mybatisr如何解析CRUD

解析入口buildStatementFromContext(List<XNode> list)。

list对应和各种CRUD节点，可以看出，对应着6个sql语句：

然后将configuration(前面加载的配置信息都存储在这),builderAssistant,list里存储的各个CRUD存入节点出入XMLStatementBuilder的构造函数，初始化。

接下来，我们看看XMLStatementBuilder，主要关注parseStatementNode():

public void parseStatementNode() {

    //这里拿到了CRUD节点的id,比如我的就是findPosts

    String id =context.getStringAttribute("id");

    String databaseId =context.getStringAttribute("databaseId");

     if (!databaseIdMatchesCurrent(id, databaseId, this.requiredDatabaseId)) {

        return;

      }

    //获取fetchSize与timeout.......各个属性

    Integer fetchSize =context.getIntAttribute("fetchSize");

    Integer timeout =context.getIntAttribute("timeout");

    String parameterMap =context.getStringAttribute("parameterMap");

    String parameterType =context.getStringAttribute("parameterType");

    Class parameterTypeClass = resolveClass(parameterType);

    String resultMap =context.getStringAttribute("resultMap");

    String resultType =context.getStringAttribute("resultType");

    String lang =context.getStringAttribute("lang");

    LanguageDriver langDriver = getLanguageDriver(lang);

    Class resultTypeClass = resolveClass(resultType);

    String resultSetType =context.getStringAttribute("resultSetType");

    StatementType statementType = StatementType.valueOf(context.getStringAttribute("statementType",         StatementType.PREPARED.toString()));

  ResultSetType resultSetTypeEnum = resolveResultSetType(resultSetType);

//nodename 如何update,insert,delete,select

  String nodeName =context.getNode().getNodeName();

  SqlCommandType sqlCommandType = SqlCommandType.valueOf(nodeName.toUpperCase(Locale.ENGLISH));

//是否select及对应缓存情况

  boolean isSelect = sqlCommandType == SqlCommandType.SELECT;

  boolean flushCache =context.getBooleanAttribute("flushCache", !isSelect);

  boolean useCache =context.getBooleanAttribute("useCache", isSelect);

  boolean resultOrdered =context.getBooleanAttribute("resultOrdered", false);

  // Include Fragments before parsing

  XMLIncludeTransformer includeParser =new XMLIncludeTransformer(configuration, builderAssistant);

  includeParser.applyIncludes(context.getNode());

  // Parse selectKey after includes and remove them.

  processSelectKeyNodes(id, parameterTypeClass, langDriver);

  // Parse the SQL (pre: and were parsed and removed)

//这里有一个sqlsource的实例，这就是保存sql语句的对象，具体有静态和动态两种类型

  SqlSource sqlSource = langDriver.createSqlSource(configuration, context, parameterTypeClass);

  String resultSets =context.getStringAttribute("resultSets");

  String keyProperty =context.getStringAttribute("keyProperty");

  String keyColumn =context.getStringAttribute("keyColumn");

  KeyGenerator keyGenerator;

  String keyStatementId = id + SelectKeyGenerator.SELECT_KEY_SUFFIX;

  keyStatementId =builderAssistant.applyCurrentNamespace(keyStatementId, true);

  if (configuration.hasKeyGenerator(keyStatementId)) {

keyGenerator =configuration.getKeyGenerator(keyStatementId);

  }else {

keyGenerator =context.getBooleanAttribute("useGeneratedKeys",

        configuration.isUseGeneratedKeys() && SqlCommandType.INSERT.equals(sqlCommandType))

?new Jdbc3KeyGenerator() :new NoKeyGenerator();

  }

builderAssistant.addMappedStatement(id, sqlSource, statementType, sqlCommandType,

      fetchSize, timeout, parameterMap, parameterTypeClass, resultMap, resultTypeClass,

      resultSetTypeEnum, flushCache, useCache, resultOrdered,

      keyGenerator, keyProperty, keyColumn, databaseId, langDriver, resultSets);

}

最后调用builderAssistant.addMappedStatement(),将组装的statement放入configuration。

然后再看看langDriver.createSqlSource(configuration, context, parameterTypeClass)如何具体的sql语句进行解析。

重点关注builder.parseScriptNode()都做了什么。

这里首先会将CRUD节点进行分解成MixedSqlNode,这个对象托管了一个SqlNode的List对象。

进而判断是动态sql还是静态sql，分别初始化DynamicSqlSource或RawSqlSource。。。。。待续。