@Intercepts(@Signature(type = Executor.class, method = "query", args = {MappedStatement.class, Object.class, RowBounds.class, ResultHandler.class}))
public class PageHelper implements Interceptor {
//sql工具类
private SqlUtil sqlUtil;
//属性参数信息
private Properties properties;
//配置对象方式
private SqlUtilConfig sqlUtilConfig;
//自动获取dialect,如果没有setProperties或setSqlUtilConfig,也可以正常进行
private boolean autoDialect = true;
//运行时自动获取dialect
private boolean autoRuntimeDialect;
//多数据源时,获取jdbcurl后是否关闭数据源
private boolean closeConn = true;
//缓存
private Map<String, SqlUtil> urlSqlUtilMap = new ConcurrentHashMap<String, SqlUtil>();
private ReentrantLock lock = new ReentrantLock();
// ...
}
SqlUtil
:数据库类型专用sql工具类,一个数据库url对应一个SqlUtil实例,SqlUtil内有一个Parser对象,如果是mysql,它是MysqlParser,如果是oracle,它是OracleParser,这个Parser对象是SqlUtil不同实例的主要存在价值。执行count查询、设置Parser对象、执行分页查询、保存Page分页对象等功能,均由SqlUtil来完成。SqlUtilConfig
:Spring Boot中使用。autoRuntimeDialect
:多个数据源切换时,比如mysql和oracle数据源同时存在,就不能简单指定dialect,这个时候就需要运行时自动检测当前的dialect。Map<String, SqlUtil> urlSqlUtilMap
:它就用来缓存autoRuntimeDialect自动检测结果的,key是数据库的url,value是SqlUtil。由于这种自动检测只需要执行1次,所以做了缓存。ReentrantLock lock
:这个lock对象是比较有意思的现象,urlSqlUtilMap明明是一个同步ConcurrentHashMap,又搞了一个lock出来同步ConcurrentHashMap做什么呢?简单的说,ConcurrentHashMap可以保证put或者remove方法一定是线程安全的,但它不能保证put、get、remove的组合操作是线程安全的,为了保证组合操作也是线程安全的,所以使用了lock。// Mybatis拦截器方法
public Object intercept(Invocation invocation) throws Throwable {
if (autoRuntimeDialect) {
// 多数据源
SqlUtil sqlUtil = getSqlUtil(invocation);
return sqlUtil.processPage(invocation);
} else {
// 单数据源
if (autoDialect) {
initSqlUtil(invocation);
}
// 指定了dialect
return sqlUtil.processPage(invocation);
}
}
public synchronized void initSqlUtil(Invocation invocation) {
if (this.sqlUtil == null) {
this.sqlUtil = getSqlUtil(invocation);
if (!autoRuntimeDialect) {
properties = null;
sqlUtilConfig = null;
}
autoDialect = false;
}
}
public void setProperties(Properties p) {
checkVersion();
//多数据源时,获取jdbcurl后是否关闭数据源
String closeConn = p.getProperty("closeConn");
//解决#97
if(StringUtil.isNotEmpty(closeConn)){
this.closeConn = Boolean.parseBoolean(closeConn);
}
//初始化SqlUtil的PARAMS
SqlUtil.setParams(p.getProperty("params"));
//数据库方言
String dialect = p.getProperty("dialect");
String runtimeDialect = p.getProperty("autoRuntimeDialect");
if (StringUtil.isNotEmpty(runtimeDialect) && runtimeDialect.equalsIgnoreCase("TRUE")) {
this.autoRuntimeDialect = true;
this.autoDialect = false;
this.properties = p;
} else if (StringUtil.isEmpty(dialect)) {
autoDialect = true;
this.properties = p;
} else {
autoDialect = false;
sqlUtil = new SqlUtil(dialect);
sqlUtil.setProperties(p);
}
}
public SqlUtil getSqlUtil(Invocation invocation) {
MappedStatement ms = (MappedStatement) invocation.getArgs()[0];
//改为对dataSource做缓存
DataSource dataSource = ms.getConfiguration().getEnvironment().getDataSource();
String url = getUrl(dataSource);
if (urlSqlUtilMap.containsKey(url)) {
return urlSqlUtilMap.get(url);
}
try {
lock.lock();
if (urlSqlUtilMap.containsKey(url)) {
return urlSqlUtilMap.get(url);
}
if (StringUtil.isEmpty(url)) {
throw new RuntimeException("无法自动获取jdbcUrl,请在分页插件中配置dialect参数!");
}
String dialect = Dialect.fromJdbcUrl(url);
if (dialect == null) {
throw new RuntimeException("无法自动获取数据库类型,请通过dialect参数指定!");
}
SqlUtil sqlUtil = new SqlUtil(dialect);
if (this.properties != null) {
sqlUtil.setProperties(properties);
} else if (this.sqlUtilConfig != null) {
sqlUtil.setSqlUtilConfig(this.sqlUtilConfig);
}
urlSqlUtilMap.put(url, sqlUtil);
return sqlUtil;
} finally {
lock.unlock();
}
}
public class PageStaticSqlSource extends PageSqlSource {
private String sql;
private List<ParameterMapping> parameterMappings;
private Configuration configuration;
private SqlSource original;
@Override
protected BoundSql getDefaultBoundSql(Object parameterObject) {
String tempSql = sql;
String orderBy = PageHelper.getOrderBy();
if (orderBy != null) {
tempSql = OrderByParser.converToOrderBySql(sql, orderBy);
}
return new BoundSql(configuration, tempSql, parameterMappings, parameterObject);
}
@Override
protected BoundSql getCountBoundSql(Object parameterObject) {
// localParser指的就是MysqlParser或者OracleParser
// localParser.get().getCountSql(sql),可以根据原始的sql,生成一个count查询的sql
return new BoundSql(configuration, localParser.get().getCountSql(sql), parameterMappings, parameterObject);
}
@Override
protected BoundSql getPageBoundSql(Object parameterObject) {
String tempSql = sql;
String orderBy = PageHelper.getOrderBy();
if (orderBy != null) {
tempSql = OrderByParser.converToOrderBySql(sql, orderBy);
}
// getPageSql可以根据原始的sql,生成一个带有分页参数信息的sql,比如 limit ?, ?
tempSql = localParser.get().getPageSql(tempSql);
// 由于sql增加了分页参数的?号占位符,getPageParameterMapping()就是在原有List<ParameterMapping>基础上,增加两个分页参数对应的ParameterMapping对象,为分页参数赋值使用
return new BoundSql(configuration, tempSql, localParser.get().getPageParameterMapping(configuration, original.getBoundSql(parameterObject)), parameterObject);
}
}
```java
public abstract class AbstractParser implements Parser, Constant { public String getCountSql(final String sql) { return sqlParser.getSmartCountSql(sql); } }
public class MysqlParser extends AbstractParser { @Override public String getPageSql(String sql) { StringBuilder sqlBuilder = new StringBuilder(sql.length() + 14); sqlBuilder.append(sql); sqlBuilder.append(“ limit ?,?”); return sqlBuilder.toString(); }
@Override
public Map<String, Object> setPageParameter(MappedStatement ms, Object parameterObject, BoundSql boundSql, Page<?> page) {
Map<String, Object> paramMap = super.setPageParameter(ms, parameterObject, boundSql, page);
paramMap.put(PAGEPARAMETER_FIRST, page.getStartRow());
paramMap.put(PAGEPARAMETER_SECOND, page.getPageSize());
return paramMap;
}
}
```java
// PageSqlSource装饰原SqlSource
public void processMappedStatement(MappedStatement ms) throws Throwable {
SqlSource sqlSource = ms.getSqlSource();
MetaObject msObject = SystemMetaObject.forObject(ms);
SqlSource pageSqlSource;
if (sqlSource instanceof StaticSqlSource) {
pageSqlSource = new PageStaticSqlSource((StaticSqlSource) sqlSource);
} else if (sqlSource instanceof RawSqlSource) {
pageSqlSource = new PageRawSqlSource((RawSqlSource) sqlSource);
} else if (sqlSource instanceof ProviderSqlSource) {
pageSqlSource = new PageProviderSqlSource((ProviderSqlSource) sqlSource);
} else if (sqlSource instanceof DynamicSqlSource) {
pageSqlSource = new PageDynamicSqlSource((DynamicSqlSource) sqlSource);
} else {
throw new RuntimeException("无法处理该类型[" + sqlSource.getClass() + "]的SqlSource");
}
msObject.setValue("sqlSource", pageSqlSource);
//由于count查询需要修改返回值,因此这里要创建一个Count查询的MS
msCountMap.put(ms.getId(), MSUtils.newCountMappedStatement(ms));
}
// 执行分页查询
private Page doProcessPage(Invocation invocation, Page page, Object[] args) throws Throwable {
//保存RowBounds状态
RowBounds rowBounds = (RowBounds) args[2];
//获取原始的ms
MappedStatement ms = (MappedStatement) args[0];
//判断并处理为PageSqlSource
if (!isPageSqlSource(ms)) {
processMappedStatement(ms);
}
//设置当前的parser,后面每次使用前都会set,ThreadLocal的值不会产生不良影响
((PageSqlSource)ms.getSqlSource()).setParser(parser);
try {
//忽略RowBounds-否则会进行Mybatis自带的内存分页
args[2] = RowBounds.DEFAULT;
//如果只进行排序 或 pageSizeZero的判断
if (isQueryOnly(page)) {
return doQueryOnly(page, invocation);
}
//简单的通过total的值来判断是否进行count查询
if (page.isCount()) {
page.setCountSignal(Boolean.TRUE);
//替换MS
args[0] = msCountMap.get(ms.getId());
//查询总数
Object result = invocation.proceed();
//还原ms
args[0] = ms;
//设置总数
page.setTotal((Integer) ((List) result).get(0));
if (page.getTotal() == 0) {
return page;
}
} else {
page.setTotal(-1l);
}
//pageSize>0的时候执行分页查询,pageSize<=0的时候不执行相当于可能只返回了一个count
if (page.getPageSize() > 0 &&
((rowBounds == RowBounds.DEFAULT && page.getPageNum() > 0)
|| rowBounds != RowBounds.DEFAULT)) {
//将参数中的MappedStatement替换为新的qs
page.setCountSignal(null);
BoundSql boundSql = ms.getBoundSql(args[1]);
args[1] = parser.setPageParameter(ms, args[1], boundSql, page);
page.setCountSignal(Boolean.FALSE);
//执行分页查询
Object result = invocation.proceed();
//得到处理结果
page.addAll((List) result);
}
} finally {
((PageSqlSource)ms.getSqlSource()).removeParser();
}
//返回结果
return page;