001/*
002 *  Copyright 2010 Anyware Services
003 *
004 *  Licensed under the Apache License, Version 2.0 (the "License");
005 *  you may not use this file except in compliance with the License.
006 *  You may obtain a copy of the License at
007 *
008 *      http://www.apache.org/licenses/LICENSE-2.0
009 *
010 *  Unless required by applicable law or agreed to in writing, software
011 *  distributed under the License is distributed on an "AS IS" BASIS,
012 *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
013 *  See the License for the specific language governing permissions and
014 *  limitations under the License.
015 */
016package org.ametys.web.robots;
017
018import java.io.IOException;
019
020import org.apache.avalon.framework.parameters.ParameterException;
021import org.apache.cocoon.ProcessingException;
022import org.apache.cocoon.xml.XMLUtils;
023import org.xml.sax.SAXException;
024
025import org.ametys.plugins.repository.AmetysObjectIterable;
026import org.ametys.web.repository.page.Page;
027import org.ametys.web.repository.page.jcr.DefaultPage;
028import org.ametys.web.repository.site.Site;
029import org.ametys.web.repository.sitemap.Sitemap;
030
031/**
032 * Generates robots.txt file
033 */
034public class RobotsGenerator extends AbstractRobotsGenerator
035{
036
037    private static final String DISALLOW_LABEL = "Disallow:";
038
039    private static final String NEW_LINE = System.getProperty("line.separator");
040
041    private static final String SITEMAP_LABEL = "Sitemap:";
042
043    private static final String SPACE_CHAR = " ";
044
045    private static final String TEXT_TAG = "text";
046
047    private static final String USER_AGENT_LABEL = "User-agent:";
048
049    private void _serializePageInformation(Page page, StringBuilder builder)
050    {
051        boolean disallow = page.getValue(DefaultPage.METADATA_ROBOTS_DISALLOW, false);
052
053        if (disallow)
054        {
055            builder.append(DISALLOW_LABEL).append(SPACE_CHAR).append("/" + page.getSitemapName() + "/" + page.getPathInSitemap() + ".html" + NEW_LINE);
056            builder.append(DISALLOW_LABEL).append(SPACE_CHAR).append("/" + page.getSitemapName() + "/" + page.getPathInSitemap() + "/" + NEW_LINE);
057        }
058        else
059        {
060            AmetysObjectIterable< ? extends Page> childrenPages = page.getChildrenPages();
061            for (Page child : childrenPages)
062            {
063                _serializePageInformation(child, builder);
064            }
065        }
066
067    }
068
069    private void _serializeSitemapProperty(Site site, StringBuilder builder)
070    {
071        builder.append(SITEMAP_LABEL).append(" ").append(site.getUrl()).append("/sitemap.xml");
072    }
073
074    @Override
075    public void generate() throws IOException, SAXException, ProcessingException
076    {
077        String siteName;
078        try
079        {
080            siteName = parameters.getParameter("siteName");
081
082            Site site = getSite(siteName);
083
084            StringBuilder builder = new StringBuilder();
085            // User agent information
086            builder.append(USER_AGENT_LABEL).append(" *").append(NEW_LINE);
087            // Pages informations
088            for (Sitemap sitemap : site.getSitemaps())
089            {
090                boolean disallow = sitemap.getValue(DefaultPage.METADATA_ROBOTS_DISALLOW, false);
091                
092                if (disallow)
093                {
094                    builder.append(DISALLOW_LABEL).append(SPACE_CHAR).append("/" + sitemap.getSitemapName() + "/" + NEW_LINE);
095                }
096                else
097                {
098                    for (Page page : sitemap.getChildrenPages())
099                    {
100                        _serializePageInformation(page, builder);
101                    }
102                }
103                
104            }
105            // Sitemap URL
106            _serializeSitemapProperty(site, builder);
107            contentHandler.startDocument();
108            XMLUtils.createElement(contentHandler, TEXT_TAG, builder.toString());
109            contentHandler.endDocument();
110        }
111        catch (ParameterException e)
112        {
113            getLogger().error("Error when getting site name", e);
114        }
115    }
116}