如果您正在使用hibernate 3.2。X使用这个:

log4j.logger.org.hibernate.SQL=trace

而不是这样:

log4j.logger.org.hibernate.SQL=debug 

你可以记录这个:net.sf.hibernate.hql.QueryTranslator

输出的例子:

2013-10-31 14:56:19,029 DEBUG [net.sf.hibernate.hql.QueryTranslator] HQL: select noti.id, noti.idmicrosite, noti.fcaducidad, noti.fpublicacion, noti.tipo, noti.imagen, noti.visible, trad.titulo, trad.subtitulo, trad.laurl, trad.urlnom, trad.fuente, trad.texto  from org.ibit.rol.sac.micromodel.Noticia noti join noti.traducciones trad where index(trad)='ca' and noti.visible='S' and noti.idmicrosite=985 and noti.tipo=3446       
2013-10-31 14:56:19,029 DEBUG [net.sf.hibernate.hql.QueryTranslator] SQL: select noticia0_.NOT_CODI as x0_0_, noticia0_.NOT_MICCOD as x1_0_, noticia0_.NOT_CADUCA as x2_0_, noticia0_.NOT_PUBLIC as x3_0_, noticia0_.NOT_TIPO as x4_0_, noticia0_.NOT_IMAGEN as x5_0_, noticia0_.NOT_VISIB as x6_0_, traduccion1_.NID_TITULO as x7_0_, traduccion1_.NID_SUBTIT as x8_0_, traduccion1_.NID_URL as x9_0_, traduccion1_.NID_URLNOM as x10_0_, traduccion1_.NID_FUENTE as x11_0_, traduccion1_.NID_TEXTO as x12_0_ from GUS_NOTICS noticia0_ inner join GUS_NOTIDI traduccion1_ on noticia0_.NOT_CODI=traduccion1_.NID_NOTCOD where (traduccion1_.NID_CODIDI='ca' )and(noticia0_.NOT_VISIB='S' )and(noticia0_.NOT_MICCOD=985 )and(noticia0_.NOT_TIPO=3446 )

您可以在log4j.xml中添加类别行:

<category name="org.hibernate.type">
    <priority value="TRACE"/>
</category>

并添加hibernate属性:

<property name="show_sql">true</property>
<property name="format_sql">true</property>
<property name="use_sql_comments">true</property>

Log4Jdbc插件最适合您的需求。它显示如下-

1. Complete SQL query being hit to the db
2. Parameter values being passed to the query
3. Execution time taken by each query

参考下面的链接来配置Log4Jdbc-

https://code.google.com/p/log4jdbc/

为了方便起见，下面是Logback (SLF4J)的相同配置示例

<appender name="SQLROLLINGFILE">
 <File>/tmp/sql.log</File>
 <rollingPolicy>
  <FileNamePattern>logFile.%d{yyyy-MM-dd}.log</FileNamePattern>
 </rollingPolicy>
 <layout>
  <Pattern>%-4date | %msg %n</Pattern>
 </layout>
</appender>

<logger name="org.hibernate.SQL" additivity="false" >   
 <level value="DEBUG" />    
 <appender-ref ref="SQLROLLINGFILE" />
</logger>

<logger name="org.hibernate.type" additivity="false" >
 <level value="TRACE" />
 <appender-ref ref="SQLROLLINGFILE" />
</logger>

sql.log(示例)中的输出如下所示:

2013-08-30 18:01:15,083 | update stepprovider set created_at=?, lastupdated_at=?, version=?, bundlelocation=?, category_id=?, customer_id=?, description=?, icon_file_id=?, name=?, shareStatus=?, spversion=?, status=?, title=?, type=?, num_used=? where id=?
2013-08-30 18:01:15,084 | binding parameter [1] as [TIMESTAMP] - 2012-07-11 09:57:32.0
2013-08-30 18:01:15,085 | binding parameter [2] as [TIMESTAMP] - Fri Aug 30 18:01:15 CEST 2013
2013-08-30 18:01:15,086 | binding parameter [3] as [INTEGER] -
2013-08-30 18:01:15,086 | binding parameter [4] as [VARCHAR] - com.mypackage.foo
2013-08-30 18:01:15,087 | binding parameter [5] as [VARCHAR] -
2013-08-30 18:01:15,087 | binding parameter [6] as [VARCHAR] -
2013-08-30 18:01:15,087 | binding parameter [7] as [VARCHAR] - TODO
2013-08-30 18:01:15,087 | binding parameter [8] as [VARCHAR] -
2013-08-30 18:01:15,088 | binding parameter [9] as [VARCHAR] - MatchingStep@com.mypackage.foo
2013-08-30 18:01:15,088 | binding parameter [10] as [VARCHAR] - PRIVATE
2013-08-30 18:01:15,088 | binding parameter [11] as [VARCHAR] - 1.0
2013-08-30 18:01:15,088 | binding parameter [12] as [VARCHAR] - 32
2013-08-30 18:01:15,088 | binding parameter [13] as [VARCHAR] - MatchingStep
2013-08-30 18:01:15,089 | binding parameter [14] as [VARCHAR] -
2013-08-30 18:01:15,089 | binding parameter [15] as [INTEGER] - 0
2013-08-30 18:01:15,089 | binding parameter [16] as [VARCHAR] - 053c2e65-5d51-4c09-85f3-2281a1024f64

对我来说，最简单的解决方案是实现一个常规的字符串替换，将参数输入替换为参数值(为了简单起见，将所有参数都视为字符串):

String debuggedSql = sql;
// then, for each named parameter
debuggedSql = debuggedSql.replaceAll(":"+key, "'"+value.toString()+"'");
// and finally
System.out.println(debuggedSql);

或者类似的位置参数(?)。

如果您想要记录运行就绪的SQL，请注意空值和特定的值类型(如date)。