iBatis 簡介：
iBatis 是apache 的一個開源項目，一個O/R Mapping 解決方案，iBatis 最大的特點就是小巧，上手很快。如果不需要太多復雜的功能，iBatis 是能夠滿足你的要求又足夠靈活的最簡單的解決方案，現在的iBatis 已經改名為Mybatis 了。
1.輸入參數為單個值
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
<delete id="com.fashionfree.stat.accesslog.deleteMemberAccessLogsBefore"parameterClass="long"> delete from MemberAccessLog where accessTimestamp = #value# </delete> <delete id="com.fashionfree.stat.accesslog.deleteMemberAccessLogsBefore"parameterClass="long"> delete from MemberAccessLog where accessTimestamp = #value# </delete> |
2.輸入參數為一個對象
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
|
<insert id="com.fashionfree.stat.accesslog.MemberAccessLog.insert"parameterClass="com.fashionfree.stat.accesslog.model.MemberAccessLog> insert into MemberAccessLog ( accessLogId, memberId, clientIP, httpMethod, actionId, requestURL, accessTimestamp, extend1, extend2, extend3 ) values ( #accessLogId#, #memberId#, #clientIP#, #httpMethod#, #actionId#, #requestURL#, #accessTimestamp#, #extend1#, #extend2#, #extend3# ) </insert> <insert id="com.fashionfree.stat.accesslog.MemberAccessLog.insert"parameterClass="com.fashionfree.stat.accesslog.model.MemberAccessLog> insert into MemberAccessLog ( accessLogId, memberId, clientIP, httpMethod, actionId, requestURL, accessTimestamp, extend1, extend2, extend3 ) values ( #accessLogId#, #memberId#, #clientIP#, #httpMethod#, #actionId#, #requestURL#, #accessTimestamp#, #extend1#, #extend2#, #extend3# ) </insert> |
3.輸入參數為一個java.util.HashMap
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
|
<select id="com.fashionfree.stat.accesslog.selectActionIdAndActionNumber"parameterClass="hashMap"resultMap="getActionIdAndActionNumber"> select actionId, count(*) as count from MemberAccessLog where memberId = #memberId# and accessTimestamp > #start# and accessTimestamp <= #end# group by actionId </select><select id="com.fashionfree.stat.accesslog.selectActionIdAndActionNumber"parameterClass="hashMap"resultMap="getActionIdAndActionNumber"> select actionId, count(*) as count from MemberAccessLog where memberId = #memberId# and accessTimestamp > #start# and accessTimestamp <= #end# group by actionId </select> |
4.輸入參數中含有數組
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
|
<insert id="updateStatusBatch" parameterClass="hashMap"> update Question set status = #status# <dynamic prepend="where questionId in"> <isNotNull property="actionIds"> <iterate property="actionIds" open="(" close=")" conjunction=","> #actionIds[]# </iterate> </isNotNull> </dynamic> </insert> <insert id="updateStatusBatch" parameterClass="hashMap"> update Question set status = #status# <dynamic prepend="where questionId in"> <isNotNull property="actionIds"> <iterate property="actionIds" open="(" close=")" conjunction=","> #actionIds[]# </iterate> </isNotNull> </dynamic> </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
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
|
<select id="com.fashionfree.stat.accesslog.model.StatMemberAction.selectActionIdsOfModule"resultClass="hashMap"> select moduleId, actionId from StatMemberAction <dynamic prepend="where moduleId in"> <iterate open="(" close=")" conjunction=","> #[]# </iterate> </dynamic> order by moduleId </select><select id="com.fashionfree.stat.accesslog.model.StatMemberAction.selectActionIdsOfModule"resultClass="hashMap"> select moduleId, actionId from StatMemberAction <dynamic prepend="where moduleId in"> <iterate open="(" close=")" conjunction=","> #[]# </iterate> </dynamic> order by moduleId </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
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
|
<select id="com.fashionfree.stat.accesslog.selectSumDistinctCountOfAccessMemberNum"parameterClass="hashMap" resultClass="int"> select count(distinct memberId) from MemberAccessLog where accessTimestamp >= #start# and accessTimestamp < #end# and actionId in $actionIdString$ </select> <select id="com.fashionfree.stat.accesslog.selectSumDistinctCountOfAccessMemberNum"parameterClass="hashMap" resultClass="int"> select count(distinct memberId) from MemberAccessLog where accessTimestamp >= #start# and accessTimestamp < #end# and actionId in $actionIdString$ </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)
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
|
<select id="com.fashionfree.stat.accesslog.selectMemberAccessLogBy"parameterClass="hashMap" resultMap="MemberAccessLogMap"> <include refid="selectAllSql"/> <include refid="whereSql"/> <include refid="pageSql"/> </select> <select id="com.fashionfree.stat.accesslog.selectMemberAccessLogBy.Count"parameterClass="hashMap" resultClass="int"> <include refid="countSql"/> <include refid="whereSql"/> </select> <sql id="selectAllSql"> select accessLogId, memberId, clientIP, httpMethod, actionId, requestURL, accessTimestamp, extend1, extend2, extend3 from MemberAccessLog </sql> <sql id="whereSql"> accessTimestamp <= #accessTimestamp# </sql> <sql id="countSql"> select count(*) from MemberAccessLog </sql> <sql id="pageSql"> <dynamic> <isNotNull property="startIndex"> <isNotNull property="pageSize"> limit #startIndex# , #pageSize# </isNotNull> </isNotNull> </dynamic> </sql><select id="com.fashionfree.stat.accesslog.selectMemberAccessLogBy"parameterClass="hashMap" resultMap="MemberAccessLogMap"> <include refid="selectAllSql"/> <include refid="whereSql"/> <include refid="pageSql"/> </select> <select id="com.fashionfree.stat.accesslog.selectMemberAccessLogBy.Count"parameterClass="hashMap" resultClass="int"> <include refid="countSql"/> <include refid="whereSql"/> </select> <sql id="selectAllSql"> select accessLogId, memberId, clientIP, httpMethod, actionId, requestURL, accessTimestamp, extend1, extend2, extend3 from MemberAccessLog </sql> <sql id="whereSql"> accessTimestamp <= #accessTimestamp# </sql> <sql id="countSql"> select count(*) from MemberAccessLog </sql> <sql id="pageSql"> <dynamic> <isNotNull property="startIndex"> <isNotNull property="pageSize"> limit #startIndex# , #pageSize# </isNotNull> </isNotNull> </dynamic> </sql> |
說明：本例中，代碼應為：
|
1
2
3
|
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語句中含有大于號>、小于號< 1. 將大于號、小于號寫為： > < 如：
|
1
2
3
4
5
6
7
8
9
10
11
12
13
|
<delete id="com.fashionfree.stat.accesslog.deleteMemberAccessLogsBefore" parameterClass="long"> delete from MemberAccessLog where accessTimestamp <= #value# </delete> Xml代碼 <delete id="com.fashionfree.stat.accesslog.deleteMemberAccessLogsBefore" parameterClass="long"> delete from MemberAccessLog where accessTimestamp <= #value# </delete> |
將特殊字符放在xml的CDATA區內：
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
<delete id="com.fashionfree.stat.accesslog.deleteMemberAccessLogsBefore" parameterClass="long"> <![CDATA[ delete from MemberAccessLog where accessTimestamp <= #value# ]]> </delete> <delete id="com.fashionfree.stat.accesslog.deleteMemberAccessLogsBefore" parameterClass="long"> <![CDATA[ delete from MemberAccessLog where accessTimestamp <= #value# ]]> </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，便于共用：
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
|
<sql id="selectBasicSql"> select samplingTimestamp,onlineNum,year, month,week,day,hour from OnlineMemberNum </sql> <sql id="whereSqlBefore"> where samplingTimestamp <= #samplingTimestamp# </sql> <select id="com.fashionfree.accesslog.selectOnlineMemberNumsBeforeSamplingTimestamp" parameterClass="hashmap" resultClass="OnlineMemberNum"> <include refid="selectBasicSql" /> <include refid="whereSqlBefore" /> </select> <sql id="selectBasicSql"> select samplingTimestamp,onlineNum,year, month,week,day,hour from OnlineMemberNum </sql> <sql id="whereSqlBefore"> where samplingTimestamp <= #samplingTimestamp# </sql> <select id="com.fashionfree.accesslog.selectOnlineMemberNumsBeforeSamplingTimestamp" parameterClass="hashmap" resultClass="OnlineMemberNum"> <include refid="selectBasicSql" /> <include refid="whereSqlBefore" /> </select> |
注意：sql標簽只能用于被引用，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:
|
1
|
sqlMapClient.queryForList(“selectBasicSql”); × |
10.隨機選取記錄
|
1
2
3
|
<sql id=”randomSql”> ORDER BY rand() LIMIT #number# </sql> |
從數據庫中隨機選取number條記錄（只適用于MySQL）
11.將SQL GROUP BY分組中的字段拼接
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
|
<sql id=”selectGroupBy> SELECT a.answererCategoryId, a.answererId, a.answererName, a.questionCategoryId, a.score, a.answeredNum, a.correctNum, a.answerSeconds, a.createdTimestamp, a.lastQuestionApprovedTimestamp, a.lastModified, GROUP_CONCAT(q.categoryName) as categoryName FROM AnswererCategory a, QuestionCategory q WHERE a.questionCategoryId = q.questionCategoryId GROUP BY a.answererId ORDER BY a.answererCategoryId </sql><sql id=”selectGroupBy> SELECT a.answererCategoryId, a.answererId, a.answererName, a.questionCategoryId, a.score, a.answeredNum, a.correctNum, a.answerSeconds, a.createdTimestamp, a.lastQuestionApprovedTimestamp, a.lastModified, GROUP_CONCAT(q.categoryName) as categoryName FROM AnswererCategory a, QuestionCategory q WHERE a.questionCategoryId = q.questionCategoryId GROUP BY a.answererId ORDER BY a.answererCategoryId </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:
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
|
<sql id=”groupByInArea”> select moduleId, moduleName, status, lastModifierId, lastModifiedName, lastModified from StatModule where moduleId in (3, 5, 1) order by instr(',3,5,1,' , ','+ltrim(moduleId)+',') </sql> <sql id=”groupByInArea”> select moduleId, moduleName, status, lastModifierId, lastModifiedName, lastModified from StatModule where moduleId in (3, 5, 1) order by instr(',3,5,1,' , ','+ltrim(moduleId)+',') </sql> |
②SQLSERVER:
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
|
<sql id=”groupByInArea”> select moduleId, moduleName, status, lastModifierId, lastModifiedName, lastModified from StatModule where moduleId in (3, 5, 1) order by charindex(','+ltrim(moduleId)+',' , ',3,5,1,') </sql> <sql id=”groupByInArea”> select moduleId, moduleName, status, lastModifierId, lastModifiedName, lastModified from StatModule where moduleId in (3, 5, 1) order by charindex(','+ltrim(moduleId)+',' , ',3,5,1,') </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
|
1
|
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
|
1
2
3
4
5
6
7
8
9
|
<resultMap class="java.util.HashMap" id="getActionIdAndActionNumber"> <result column="actionId" property="actionId" jdbcType="BIGINT" javaType="long"/> <result column="count" property="count" jdbcType="INT" javaType="int"/> </resultMap> Xml代碼 <resultMap class="java.util.HashMap" id="getActionIdAndActionNumber"> <result column="actionId" property="actionId" jdbcType="BIGINT" javaType="long"/> <result column="count" property="count" jdbcType="INT" javaType="int"/> </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
|
1
2
|
<typeAlias alias="MemberOnlineDuration" type="com.fashionfree.stat.accesslog.model.MemberOnlineDuration" /> <typeAlias> |
允許你定義別名，避免重復輸入過長的名字
15.remap
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
|
<select id="testForRemap" parameterClass="hashMap" resultClass="hashMap" remapResults="true"> select userId <isEqual property="tag" compareValue="1"> , userName </isEqual> <isEqual property="tag" compareValue="2"> , userPassword </isEqual> from UserInfo </select> <select id="testForRemap" parameterClass="hashMap" resultClass="hashMap" remapResults="true"> select userId <isEqual property="tag" compareValue="1"> , userName </isEqual> <isEqual property="tag" compareValue="2"> , userPassword </isEqual> from UserInfo </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：
|
1
2
3
4
5
|
HashMap<String, Integer> hashMap = new HashMap<String, Integer>(); hashMap.put("tag", 1); sqlClient.queryForList("testForRemap", hashMap); hashMap.put("tag", 2); sqlClient.queryForList("testForRemap", hashMap); |
Java代碼
|
1
2
3
4
5
|
HashMap<String, Integer> hashMap = new HashMap<String, Integer>(); hashMap.put("tag", 1); sqlClient.queryForList("testForRemap", hashMap); hashMap.put("tag", 2); 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：
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
|
<sql id="whereSql"> <dynamic prepend="where "> <isNotNull property="userId" prepend="BOGUS"> userId = #userId# </isNotNull> <isNotEmpty property="userName" prepend="and "> userName = #userName# </isNotEmpty> </dynamic> </sql> <sql id="whereSql"> <dynamic prepend="where "> <isNotNull property="userId" prepend="BOGUS"> userId = #userId# </isNotNull> <isNotEmpty property="userName" prepend="and "> userName = #userName# </isNotEmpty> </dynamic> </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。
以上所述是小編給大家介紹的iBatis習慣用的16條SQL語句，希望對大家有所幫助，如果大家有任何疑問請給我留言，小編會及時回復大家的。在此也非常感謝大家對服務器之家網站的支持！
原文鏈接：http://www.cnblogs.com/huiy/archive/2016/10/26/6000676.html
