全文检索引擎Solr系列——整合MySQL、MongoDB

MySQL

1. 拷贝mysql-connector-java-5.1.25-bin.jar到E:\solr-4.8.0\example\solr-webapp\webapp\WEB-INF\lib目录下面
2. 配置E:\solr-4.8.0\example\solr\collection1\conf\solrconfig.xml

< requestHandler name = "/dataimport"
     class = "org.apache.solr.handler.dataimport.DataImportHandler" >
       < lst name = "defaults" >
          < str name = "config" >data-config.xml</ str >
       </ lst >
</ requestHandler >

1. 导入依赖库文件：

<lib dir="../../../dist/" regex="solr-dataimporthandler-\d.*\.jar"/>

加在

<lib dir="../../../dist/" regex="solr-cell-\d.*\.jar" />

前面。

1. 创建E:\solr-4.8.0\example\solr\collection1\conf\data-config.xml，指定MySQL数据库地址，用户名、密码以及建立索引的数据表

   <? xml version = "1.0" encoding = "UTF-8" ?>
         < dataConfig >
                 < dataSource type = "JdbcDataSource"
                             driver = "com.mysql.jdbc.Driver"
                             url = "jdbc:mysql://localhost:3306/django_blog"
                             user = "root"
                             password = "" />
                     < document name = "blog" >
                             < entity name = "blog_blog" pk = "id"
                                     query = "select id,title,content from blog_blog"
                                     deltaImportQuery = "select id,title,content from blog_blog where ID='${dataimporter.delta.id}'"
                                     deltaQuery="select id  from blog_blog where add_time > '${dataimporter.last_index_time}'"
                                     deletedPkQuery="select id  from blog_blog where id=0">
                                  < field column = "id" name = "id" />
                                  < field column = "title" name = "title" />
                                  < field column = "content" name = "content" />
                             </ entity >
                    </ document >
         </ dataConfig >

   • query 用于初次导入到索引的sql语句。
     • 考虑到数据表中的数据量非常大，比如千万级，不可能一次索引完，因此需要分批次完成，那么查询语句query要设置两个参数：${dataimporter.request.length} ${dataimporter.request.offset}
     • query=”select id,title,content from blog_blog limit ${dataimporter.request.length} offset ${dataimporter.request.offset}”
     • 请求：http://localhost:8983/solr/collection2/dataimport?command=full-import&commit=true&clean=false&offset=0&length=10000
   • deltaImportQuery 根据ID取得需要进入的索引的单条数据。
   • deltaQuery 用于增量索引的sql语句，用于取得需要增量索引的ID。
   • deletedPkQuery 用于取出需要从索引中删除文档的的ID
2. 为数据库表字段建立域（field），编辑E:\solr-4.8.0\example\solr\collection1\conf\schema.xml:

<!-- mysql -->
   < field name = "id" type = "string" indexed = "true" stored = "true" required = "true" />
   < field name = "title" type = "text_cn" indexed = "true" stored = "true" termVectors = "true" termPositions = "true" termOffsets = "true" />
   < field name = "content" type = "text_cn" indexed = "true" stored = "true" termVectors = "true" termPositions = "true" termOffsets = "true" />
<!-- mysql -->

1. 配置增量索引更新文件

参考：

• http://josh-persistence.iteye.com/blog/2017155
• http://wiki.apache.org/solr/DataImportHandler#Using_delta-import_command

Mongodb

1. 安装mongo-connector，最好使用手动安装方式：

   <code>git clone https://github.com/10gen-labs/mongo-connector.git cd mongo-connector #安装前修改mongo_connector/constants.py的变量：设置DEFAULT_COMMIT_INTERVAL = 0 python setup.py install </code>

   默认是不会自动提交了，这里设置成自动提交，否则mongodb数据库更新，索引这边没法同时更新，或者在命令行中可以指定是否自动提交，不过我现在还没发现。

2. 配置schema.xml，把mongodb中需要加上索引的字段配置到schema.xml文件中：

   <? xml version = "1.0" encoding = "UTF-8" ?>
    < schema name = "example" version = "1.5" >
        < field name = "_version_" type = "long" indexed = "true" stored = "true" />
        < field name = "_id" type = "string" indexed = "true" stored = "true" required = "true" multiValued = "false" />
        < field name = "body" type = "string" indexed = "true" stored = "true" />
        < field name = "title" type = "string" indexed = "true" stored = "true" multiValued = "true" />
        < field name = "text" type = "text_general" indexed = "true" stored = "false" multiValued = "true" />
        < uniqueKey >_id</ uniqueKey >
        < defaultSearchField >title</ defaultSearchField >
        < solrQueryParser defaultOperator = "OR" />
        < fieldType name = "string" class = "solr.StrField" sortMissingLast = "true" />
        < fieldType name = "long" class = "solr.TrieLongField" precisionStep = "0" positionIncrementGap = "0" />
        < fieldType name = "text_general" class = "solr.TextField" positionIncrementGap = "100" >
          < analyzer type = "index" >
            < tokenizer class = "solr.StandardTokenizerFactory" />
            < filter class = "solr.StopFilterFactory" ignoreCase = "true" words = "stopwords.txt" />
            < filter class = "solr.LowerCaseFilterFactory" />
          </ analyzer >
          < analyzer type = "query" >
            < tokenizer class = "solr.StandardTokenizerFactory" />
            < filter class = "solr.StopFilterFactory" ignoreCase = "true" words = "stopwords.txt" />
            < filter class = "solr.SynonymFilterFactory" synonyms = "synonyms.txt" ignoreCase = "true" expand = "true" />
            < filter class = "solr.LowerCaseFilterFactory" />
          </ analyzer >
        </ fieldType >
    </ schema >

3. 启动Mongod：

   <code>mongod --replSet myDevReplSet --smallfiles </code>

   初始化:rs.initiate()

4. 启动mongo-connector:

   <code>E:\Users\liuzhijun\workspace\mongo-connector\mongo_connector\doc_managers&gt;mongo-connector -m localhost:27017 -t http://localhost:8983/solr/collection2 -n s_soccer.person -u id -d ./solr_doc_manager.py </code>

   • -m：mongod服务
   • -t：solr服务
   • -n：mongodb命名空间，监听database.collection，多个命名空间逗号分隔
   • -u：uniquekey
   • -d：处理文档的manager文件

   注意：mongodb通常使用_id作为uniquekey，而Solrmore使用id作为uniquekey，如果不做处理，索引文件时将会失败，有两种方式来处理这个问题：

   1. 指定参数--unique-key=id到mongo-connector，Mongo Connector 就可以翻译把_id转换到id。
   2. 把schema.xml文件中的:

      <code>&lt;uniqueKey&gt;id&lt;uniqueKey&gt; </code>

      替换成

      <code>&lt;uniqueKey&gt;_id&lt;/uniqueKey&gt; </code>

      同时还要定义一个_id的字段：

      <code>&lt;field name="_id" type="string" indexed="true" stored="true" /&gt; </code>

   3. 启动时如果报错：

      <code>2014-06-18 12:30:36,648 - ERROR - OplogThread: Last entry no longer in oplog cannot recover! Collection(Database(MongoClient('localhost', 27017), u'local'), u'oplog.rs') </code>

      清空E:\Users\liuzhijun\workspace\mongo-connector\mongo_connector\doc_managers\config.txt中的内容，需要删除索引目录下的文件重新启动

5. 测试
   mongodb中的数据变化都会同步到solr中去。