1 /*
2 * Copyright (c) 2002-2025 Gargoyle Software Inc.
3 *
4 * Licensed under the Apache License, Version 2.0 (the "License");
5 * you may not use this file except in compliance with the License.
6 * You may obtain a copy of the License at
7 * https://www.apache.org/licenses/LICENSE-2.0
8 *
9 * Unless required by applicable law or agreed to in writing, software
10 * distributed under the License is distributed on an "AS IS" BASIS,
11 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12 * See the License for the specific language governing permissions and
13 * limitations under the License.
14 */
15 package org.htmlunit.javascript.host.dom;
16
17 import java.io.IOException;
18
19 import org.htmlunit.StringWebResponse;
20 import org.htmlunit.WebClient;
21 import org.htmlunit.WebResponse;
22 import org.htmlunit.WebWindow;
23 import org.htmlunit.html.HtmlPage;
24 import org.htmlunit.html.parser.HTMLParser;
25 import org.htmlunit.javascript.HtmlUnitScriptable;
26 import org.htmlunit.javascript.JavaScriptEngine;
27 import org.htmlunit.javascript.configuration.JsxClass;
28 import org.htmlunit.javascript.configuration.JsxConstructor;
29 import org.htmlunit.javascript.configuration.JsxFunction;
30 import org.htmlunit.javascript.host.Window;
31 import org.htmlunit.javascript.host.html.HTMLDocument;
32 import org.htmlunit.javascript.host.xml.XMLDocument;
33 import org.htmlunit.util.MimeType;
34
35 /**
36 * A JavaScript object for {@code DOMParser}.
37 *
38 * @author Ahmed Ashour
39 * @author Frank Danek
40 * @author Ronald Brill
41 *
42 * @see <a href="http://www.w3.org/TR/DOM-Parsing/">W3C Spec</a>
43 * @see <a href="http://domparsing.spec.whatwg.org/">WhatWG Spec</a>
44 * @see <a href="https://developer.mozilla.org/en-US/docs/Web/API/DOMParser">Mozilla Developer Network</a>
45 * @see <a href="http://msdn.microsoft.com/en-us/library/ff975060.aspx">MSDN</a>
46 * @see <a href="http://www.xulplanet.com/references/objref/DOMParser.html">XUL Planet</a>
47 */
48 @JsxClass
49 public class DOMParser extends HtmlUnitScriptable {
50
51 /**
52 * JavaScript constructor.
53 */
54 @JsxConstructor
55 public void jsConstructor() {
56 // nothing to do
57 }
58
59 /**
60 * Parses the given Unicode string into a DOM document.
61 * @param str the Unicode string to be parsed
62 * @param type the MIME type of the string -
63 * <code>text/html</code>, <code>text/xml</code>, <code>application/xml</code>,
64 * <code>application/xhtml+xml</code>, <code>image/svg+xml</code>. Must not be {@code null}.
65 * @return the generated document
66 */
67 @JsxFunction
68 public Document parseFromString(final String str, final Object type) {
69 try {
70 final Document document = parseFromString(this, str, type);
71 if (document == null) {
72 throw JavaScriptEngine.typeError("Invalid 'type' parameter: " + type);
73 }
74 return document;
75 }
76 catch (final IOException e) {
77 throw JavaScriptEngine.syntaxError("Parsing failed" + e.getMessage());
78 }
79 }
80
81 /**
82 * <span style="color:red">INTERNAL API - SUBJECT TO CHANGE AT ANY TIME - USE AT YOUR OWN RISK.</span><br>
83 *
84 * Parses the given Unicode string into a DOM document.
85 * @param scriptable the ScriptableObject this belongs to
86 * @param str the Unicode string to be parsed
87 * @param type the MIME type of the string -
88 * <code>text/html</code>, <code>text/xml</code>, <code>application/xml</code>,
89 * <code>application/xhtml+xml</code>, <code>image/svg+xml</code>. Must not be {@code null}.
90 * @return the generated document
91 * @throws IOException in case of error
92 */
93 public static Document parseFromString(final HtmlUnitScriptable scriptable, final String str, final Object type)
94 throws IOException {
95 if (type == null || JavaScriptEngine.isUndefined(type)) {
96 throw JavaScriptEngine.typeError("Missing 'type' parameter");
97 }
98
99 if (MimeType.TEXT_XML.equals(type)
100 || MimeType.APPLICATION_XML.equals(type)
101 || MimeType.APPLICATION_XHTML.equals(type)
102 || "image/svg+xml".equals(type)) {
103 final XMLDocument document = new XMLDocument();
104 document.setParentScope(scriptable.getParentScope());
105 document.setPrototype(scriptable.getPrototype(XMLDocument.class));
106 document.loadXML(str);
107 return document;
108 }
109
110 if (MimeType.TEXT_HTML.equals(type)) {
111 final WebWindow webWindow = scriptable.getWindow().getWebWindow();
112 final WebResponse webResponse = new StringWebResponse(str, webWindow.getEnclosedPage().getUrl());
113
114 return parseHtmlDocument(scriptable, webResponse, webWindow);
115 }
116
117 return null;
118 }
119
120 /**
121 * <span style="color:red">INTERNAL API - SUBJECT TO CHANGE AT ANY TIME - USE AT YOUR OWN RISK.</span><br>
122 *
123 * Parses the given Unicode string into a DOM document.
124 * @param scriptable the ScriptableObject this belongs to
125 * @param webResponse the response to be parsed
126 * @param webWindow the window
127 * @return the generated document
128 * @throws IOException in case of error
129 */
130 public static Document parseHtmlDocument(final HtmlUnitScriptable scriptable, final WebResponse webResponse,
131 final WebWindow webWindow)
132 throws IOException {
133 // a similar impl is in
134 // org.htmlunit.javascript.host.dom.DOMImplementation.createHTMLDocument(Object)
135 final HtmlPage page = new HtmlPage(webResponse, webWindow);
136 page.setEnclosingWindow(null);
137 final Window window = webWindow.getScriptableObject();
138
139 // document knows the window but is not the windows document
140 final HTMLDocument document = new HTMLDocument();
141 document.setParentScope(window);
142 document.setPrototype(window.getPrototype(document.getClass()));
143 // document.setWindow(window);
144 document.setDomNode(page);
145
146 final WebClient webClient = webWindow.getWebClient();
147 final HTMLParser htmlParser = webClient.getPageCreator().getHtmlParser();
148 htmlParser.parse(webResponse, page, false, true);
149 return page.getScriptableObject();
150 }
151 }