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
  33 /**
  34  * Parse the configuration of desired programs.
  35  */
  36 class ProgramConfigurationParser {
  37
  38     private static final String ELEM_PROGRAM = "program";
  39
  40     private static final String ELEM_PATTERN = "match";
  41
  42     private static final String ELEM_ACTION = "action";
  43
  44     private static final String ACTION_NOTIFY = "notify";
  45
  46     /**
  47      * Parses the condition used to match the desired programs.
  48      *
  49      * @param aStream
  50      *            Input stream to parse from.
  51      * @return Condition.
  52      */
  53     List<ProgramFilter> parse(InputStream aStream) {
  54         List<ProgramFilter> filters = new ArrayList<ProgramFilter>();
  55         try {
  56             SAXReader reader = new SAXReader();
  57             Document document = reader.read(aStream);
  58
  59             Element root = document.getRootElement();
  60
  61             for (Iterator i = root.elementIterator(ELEM_PROGRAM); i.hasNext();) {
  62                 Element program = (Element) i.next();
  63
  64                 Element actionElem = program.element(ELEM_ACTION);
  65                 ProgramAction action = new RecordProgramAction();
  66                 if (actionElem != null) {
  67                     if (actionElem.getText().equals(ACTION_NOTIFY)) {
  68                         action = new InterestingProgramAction("");
  69                     }
  70                 }
  71                 List<Condition<Program>> regexConditions =
  72                     new ArrayList<Condition<Program>>();
  73                 for (Iterator j = program.elementIterator(ELEM_PATTERN); j.hasNext(); ) {
  74                     Element patternElem = (Element)j.next();
  75                     String fieldName = "name";
  76                     Attribute fieldAttribute = patternElem.attribute("field");
  77                     if ( fieldAttribute != null ) {
  78                         fieldName = fieldAttribute.getText();
  79                     }
  80                     String pattern = ".*(" + patternElem.getText()
  81                     + ").*";
  82                     regexConditions.add(new PropertyRegexCondition<Program>(fieldName, pattern, true));
  83                 }
  84                 Condition<Program> condition = new AndCondition<Program>(regexConditions);
  85                 filters.add(new ProgramFilter(condition, action));
  86             }
  87             return filters;
  88         } catch (DocumentException e) {
  89             throw new RuntimeException("Error parsing program configuraiton", e);
  90         }
  91     }
  92 }