1 | // Copyright 2012 Georg-August-Universität Göttingen, Germany |
---|
2 | // |
---|
3 | // Licensed under the Apache License, Version 2.0 (the "License"); |
---|
4 | // you may not use this file except in compliance with the License. |
---|
5 | // You may obtain a copy of the License at |
---|
6 | // |
---|
7 | // http://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 de.ugoe.cs.autoquest.plugin.html; |
---|
16 | |
---|
17 | import java.util.List; |
---|
18 | import java.util.Map; |
---|
19 | import java.util.regex.Matcher; |
---|
20 | import java.util.regex.Pattern; |
---|
21 | |
---|
22 | import org.xml.sax.SAXException; |
---|
23 | |
---|
24 | import de.ugoe.cs.autoquest.eventcore.Event; |
---|
25 | import de.ugoe.cs.autoquest.eventcore.IEventType; |
---|
26 | import de.ugoe.cs.autoquest.eventcore.guimodel.GUIModel; |
---|
27 | import de.ugoe.cs.autoquest.eventcore.guimodel.IGUIElement; |
---|
28 | import de.ugoe.cs.autoquest.plugin.html.eventcore.HTMLEventTypeFactory; |
---|
29 | import de.ugoe.cs.autoquest.plugin.html.guimodel.HTMLDocumentSpec; |
---|
30 | import de.ugoe.cs.autoquest.plugin.html.guimodel.HTMLGUIElement; |
---|
31 | import de.ugoe.cs.autoquest.plugin.html.guimodel.HTMLGUIElementSpec; |
---|
32 | import de.ugoe.cs.autoquest.plugin.html.guimodel.HTMLPageElementSpec; |
---|
33 | import de.ugoe.cs.autoquest.plugin.html.guimodel.HTMLServerSpec; |
---|
34 | |
---|
35 | /** |
---|
36 | * <p> |
---|
37 | * This class provides the functionality to parse XML log files generated by the HTMLMonitor of |
---|
38 | * AutoQUEST. The result of parsing a file is a collection of event sequences and a GUI model |
---|
39 | * </p> |
---|
40 | * |
---|
41 | * @author Fabian Glaser, Patrick Harms |
---|
42 | * @version 1.0 |
---|
43 | * |
---|
44 | */ |
---|
45 | public class HTMLLogParser extends AbstractDefaultLogParser { |
---|
46 | |
---|
47 | /** |
---|
48 | * |
---|
49 | */ |
---|
50 | private Pattern htmlElementPattern = |
---|
51 | Pattern.compile("(\\w+)(\\[(\\d+)\\]|\\(htmlId=([\\w-]+)\\))"); |
---|
52 | |
---|
53 | /* (non-Javadoc) |
---|
54 | * @see de.ugoe.cs.autoquest.plugin.html.AbstractDefaultLogParser#handleGUIElement(java.lang.String, java.util.Map) |
---|
55 | */ |
---|
56 | @Override |
---|
57 | protected boolean handleGUIElement(String id, Map<String, String> parameters) |
---|
58 | throws SAXException |
---|
59 | { |
---|
60 | HTMLGUIElementSpec specification = null; |
---|
61 | |
---|
62 | String parentId = parameters.get("parent"); |
---|
63 | IGUIElement parent = super.getGUIElementTree().find(parentId); |
---|
64 | |
---|
65 | if (parameters.containsKey("host")) { |
---|
66 | // this is a server specification |
---|
67 | int port = 80; |
---|
68 | String portStr = parameters.get(port); |
---|
69 | |
---|
70 | if (portStr != null) { |
---|
71 | port = Integer.parseInt(portStr); |
---|
72 | } |
---|
73 | |
---|
74 | specification = new HTMLServerSpec(parameters.get("host"), port); |
---|
75 | } |
---|
76 | else if (parameters.containsKey("path")) { |
---|
77 | // this is a document specification |
---|
78 | |
---|
79 | if (parent != null) { |
---|
80 | if (!(parent.getSpecification() instanceof HTMLServerSpec)) { |
---|
81 | throw new SAXException |
---|
82 | ("invalid log: parent GUI element of a document is not of type server"); |
---|
83 | } |
---|
84 | |
---|
85 | specification = new HTMLDocumentSpec |
---|
86 | ((HTMLServerSpec) parent.getSpecification(), parameters.get("path"), |
---|
87 | parameters.get("query"), parameters.get("title")); |
---|
88 | } |
---|
89 | else if (parentId == null) { |
---|
90 | throw new SAXException("invalid log: a document has no parent id"); |
---|
91 | } |
---|
92 | } |
---|
93 | else if (parameters.containsKey("tagname")) { |
---|
94 | String tagName = parameters.get("tagname"); |
---|
95 | |
---|
96 | if (!tagNameMustBeConsidered(tagName)) { |
---|
97 | return true; |
---|
98 | } |
---|
99 | |
---|
100 | if (parent != null) { |
---|
101 | IGUIElement document = parent; |
---|
102 | |
---|
103 | while ((document != null) && |
---|
104 | (!(document.getSpecification() instanceof HTMLDocumentSpec))) |
---|
105 | { |
---|
106 | document = document.getParent(); |
---|
107 | } |
---|
108 | |
---|
109 | if (document == null) { |
---|
110 | throw new SAXException |
---|
111 | ("invalid log: parent hierarchy of a page element does not contain a " + |
---|
112 | "document"); |
---|
113 | } |
---|
114 | |
---|
115 | int index = -1; |
---|
116 | String indexStr = parameters.get("index"); |
---|
117 | |
---|
118 | if ((indexStr != null) && (!"".equals(indexStr))) { |
---|
119 | index = Integer.parseInt(indexStr); |
---|
120 | } |
---|
121 | |
---|
122 | specification = new HTMLPageElementSpec |
---|
123 | ((HTMLDocumentSpec) document.getSpecification(), tagName, |
---|
124 | parameters.get("htmlid"), index); |
---|
125 | } |
---|
126 | else if (parentId == null) { |
---|
127 | throw new SAXException("invalid log: a page element has no parent id"); |
---|
128 | } |
---|
129 | } |
---|
130 | else { |
---|
131 | throw new SAXException("invalid log: unknown GUI element"); |
---|
132 | } |
---|
133 | |
---|
134 | if (specification != null) { |
---|
135 | super.getGUIElementTree().add(id, parentId, specification); |
---|
136 | return true; |
---|
137 | } |
---|
138 | else { |
---|
139 | return false; |
---|
140 | } |
---|
141 | } |
---|
142 | |
---|
143 | /* (non-Javadoc) |
---|
144 | * @see de.ugoe.cs.autoquest.plugin.html.AbstractDefaultLogParser#handleEvent(java.lang.String, java.util.Map) |
---|
145 | */ |
---|
146 | @Override |
---|
147 | protected boolean handleEvent(String type, Map<String, String> parameters) throws SAXException { |
---|
148 | String targetId = parameters.get("target"); |
---|
149 | |
---|
150 | if (targetId == null) { |
---|
151 | String targetDocument = parameters.get("targetDocument"); |
---|
152 | String targetDOMPath = parameters.get("targetDOMPath"); |
---|
153 | |
---|
154 | if ((targetDocument == null) || (targetDOMPath == null)) { |
---|
155 | throw new SAXException("event has no target defined"); |
---|
156 | } |
---|
157 | |
---|
158 | targetId = determineTargetId(targetDocument, targetDOMPath); |
---|
159 | |
---|
160 | if (targetId == null) { |
---|
161 | // the target id can not be determined yet |
---|
162 | return false; |
---|
163 | } |
---|
164 | } |
---|
165 | |
---|
166 | IGUIElement target = super.getGUIElementTree().find(targetId); |
---|
167 | |
---|
168 | if (target == null) { |
---|
169 | // event not processable yet |
---|
170 | return false; |
---|
171 | } |
---|
172 | |
---|
173 | IEventType eventType = |
---|
174 | HTMLEventTypeFactory.getInstance().getEventType(type, parameters, target); |
---|
175 | |
---|
176 | Event event = new Event(eventType, target); |
---|
177 | |
---|
178 | String timestampStr = parameters.get("timestamp"); |
---|
179 | |
---|
180 | if (timestampStr != null) { |
---|
181 | event.setTimestamp(Long.parseLong(timestampStr)); |
---|
182 | } |
---|
183 | |
---|
184 | ((HTMLGUIElement) event.getTarget()).markUsed(); |
---|
185 | |
---|
186 | super.addToSequence(event); |
---|
187 | |
---|
188 | return true; |
---|
189 | } |
---|
190 | |
---|
191 | /** |
---|
192 | * <p> |
---|
193 | * TODO: comment |
---|
194 | * </p> |
---|
195 | * |
---|
196 | * @param targetDocument |
---|
197 | * @param targetDOMPath |
---|
198 | * @return |
---|
199 | */ |
---|
200 | private String determineTargetId(String targetDocument, String targetDOMPath) |
---|
201 | throws SAXException |
---|
202 | { |
---|
203 | IGUIElement document = super.getGUIElementTree().find(targetDocument); |
---|
204 | |
---|
205 | if (document == null) { |
---|
206 | return null; |
---|
207 | } |
---|
208 | |
---|
209 | if (!(document.getSpecification() instanceof HTMLDocumentSpec)) { |
---|
210 | throw new SAXException("an id that should refer to an HTML document refers to" + |
---|
211 | "something else"); |
---|
212 | } |
---|
213 | |
---|
214 | GUIModel model = super.getGUIElementTree().getGUIModel(); |
---|
215 | IGUIElement child = document; |
---|
216 | String[] pathElements = targetDOMPath.split("/"); |
---|
217 | int pathIndex = 0; |
---|
218 | |
---|
219 | HTMLPageElementSpec compareSpec; |
---|
220 | String tagName; |
---|
221 | int index; |
---|
222 | String htmlId; |
---|
223 | |
---|
224 | while ((pathIndex < pathElements.length) && (child != null)) { |
---|
225 | if ((pathElements[pathIndex] != null) && (!"".equals(pathElements[pathIndex]))) { |
---|
226 | Matcher matcher = htmlElementPattern.matcher(pathElements[pathIndex]); |
---|
227 | if (!matcher.matches()) { |
---|
228 | throw new SAXException |
---|
229 | ("could not parse target DOM path element " + pathElements[pathIndex]); |
---|
230 | } |
---|
231 | |
---|
232 | tagName = matcher.group(1); |
---|
233 | String indexStr = matcher.group(3); |
---|
234 | htmlId = matcher.group(4); |
---|
235 | |
---|
236 | index = -1; |
---|
237 | if ((indexStr != null) && (!"".equals(indexStr))) { |
---|
238 | index = Integer.parseInt(indexStr); |
---|
239 | } |
---|
240 | |
---|
241 | compareSpec = new HTMLPageElementSpec |
---|
242 | ((HTMLDocumentSpec) document.getSpecification(), tagName, htmlId, index); |
---|
243 | |
---|
244 | List<IGUIElement> children = model.getChildren(child); |
---|
245 | child = null; |
---|
246 | |
---|
247 | for (IGUIElement candidate : children) { |
---|
248 | if (compareSpec.getSimilarity(candidate.getSpecification())) { |
---|
249 | child = candidate; |
---|
250 | break; |
---|
251 | } |
---|
252 | } |
---|
253 | } |
---|
254 | |
---|
255 | pathIndex++; |
---|
256 | } |
---|
257 | |
---|
258 | if (child != null) { |
---|
259 | return super.getGUIElementTree().find(child); |
---|
260 | } |
---|
261 | else { |
---|
262 | return null; |
---|
263 | } |
---|
264 | } |
---|
265 | |
---|
266 | /** |
---|
267 | * <p> |
---|
268 | * checks if tags with the provided name must be handled in the GUI model. As an example, |
---|
269 | * it is not necessary to handle "head" tags and anything included in them. |
---|
270 | * </p> |
---|
271 | * |
---|
272 | * @param tagName |
---|
273 | * @return |
---|
274 | */ |
---|
275 | private boolean tagNameMustBeConsidered(String tagName) { |
---|
276 | return |
---|
277 | !"head".equals(tagName) && !"title".equals(tagName) && !"script".equals(tagName) && |
---|
278 | !"style".equals(tagName) && !"link".equals(tagName) && !"meta".equals(tagName) && |
---|
279 | !"iframe".equals(tagName) && !"input_hidden".equals(tagName) && |
---|
280 | !"option".equals(tagName) && !"tt".equals(tagName) && !"br".equals(tagName) && |
---|
281 | !"colgroup".equals(tagName) && !"col".equals(tagName); |
---|
282 | |
---|
283 | } |
---|
284 | |
---|
285 | } |
---|