ibatis常用16條SQL語句
博客分類:
• ibatis專欄
(1) 輸入參數為單個值
Xml代碼
1. <delete id="com.fashionfree.stat.accesslog.deleteMemberAccessLogsBefore"
2. parameterClass="long">
3. delete from
4. MemberAccessLog
5. where
6. accessTimestamp = #value#
7. </delete>
(2) 輸入參數為一個對象
Xml代碼
1. <insert id="com.fashionfree.stat.accesslog.MemberAccessLog.insert"
2. parameterClass="com.fashionfree.stat.accesslog.model.MemberAccessLog>
3. insert into MemberAccessLog
4. (
5. accessLogId, memberId, clientIP,
6. httpMethod, actionId, requestURL,
7. accessTimestamp, extend1, extend2,
8. extend3
9. )
10. values
11. (
12. #accessLogId#, #memberId#,
13. #clientIP#, #httpMethod#,
14. #actionId#, #requestURL#,
15. #accessTimestamp#, #extend1#,
16. #extend2#, #extend3#
17. )
18. </insert>
(3) 輸入參數為一個java.util.HashMap
Xml代碼
1. <select id="com.fashionfree.stat.accesslog.selectActionIdAndActionNumber"
2. parameterClass="hashMap"
3. resultMap="getActionIdAndActionNumber">
4. select
5. actionId, count(*) as count
6. from
7. MemberAccessLog
8. where
9. memberId = #memberId#
10. and accessTimestamp > #start#
11. and accessTimestamp <= #end#
12. group by actionId
13. </select>
(4) 輸入參數中含有數組
Xml代碼
1. <insert id="updateStatusBatch" parameterClass="hashMap">
2. update
3. Question
4. set
5. status = #status#
6. <dynamic prepend="where questionId in">
7. <isNotNull property="actionIds">
8. <iterate property="actionIds" open="(" close=")" conjunction=",">
9. #actionIds[]#
10. </iterate>
11. </isNotNull>
12. </dynamic>
13. </insert>
說明:actionIds為傳入的數組的名字;
使用dynamic標簽避免數組為空時導致sql語句語法出錯;
使用isNotNull標簽避免數組為null時ibatis解析出錯
(5)傳遞參數只含有一個數組
Xml代碼
1. <select id="com.fashionfree.stat.accesslog.model.StatMemberAction.selectActionIdsOfModule"
2. resultClass="hashMap">
3. select
4. moduleId, actionId
5. from
6. StatMemberAction
7. <dynamic prepend="where moduleId in">
8. <iterate open="(" close=")" conjunction=",">
9. #[]#
10. </iterate>
11. </dynamic>
12. order by
13. moduleId
14. </select>
說明:注意select的標簽中沒有parameterClass一項
另:這里也可以把數組放進一個hashMap中,但增加額外開銷,不建議使用
(6)讓ibatis把參數直接解析成字符串
Xml代碼
1. <select id="com.fashionfree.stat.accesslog.selectSumDistinctCountOfAccessMemberNum"
2. parameterClass="hashMap" resultClass="int">
3. select
4. count(distinct memberId)
5. from
6. MemberAccessLog
7. where
8. accessTimestamp >= #start#
9. and accessTimestamp < #end#
10. and actionId in $actionIdString$
11. </select>
說明:使用這種方法存在sql注入的風險,不推薦使用
(7)分頁查詢 (pagedQuery)
Java代碼
1. <select id="com.fashionfree.stat.accesslog.selectMemberAccessLogBy"
2. parameterClass="hashMap" resultMap="MemberAccessLogMap">
3. <include refid="selectAllSql"/>
4. <include refid="whereSql"/>
5. <include refid="pageSql"/>
6. </select>
7. <select id="com.fashionfree.stat.accesslog.selectMemberAccessLogBy.Count"
8. parameterClass="hashMap" resultClass="int">
9. <include refid="countSql"/>
10. <include refid="whereSql"/>
11. </select>
12. <sql id="selectAllSql">
13. select
14. accessLogId, memberId, clientIP,
15. httpMethod, actionId, requestURL,
16. accessTimestamp, extend1, extend2,
17. extend3
18. from
19. MemberAccessLog
20. </sql>
21. <sql id="whereSql">
22. accessTimestamp <= #accessTimestamp#
23. </sql>
24. <sql id="countSql">
25. select
26. count(*)
27. from
28. MemberAccessLog
29. </sql>
30. <sql id="pageSql">
31. <dynamic>
32. <isNotNull property="startIndex">
33. <isNotNull property="pageSize">
34. limit #startIndex# , #pageSize#
35. </isNotNull>
36. </isNotNull>
37. </dynamic>
38. </sql>
說明:本例中,代碼應為:
HashMap hashMap = new HashMap();
hashMap.put(“accessTimestamp”, someValue);
pagedQuery(“com.fashionfree.stat.accesslog.selectMemberAccessLogBy”, hashMap);
pagedQuery方法首先去查找名為com.fashionfree.stat.accesslog.selectMemberAccessLogBy.Count 的mapped statement來進行sql查詢,從而得到com.fashionfree.stat.accesslog.selectMemberAccessLogBy查詢的記錄個數,
再進行所需的paged sql查詢(com.fashionfree.stat.accesslog.selectMemberAccessLogBy),具體過程參見utils類中的相關代碼
(8)sql語句中含有大於號>、小於號<
1. 將大於號、小於號寫為: > < 如:
Xml代碼
1. <delete id="com.fashionfree.stat.accesslog.deleteMemberAccessLogsBefore" parameterClass="long">
2. delete from
3. MemberAccessLog
4. where
5. accessTimestamp <= #value#
6. </delete>
2. 將特殊字符放在xml的CDATA區內:
Xml代碼
1. <delete id="com.fashionfree.stat.accesslog.deleteMemberAccessLogsBefore" parameterClass="long">
2. <![CDATA[
3. delete from
4. MemberAccessLog
5. where
6. accessTimestamp <= #value#
7. ]]>
8. </delete>
推薦使用第一種方式,寫為< 和 > (XML不對CDATA里的內容進行解析,因此如果CDATA中含有dynamic標簽,將不起作用)
(9)include和sql標簽
將常用的sql語句整理在一起,便於共用:
Xml代碼
1. <sql id="selectBasicSql">
2. select
3. samplingTimestamp,onlineNum,year,
4. month,week,day,hour
5. from
6. OnlineMemberNum
7. </sql>
8. <sql id="whereSqlBefore">
9. where samplingTimestamp <= #samplingTimestamp#
10. </sql>
11. <select id="com.fashionfree.accesslog.selectOnlineMemberNumsBeforeSamplingTimestamp" parameterClass="hashmap" resultClass="OnlineMemberNum">
12. <include refid="selectBasicSql" />
13. <include refid="whereSqlBefore" />
14. </select>
注意:sql標簽只能用於被引用,不能當作mapped statement。如上例中有名為selectBasicSql的sql元素,試圖使用其作為sql語句執行是錯誤的:
sqlMapClient.queryForList(“selectBasicSql”); ×
(10)隨機選取記錄
Xml代碼
1. <sql id=”randomSql”>
2. ORDER BY rand() LIMIT #number#
3. </sql>
從數據庫中隨機選取number條記錄(只適用於MySQL)
(11)將SQL GROUP BY分組中的字段拼接
Xml代碼
1. <sql id=”selectGroupBy>
2. SELECT
3. a.answererCategoryId, a.answererId, a.answererName,
4. a.questionCategoryId, a.score, a.answeredNum,
5. a.correctNum, a.answerSeconds, a.createdTimestamp,
6. a.lastQuestionApprovedTimestamp, a.lastModified, GROUP_CONCAT(q.categoryName) as categoryName
7. FROM
8. AnswererCategory a, QuestionCategory q
9. WHERE a.questionCategoryId = q.questionCategoryId
10. GROUP BY a.answererId
11. ORDER BY a.answererCategoryId
12. </sql>
注:SQL中使用了MySQL的GROUP_CONCAT函數
(12) 按照IN里面的順序進行排序
①MySQL:
Xml代碼
1. <sql id=”groupByInArea”>
2. select
3. moduleId, moduleName,
4. status, lastModifierId, lastModifiedName,
5. lastModified
6. from
7. StatModule
8. where
9. moduleId in (3, 5, 1)
10. order by
11. instr(',3,5,1,' , ','+ltrim(moduleId)+',')
12. </sql>
②SQLSERVER:
Xml代碼
1. <sql id=”groupByInArea”>
2. select
3. moduleId, moduleName,
4. status, lastModifierId, lastModifiedName,
5. lastModified
6. from
7. StatModule
8. where
9. moduleId in (3, 5, 1)
10. order by
11. charindex(','+ltrim(moduleId)+',' , ',3,5,1,')
12. </sql>
說明:查詢結果將按照moduleId在in列表中的順序(3, 5, 1)來返回
MySQL : instr(str, substr)
SQLSERVER: charindex(substr, str)
返回字符串str 中子字符串的第一個出現位置
ltrim(str)
返回字符串str, 其引導(左面的)空格字符被刪除
(13) resultMap
resultMap負責將SQL查詢結果集的列值映射成Java Bean的屬性值。
Xml代碼
1. <resultMap class="java.util.HashMap" id="getActionIdAndActionNumber">
2. <result column="actionId" property="actionId" jdbcType="BIGINT" javaType="long"/>
3. <result column="count" property="count" jdbcType="INT" javaType="int"/>
4. </resultMap>
使用resultMap稱為顯式結果映射,與之對應的是resultClass(內聯結果映射),使用resultClass的最大好處便是簡單、方便,不需顯示指定結果,由iBATIS根據反射來確定自行決定。而resultMap則可以通過指定jdbcType和javaType,提供更嚴格的配置認證。
(14) typeAlias
Xml代碼
1. <typeAlias alias="MemberOnlineDuration" type="com.fashionfree.stat.accesslog.model.MemberOnlineDuration" />
2. <typeAlias>允許你定義別名,避免重復輸入過長的名字。
(15) remap
Xml代碼
1. <select id="testForRemap" parameterClass="hashMap" resultClass="hashMap" remapResults="true">
2. select
3. userId
4. <isEqual property="tag" compareValue="1">
5. , userName
6. </isEqual>
7. <isEqual property="tag" compareValue="2">
8. , userPassword
9. </isEqual>
10. from
11. UserInfo
12. </select>
此例中,根據參數tag值的不同,會獲得不同的結果集,如果沒有remapResults="true"屬性,iBatis會將第一次查詢時的結果集緩存,下次再執行時(必須還是該進程中)不會再執行結果集映射,而是會使用緩存的結果集。
因此,如果上面的例子中remapResult為默認的false屬性,而有一段程序這樣書寫:
Java代碼
1. HashMap<String, Integer> hashMap = new HashMap<String, Integer>();
2. hashMap.put("tag", 1);
3. sqlClient.queryForList("testForRemap", hashMap);
4. hashMap.put("tag", 2);
5. sqlClient.queryForList("testForRemap", hashMap);
則程序會在執行最后一句的query查詢時報錯,原因就是iBATIS使用了第一次查詢時的結果集,而前后兩次的結果集是不同的:(userId, userName)和(userId, userPassword),所以導致出錯。如果使用了remapResults="true"這一屬性,iBATIS會在每次執行查詢時都執行結果集映射,從而避免錯誤的發生(此時會有較大的開銷)。
(16) dynamic標簽的prepend
dynamic標簽的prepend屬性作為前綴添加到結果內容前面,當標簽的結果內容為空時,prepend屬性將不起作用。
當dynamic標簽中存在prepend屬性時,將會把其嵌套子標簽的第一個prepend屬性忽略。例如:
Xml代碼
1. <sql id="whereSql">
2. <dynamic prepend="where ">
3. <isNotNull property="userId" prepend="BOGUS">
4. userId = #userId#
5. </isNotNull>
6. <isNotEmpty property="userName" prepend="and ">
7. userName = #userName#
8. </isNotEmpty>
9. </dynamic>
10. </sql>
此例中,dynamic標簽中含有兩個子標簽<isNotNull>和<isNotEmpty>。根據前面敘述的原則,如果<isNotNull>標簽中沒有prepend="BOGUS" 這一假的屬性來讓dynamic去掉的話,<isNotEmpty>標簽中的and就會被忽略,會造成sql語法錯誤。
注意:當dynamic標簽沒有prepend屬性時,不會自動忽略其子標簽的第一個prepend屬性。