wamblee.org Git - utils/blob - crawler/kiss/src/org/wamblee/crawler/kiss/ProgramConfigurationParser.java

   1 /*
   2  * Copyright 2005 the original author or authors.
   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  *
   8  *      http://www.apache.org/licenses/LICENSE-2.0
   9  *
  10  * Unless required by applicable law or agreed to in writing, software
  11  * distributed under the License is distributed on an "AS IS" BASIS,
  12  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  13  * See the License for the specific language governing permissions and
  14  * limitations under the License.
  15  */
  16
  17 package org.wamblee.crawler.kiss;
  18
  19 import java.io.InputStream;
  20 import java.util.ArrayList;
  21 import java.util.Iterator;
  22 import java.util.List;
  23
  24 import org.dom4j.Attribute;
  25 import org.dom4j.Document;
  26 import org.dom4j.DocumentException;
  27 import org.dom4j.Element;
  28 import org.dom4j.io.SAXReader;
  29 import org.wamblee.conditions.AndCondition;
  30 import org.wamblee.conditions.Condition;
  31 import org.wamblee.conditions.PropertyRegexCondition;
  32 import org.wamblee.crawler.kiss.ProgramFilter.ProgramAction;
  33
  34 /**
  35  * Parse the configuration of desired programs.
  36  */
  37 class ProgramConfigurationParser {
  38
  39     private static final String ELEM_PROGRAM = "program";
  40
  41     private static final String ELEM_PATTERN = "match";
  42
  43     private static final String ELEM_ACTION = "action";
  44
  45     private static final String ACTION_NOTIFY = "notify";
  46
  47     /**
  48      * Parses the condition used to match the desired programs.
  49      *
  50      * @param aStream
  51      *            Input stream to parse from.
  52      * @return Condition.
  53      */
  54     List<ProgramFilter> parse(InputStream aStream) {
  55         List<ProgramFilter> filters = new ArrayList<ProgramFilter>();
  56         try {
  57             SAXReader reader = new SAXReader();
  58             Document document = reader.read(aStream);
  59
  60             Element root = document.getRootElement();
  61
  62             for (Iterator i = root.elementIterator(ELEM_PROGRAM); i.hasNext();) {
  63                 Element program = (Element) i.next();
  64
  65                 Element actionElem = program.element(ELEM_ACTION);
  66                 ProgramAction action = ProgramAction.RECORD;
  67                 if (actionElem != null) {
  68                     if (actionElem.getText().equals(ACTION_NOTIFY)) {
  69                         action = ProgramAction.NOTIFY;
  70                     }
  71                 }
  72                 List<Condition<Program>> regexConditions =
  73                     new ArrayList<Condition<Program>>();
  74                 for (Iterator j = program.elementIterator(ELEM_PATTERN); j.hasNext(); ) {
  75                     Element patternElem = (Element)j.next();
  76                     String fieldName = "name";
  77                     Attribute fieldAttribute = patternElem.attribute("field");
  78                     if ( fieldAttribute != null ) {
  79                         fieldName = fieldAttribute.getText();
  80                     }
  81                     String pattern = ".*" + patternElem.getText()
  82                     + ".*";
  83                     regexConditions.add(new PropertyRegexCondition<Program>(fieldName, pattern, true));
  84                 }
  85                 Condition<Program> condition = new AndCondition<Program>(regexConditions);
  86                 filters.add(new ProgramFilter(condition, action));
  87             }
  88             return filters;
  89         } catch (DocumentException e) {
  90             throw new RuntimeException("Error parsing program configuraiton", e);
  91         }
  92     }
  93 }