wamblee.org Git - utils/blob - crawler/kiss/src/main/java/org/wamblee/crawler/kiss/main/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.main;
  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.guide.Program;
  33
  34 /**
  35  * Parse the configuration of desired programs.
  36  *
  37  * @author Erik Brakkee
  38  */
  39 class ProgramConfigurationParser {
  40     private static final int DEFAULT_PRIORITY = 1;
  41
  42     // Configuration of interesting programs.
  43
  44     private static final String ELEM_PROGRAM = "program";
  45
  46     private static final String ELEM_PRIORITY = "priority";
  47
  48     private static final String ELEM_PATTERN = "match";
  49
  50     private static final String ELEM_ACTION = "action";
  51
  52     private static final String ELEM_CATEGORY = "category";
  53
  54     private static final String ACTION_NOTIFY = "notify";
  55
  56     private List<ProgramFilter> _filters;
  57
  58     ProgramConfigurationParser() {
  59         _filters = null;
  60     }
  61
  62     /**
  63      * Parses the condition used to match the desired programs.
  64      *
  65      * @param aStream
  66      *            Input stream to parse from.
  67      * @return Condition.
  68      */
  69     void parse(InputStream aStream) {
  70         List<ProgramFilter> filters = new ArrayList<ProgramFilter>();
  71         try {
  72             SAXReader reader = new SAXReader();
  73             Document document = reader.read(aStream);
  74
  75             Element root = document.getRootElement();
  76
  77             for (Iterator i = root.elementIterator(ELEM_PROGRAM); i.hasNext();) {
  78                 Element program = (Element) i.next();
  79
  80                 Element categoryElem = program.element(ELEM_CATEGORY);
  81                 String category = "";
  82                 if (categoryElem != null) {
  83                     category = categoryElem.getText().trim();
  84                 }
  85
  86                 Element actionElem = program.element(ELEM_ACTION);
  87                 int priority = DEFAULT_PRIORITY;
  88                 String priorityString = program.elementTextTrim(ELEM_PRIORITY);
  89                 if ( priorityString != null ) {
  90                     priority = Integer.valueOf(priorityString);
  91                 }
  92                 ProgramAction action = new RecordProgramAction(priority);
  93                 if (actionElem != null) {
  94                     if (actionElem.getText().equals(ACTION_NOTIFY)) {
  95                         action = new InterestingProgramAction(category);
  96                     }
  97                 }
  98
  99                 List<Condition<Program>> regexConditions = new ArrayList<Condition<Program>>();
 100                 for (Iterator j = program.elementIterator(ELEM_PATTERN); j
 101                         .hasNext();) {
 102                     Element patternElem = (Element) j.next();
 103                     String fieldName = "name";
 104                     Attribute fieldAttribute = patternElem.attribute("field");
 105                     if (fieldAttribute != null) {
 106                         fieldName = fieldAttribute.getText();
 107                     }
 108                     String pattern = ".*(" + patternElem.getText() + ").*";
 109                     regexConditions.add(new PropertyRegexCondition<Program>(
 110                             fieldName, pattern, true));
 111                 }
 112                 Condition<Program> condition = new AndCondition<Program>(
 113                         regexConditions);
 114                 filters.add(new ProgramFilter(condition, action));
 115             }
 116             _filters = filters;
 117         } catch (DocumentException e) {
 118             throw new RuntimeException("Error parsing program configuraiton", e);
 119         }
 120     }
 121
 122     /**
 123      * Returns the list of program filters.
 124      *
 125      * @return Filter list.
 126      */
 127     public List<ProgramFilter> getFilters() {
 128         return _filters;
 129     }
 130 }