0,0 +1,106 @@
+"use strict";
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+var type_1 = __importDefault(require("./nodes/type"));
+function isTag(node) {
+    return node && node.nodeType === type_1.default.ELEMENT_NODE;
+}
+function getAttributeValue(elem, name) {
+    return isTag(elem) ? elem.getAttribute(name) : undefined;
+}
+function getName(elem) {
+    return ((elem && elem.rawTagName) || '').toLowerCase();
+}
+function getChildren(node) {
+    return node && node.childNodes;
+}
+function getParent(node) {
+    return node ? node.parentNode : null;
+}
+function getText(node) {
+    return node.text;
+}
+function removeSubsets(nodes) {
+    var idx = nodes.length;
+    var node;
+    var ancestor;
+    var replace;
+    // Check if each node (or one of its ancestors) is already contained in the
+    // array.
+    while (--idx > -1) {
+        node = ancestor = nodes[idx];
+        // Temporarily remove the node under consideration
+        nodes[idx] = null;
+        replace = true;
+        while (ancestor) {
+            if (nodes.indexOf(ancestor) > -1) {
+                replace = false;
+                nodes.splice(idx, 1);
+                break;
+            }
+            ancestor = getParent(ancestor);
+        }
+        // If the node has been found to be unique, re-insert it.
+        if (replace) {
+            nodes[idx] = node;
+        }
+    }
+    return nodes;
+}
+function existsOne(test, elems) {
+    return elems.some(function (elem) {
+        return isTag(elem) ? test(elem) || existsOne(test, getChildren(elem)) : false;
+    });
+}
+function getSiblings(node) {
+    var parent = getParent(node);
+    return parent && getChildren(parent);
+}
+function hasAttrib(elem, name) {
+    return getAttributeValue(elem, name) !== undefined;
+}
+function findOne(test, elems) {
+    var elem = null;
+    for (var i = 0, l = elems.length; i < l && !elem; i++) {
+        var el = elems[i];
+        if (test(el)) {
+            elem = el;
+        }
+        else {
+            var childs = getChildren(el);
+            if (childs && childs.length > 0) {
+                elem = findOne(test, childs);
+            }
+        }
+    }
+    return elem;
+}
+function findAll(test, nodes) {
+    var result = [];
+    for (var i = 0, j = nodes.length; i < j; i++) {
+        if (!isTag(nodes[i]))
+            continue;
+        if (test(nodes[i]))
+            result.push(nodes[i]);
+        var childs = getChildren(nodes[i]);
+        if (childs)
+            result = result.concat(findAll(test, childs));
+    }
+    return result;
+}
+exports.default = {
+    isTag: isTag,
+    getAttributeValue: getAttributeValue,
+    getName: getName,
+    getChildren: getChildren,
+    getParent: getParent,
+    getText: getText,
+    removeSubsets: removeSubsets,
+    existsOne: existsOne,
+    getSiblings: getSiblings,
+    hasAttrib: hasAttrib,
+    findOne: findOne,
+    findAll: findAll
+};

package/dist/nodes/comment.d.ts

@@ -0,0 +1,18 @@
+import Node from './node';
+import NodeType from './type';
+import HTMLElement from './html';
+export default class CommentNode extends Node {
+    rawText: string;
+    constructor(rawText: string, parentNode: HTMLElement, range?: [number, number]);
+    /**
+     * Node Type declaration.
+     * @type {Number}
+     */
+    nodeType: NodeType;
+    /**
+     * Get unescaped text value of current node and its children.
+     * @return {string} text content
+     */
+    get text(): string;
+    toString(): string;
+}

package/dist/nodes/comment.js

@@ -0,0 +1,51 @@
+"use strict";
+var __extends = (this && this.__extends) || (function () {
+    var extendStatics = function (d, b) {
+        extendStatics = Object.setPrototypeOf ||
+            ({ __proto__: [] } instanceof Array && function (d, b) { d.__proto__ = b; }) ||
+            function (d, b) { for (var p in b) if (Object.prototype.hasOwnProperty.call(b, p)) d[p] = b[p]; };
+        return extendStatics(d, b);
+    };
+    return function (d, b) {
+        if (typeof b !== "function" && b !== null)
+            throw new TypeError("Class extends value " + String(b) + " is not a constructor or null");
+        extendStatics(d, b);
+        function __() { this.constructor = d; }
+        d.prototype = b === null ? Object.create(b) : (__.prototype = b.prototype, new __());
+    };
+})();
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+var node_1 = __importDefault(require("./node"));
+var type_1 = __importDefault(require("./type"));
+var CommentNode = /** @class */ (function (_super) {
+    __extends(CommentNode, _super);
+    function CommentNode(rawText, parentNode, range) {
+        var _this = _super.call(this, parentNode, range) || this;
+        _this.rawText = rawText;
+        /**
+         * Node Type declaration.
+         * @type {Number}
+         */
+        _this.nodeType = type_1.default.COMMENT_NODE;
+        return _this;
+    }
+    Object.defineProperty(CommentNode.prototype, "text", {
+        /**
+         * Get unescaped text value of current node and its children.
+         * @return {string} text content
+         */
+        get: function () {
+            return this.rawText;
+        },
+        enumerable: false,
+        configurable: true
+    });
+    CommentNode.prototype.toString = function () {
+        return "<!--" + this.rawText + "-->";
+    };
+    return CommentNode;
+}(node_1.default));
+exports.default = CommentNode;

package/dist/nodes/html.d.ts

@@ -0,0 +1,206 @@
+import Node from './node';
+import NodeType from './type';
+export interface KeyAttributes {
+    id?: string;
+    class?: string;
+}
+export interface Attributes {
+    [key: string]: string;
+}
+export interface RawAttributes {
+    [key: string]: string;
+}
+export declare type InsertPosition = 'beforebegin' | 'afterbegin' | 'beforeend' | 'afterend';
+declare class DOMTokenList {
+    private _set;
+    private _afterUpdate;
+    private _validate;
+    constructor(valuesInit?: string[], afterUpdate?: (t: DOMTokenList) => void);
+    add(c: string): void;
+    replace(c1: string, c2: string): void;
+    remove(c: string): void;
+    toggle(c: string): void;
+    contains(c: string): boolean;
+    get length(): number;
+    values(): IterableIterator<string>;
+    get value(): string[];
+    toString(): string;
+}
+/**
+ * HTMLElement, which contains a set of children.
+ *
+ * Note: this is a minimalist implementation, no complete tree
+ *   structure provided (no parentNode, nextSibling,
+ *   previousSibling etc).
+ * @class HTMLElement
+ * @extends {Node}
+ */
+export default class HTMLElement extends Node {
+    private rawAttrs;
+    private _attrs;
+    private _rawAttrs;
+    rawTagName: string;
+    id: string;
+    classList: DOMTokenList;
+    /**
+     * Node Type declaration.
+     */
+    nodeType: NodeType;
+    /**
+     * Quote attribute values
+     * @param attr attribute value
+     * @returns {string} quoted value
+     */
+    private quoteAttribute;
+    /**
+     * Creates an instance of HTMLElement.
+     * @param keyAttrs	id and class attribute
+     * @param [rawAttrs]	attributes in string
+     *
+     * @memberof HTMLElement
+     */
+    constructor(tagName: string, keyAttrs: KeyAttributes, rawAttrs: string, parentNode: HTMLElement | null, range?: [number, number]);
+    /**
+     * Remove current element
+     */
+    remove(): void;
+    /**
+     * Remove Child element from childNodes array
+     * @param {HTMLElement} node     node to remove
+     */
+    removeChild(node: Node): void;
+    /**
+     * Exchanges given child with new child
+     * @param {HTMLElement} oldNode     node to exchange
+     * @param {HTMLElement} newNode     new node
+     */
+    exchangeChild(oldNode: Node, newNode: Node): void;
+    get tagName(): string;
+    get localName(): string;
+    /**
+     * Get escpaed (as-it) text value of current node and its children.
+     * @return {string} text content
+     */
+    get rawText(): string;
+    get textContent(): string;
+    set textContent(val: string);
+    /**
+     * Get unescaped text value of current node and its children.
+     * @return {string} text content
+     */
+    get text(): string;
+    /**
+     * Get structured Text (with '\n' etc.)
+     * @return {string} structured text
+     */
+    get structuredText(): string;
+    toString(): string;
+    get innerHTML(): string;
+    set innerHTML(content: string);
+    set_content(content: string | Node | Node[], options?: Options): void;
+    replaceWith(...nodes: (string | Node)[]): void;
+    get outerHTML(): string;
+    /**
+     * Trim element from right (in block) after seeing pattern in a TextNode.
+     * @param  {RegExp} pattern pattern to find
+     * @return {HTMLElement}    reference to current node
+     */
+    trimRight(pattern: RegExp): this;
+    /**
+     * Get DOM structure
+     * @return {string} strucutre
+     */
+    get structure(): string;
+    /**
+     * Remove whitespaces in this sub tree.
+     * @return {HTMLElement} pointer to this
+     */
+    removeWhitespace(): this;
+    /**
+     * Query CSS selector to find matching nodes.
+     * @param  {string}         selector Simplified CSS selector
+     * @return {HTMLElement[]}  matching elements
+     */
+    querySelectorAll(selector: string): HTMLElement[];
+    /**
+     * Query CSS Selector to find matching node.
+     * @param  {string}         selector Simplified CSS selector
+     * @return {HTMLElement}    matching node
+     */
+    querySelector(selector: string): HTMLElement;
+    /**
+     * traverses the Element and its parents (heading toward the document root) until it finds a node that matches the provided selector string. Will return itself or the matching ancestor. If no such element exists, it returns null.
+     * @param selector a DOMString containing a selector list
+     */
+    closest(selector: string): Node;
+    /**
+     * Append a child node to childNodes
+     * @param  {Node} node node to append
+     * @return {Node}      node appended
+     */
+    appendChild<T extends Node = Node>(node: T): T;
+    /**
+     * Get first child node
+     * @return {Node} first child node
+     */
+    get firstChild(): Node;
+    /**
+     * Get last child node
+     * @return {Node} last child node
+     */
+    get lastChild(): Node;
+    /**
+     * Get attributes
+     * @access private
+     * @return {Object} parsed and unescaped attributes
+     */
+    get attrs(): Attributes;
+    get attributes(): Record<string, string>;
+    /**
+     * Get escaped (as-it) attributes
+     * @return {Object} parsed attributes
+     */
+    get rawAttributes(): RawAttributes;
+    removeAttribute(key: string): void;
+    hasAttribute(key: string): boolean;
+    /**
+     * Get an attribute
+     * @return {string} value of the attribute
+     */
+    getAttribute(key: string): string | undefined;
+    /**
+     * Set an attribute value to the HTMLElement
+     * @param {string} key The attribute name
+     * @param {string} value The value to set, or null / undefined to remove an attribute
+     */
+    setAttribute(key: string, value: string): void;
+    /**
+     * Replace all the attributes of the HTMLElement by the provided attributes
+     * @param {Attributes} attributes the new attribute set
+     */
+    setAttributes(attributes: Attributes): void;
+    insertAdjacentHTML(where: InsertPosition, html: string): void;
+    get nextSibling(): Node;
+    get nextElementSibling(): HTMLElement;
+    get classNames(): string;
+}
+export interface Options {
+    lowerCaseTagName: boolean;
+    comment: boolean;
+    blockTextElements: {
+        [tag: string]: boolean;
+    };
+}
+/**
+ * Parses HTML and returns a root element
+ * Parse a chuck of HTML source.
+ * @param  {string} data      html
+ * @return {HTMLElement}      root element
+ */
+export declare function base_parse(data: string, options?: Partial<Options>): HTMLElement[];
+/**
+ * Parses HTML and returns a root element
+ * Parse a chuck of HTML source.
+ */
+export declare function parse(data: string, options?: Partial<Options>): HTMLElement;
+export {};