首页 > 代码库 > 金蝶BOS 7.5 SQL语句生成分析
金蝶BOS 7.5 SQL语句生成分析
今天刚好有空,就分析下 金蝶BOS 7.5 SQL语句生成 方式.(分析环境 Oracle 11.0.2 , 金蝶BOS 7.5.0)
1 操作思路
思路是这样的，在开发环境触发SQL操作，然后在数据库中查出最近执行的SQL，对其分析．
数据库里使用SQL语句：
--这里只查询JDBC操作的SQL语句 SELECT t.sql_id, t.sql_text, t.sql_fulltext, to_char(t.last_active_time, ‘yyyy-mm-dd HH24:mi:ss‘) FROM v$sql t WHERE t.last_active_time > SYSDATE - 0.0005 AND t.module = ‘JDBC Thin Client‘ ORDER BY t.last_active_time DESC； --根据SQL＿ID获取执行时参数值 SELECT child_number, NAME, position, datatype_string, value_string, last_captured FROM v$sql_bind_capture WHERE sql_id = ‘8du5umjut76wc‘ ORDER BY child_number, position;
来获取最近一分钟内执行的SQL．
2 操作数据
实体找了带有一个分录的基础资料,数据共5条. ( 实体 package-> CT_BAS_PACKAGE,分录 E1 ->CT_BAS_PACKAGEE1 ).
3 操作过程
3.1 KSQL
为了操作直观,这里不使用 EntityViewInfo 操作,而是直接从 KSQL 下手,操作代码类似:
IPackage ipackage = PackageFactory.getRemoteInstance();
package.getPackageCollection("select * where number=‘BZ0001‘");//select * where number=‘BZ0001‘ 为KSQL
分析结果:
--KSQL：select * where number=‘BZ0001‘ SELECT "T0"."FID" "ID", "T0"."FCREATORID" "T1.ID", "T0"."FCREATETIME" "CREATETIME", "T0"."FLASTUPDATEUSERID" "T2.ID", "T0"."FLASTUPDATETIME" "LASTUPDATETIME", "T0"."FCONTROLUNITID" "T3.ID", "T0"."FNAME_L2" "NAME", "T0"."FNAME_L1" "NAME_L1", "T0"."FNAME_L2" "NAME_L2", "T0"."FNAME_L3" "NAME_L3", "T0"."FNUMBER" "NUMBER", "T0"."FDESCRIPTION_L2" "DESCRIPTION", "T0"."FDESCRIPTION_L1" "DESCRIPTION_L1", "T0"."FDESCRIPTION_L2" "DESCRIPTION_L2", "T0"."FDESCRIPTION_L3" "DESCRIPTION_L3", "T0"."FSIMPLENAME" "SIMPLENAME", "T0"."CFSTATE" "STATE" FROM "CT_BAS_PACKAGE" "T0" WHERE "T0"."FNUMBER" = :1; --查询分录ID SELECT "T0"."FID" "ID", "T0"."FPARENTID" "T1.ID" FROM "CT_BAS_PACKAGEE1" "T0" INNER JOIN "CT_BAS_PACKAGE" "T1" ON "T0"."FPARENTID" = "T1"."FID" WHERE "T1"."FNUMBER" = :1 ORDER BY "T1.ID" ASC, "T0"."FSEQ" ASC; --总结:如果不指定,会查询主表数据,加分录数据中的 fid 字段 --KSQL：select number where number=‘BZ0001‘ SELECT "T0"."FID" "ID", "T0"."FNUMBER" "NUMBER" FROM "CT_BAS_PACKAGE" "T0" WHERE "T0"."FNUMBER" = :1; -- SELECT "T0"."FID" "ID", "T0"."FPARENTID" "T1.ID" FROM "CT_BAS_PACKAGEE1" "T0" INNER JOIN "CT_BAS_PACKAGE" "T1" ON "T0"."FPARENTID" = "T1"."FID" WHERE "T1"."FNUMBER" = :1 ORDER BY "T1.ID" ASC, "T0"."FSEQ" ASC; --总结:即使在主表中指定了主数据的查询字段,仍然会查询分录 fid ,真是让人失望 --KSQL：select E1.id where number=‘BZ0001‘ SELECT "T0"."FID" "ID" FROM "CT_BAS_PACKAGE" "T0" WHERE "T0"."FNUMBER" = :1; -- SELECT "T0"."FID" "ID", "T0"."FPARENTID" "T1.ID" FROM "CT_BAS_PACKAGEE1" "T0" INNER JOIN "CT_BAS_PACKAGE" "T1" ON "T0"."FPARENTID" = "T1"."FID" WHERE "T1"."FNUMBER" = :1 ORDER BY "T1.ID" ASC, "T0"."FSEQ" ASC; --总结:这里也是很让人失望,指定分录中字段查询,仍会查询两次,BOS对于ID真是有很深的执念啊 --KSQL：select E1.* where number=‘BZ0001‘ SELECT "T0"."FID" "ID" FROM "CT_BAS_PACKAGE" "T0" WHERE "T0"."FNUMBER" = :1; -- SELECT "T0"."FID" "ID", "T0"."FSEQ" "SEQ", "T0"."FPARENTID" "T1.ID", "T0"."CFNAME" "NAME", "T0"."CFREMARK" "REMARK", "T0"."CFWIDTH" "WIDTH", "T0"."CFHIGTH" "HIGTH", "T0"."CFLONGNUM" "LONGNUM" FROM "CT_BAS_PACKAGEE1" "T0" INNER JOIN "CT_BAS_PACKAGE" "T1" ON "T0"."FPARENTID" = "T1"."FID" WHERE "T1"."FNUMBER" = :1 ORDER BY "T1.ID" ASC, "T0"."FSEQ" ASC; --总结:同上 --KSQL：select E1 where number=‘BZ0001‘ SELECT "T0"."FID" "ID" FROM "CT_BAS_PACKAGE" "T0" WHERE "T0"."FNUMBER" = :1; -- SELECT "T0"."FID" "ID", "T0"."FPARENTID" "T1.ID" FROM "CT_BAS_PACKAGEE1" "T0" INNER JOIN "CT_BAS_PACKAGE" "T1" ON "T0"."FPARENTID" = "T1"."FID" WHERE "T1"."FNUMBER" = :1 ORDER BY "T1.ID" ASC, "T0"."FSEQ" ASC; --总结: 这里分录 E1 转换成 E1.fid ,而不是 E1.* ,这个做得好.
3.2 exists 操作
IPackage ipackage = PackageFactory.getRemoteInstance();
FilterInfo filter=new FilterInfo();
filter.appendFilterItem("number", "BZ0001");
ipackage.exists(filter);
数据库SQL
SELECT "T0"."FID" FROM "CT_BAS_PACKAGE" "T0" WHERE "T0"."FNUMBER" = :1 AND rownum <= 1;
这个让人还算满意.
3.3 Query
大家都知道在EAS 里面 list 界面使用的是 Query的数据,用于筛选和排序.
Query里显示的SQL
SELECT "PACKAGE".FID AS "ID", "PACKAGE".FNumber AS "NUMBER", "PACKAGE".FName_l2 AS "NAME", "PACKAGE".FDescription_l2 AS "DESCRIPTION", "PACKAGE".FLastUpdateTime AS "LASTUPDATETIME", "LASTUPDATEUSER".FName_l2 AS "LASTUPDATEUSER.NAME", "PACKAGE".FCreateTime AS "CREATETIME", "CREATOR".FName_l2 AS "CREATOR.NAME", "PACKAGE".CFState AS "STATE" FROM ct_bas_package AS "PACKAGE" INNER JOIN t_pm_user AS "CREATOR" ON "PACKAGE".FCreatorID = "CREATOR".FID INNER JOIN t_pm_user AS "LASTUPDATEUSER" ON "PACKAGE".FLastUpdateUserID = "LASTUPDATEUSER".FID WHERE "PACKAGE".CFState = 1 ORDER BY "LASTUPDATETIME" ASC, "NUMBER" ASC;
实际执行的SQL:
--创建表,存入查询数据(这张表,晚点会被删掉),只有 ksql_seq, id 两列,用于保存顺序和ID
INSERT INTO vtzoggntt1c6bxfd8u1mbc4q020
(ksql_seq, id)
SELECT rownum,
TMP_SUBSELECT_ALIAS."ID"
FROM (SELECT *
FROM (SELECT "PACKAGE".FID "ID",
"PACKAGE".FLastUpdateTime "LASTUPDATETIME",
"PACKAGE".FNumber "NUMBER"
FROM ct_bas_package "PACKAGE"
INNER JOIN t_pm_user "CREATOR" ON "PACKAGE".FCreatorID =
"CREATOR".FID
INNER JOIN t_pm_user "LASTUPDATEUSER" ON "PACKAGE".FLastUpdateUserID =
"LASTUPDATEUSER".FID
WHERE ("PACKAGE".CFState = :1 AND
("PACKAGE".FControlUnitID = :2 OR
"PACKAGE".FControlUnitID = :3 OR
"PACKAGE".FControlUnitID = :4))
ORDER BY "LASTUPDATETIME" ASC,
"NUMBER" ASC)
WHERE rownum <= 10000) tmp_subselect_alias
WHERE rownum <= 10000;
--查询数据总数
SELECT *
FROM (SELECT ksql_seq FROM vtzoggntt1c6bxfd8u1mbc4q020 ORDER BY ksql_seq DESC)
WHERE rownum <= 1;
--查询前50条数据( 这个50 从哪里来的,不清楚)
SELECT *
FROM vtzoggntt1c6bxfd8u1mbc4q020
WHERE ((ksql_seq >= :1) AND (ksql_seq <= :2))
ORDER BY ksql_seq ASC;
--这里只有5条数据,为了方便看,我把参数加到SQL里了,可以看出,根据上面的查询结果,这里分两种情况,分别是ID和....真是任性
SELECT "PACKAGE".FID "ID",
"PACKAGE".FNumber "NUMBER",
"PACKAGE".FName_l2 "NAME",
"PACKAGE".FDescription_l2 "DESCRIPTION",
"PACKAGE".FLastUpdateTime "LASTUPDATETIME",
"LASTUPDATEUSER".FName_l2 "LASTUPDATEUSER.NAME",
"PACKAGE".FCreateTime "CREATETIME",
"CREATOR".FName_l2 "CREATOR.NAME",
"PACKAGE".CFState "STATE"
FROM ct_bas_package "PACKAGE"
INNER JOIN t_pm_user "CREATOR" ON "PACKAGE".FCreatorID = "CREATOR".FID
INNER JOIN t_pm_user "LASTUPDATEUSER" ON "PACKAGE".FLastUpdateUserID =
"LASTUPDATEUSER".FID
WHERE (("PACKAGE".FID IN
(‘PKIAAAABDe9S2ZK/‘, :2, :3, :4, :5, ‘NE‘ :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)) AND
("PACKAGE".FID = ‘PKIAAAABDe9S2ZK/‘ OR
(("PACKAGE".FID IS NULL) AND 0 = 1) OR "PACKAGE".FID = :53 OR
(("PACKAGE".FID IS NULL) AND :54 = 1) OR "PACKAGE".FID = :55 OR
(("PACKAGE".FID IS NULL) AND :56 = 1) OR "PACKAGE".FID = :57 OR
(("PACKAGE".FID IS NULL) AND :58 = 1) OR "PACKAGE".FID = :59 OR
(("PACKAGE".FID IS NULL) AND :60 = 1) OR "PACKAGE".FID = :61 OR
(("PACKAGE".FID IS NULL) AND :62 = 1) OR "PACKAGE".FID = :63 OR
(("PACKAGE".FID IS NULL) AND :64 = 1) OR "PACKAGE".FID = :65 OR
(("PACKAGE".FID IS NULL) AND :66 = 1) OR "PACKAGE".FID = :67 OR
(("PACKAGE".FID IS NULL) AND :68 = 1) OR "PACKAGE".FID = :69 OR
(("PACKAGE".FID IS NULL) AND :70 = 1) OR "PACKAGE".FID = :71 OR
(("PACKAGE".FID IS NULL) AND :72 = 1) OR "PACKAGE".FID = :73 OR
(("PACKAGE".FID IS NULL) AND :74 = 1) OR "PACKAGE".FID = :75 OR
(("PACKAGE".FID IS NULL) AND :76 = 1) OR "PACKAGE".FID = :77 OR
(("PACKAGE".FID IS NULL) AND :78 = 1) OR "PACKAGE".FID = :79 OR
(("PACKAGE".FID IS NULL) AND :80 = 1) OR "PACKAGE".FID = :81 OR
(("PACKAGE".FID IS NULL) AND :82 = 1) OR "PACKAGE".FID = :83 OR
(("PACKAGE".FID IS NULL) AND :84 = 1) OR "PACKAGE".FID = :85 OR
(("PACKAGE".FID IS NULL) AND :86 = 1) OR "PACKAGE".FID = :87 OR
(("PACKAGE".FID IS NULL) AND :88 = 1) OR "PACKAGE".FID = :89 OR
(("PACKAGE".FID IS NULL) AND :90 = 1) OR "PACKAGE".FID = :91 OR
(("PACKAGE".FID IS NULL) AND :92 = 1) OR "PACKAGE".FID = :93 OR
(("PACKAGE".FID IS NULL) AND :94 = 1) OR "PACKAGE".FID = :95 OR
(("PACKAGE".FID IS NULL) AND :96 = 1) OR "PACKAGE".FID = :97 OR
(("PACKAGE".FID IS NULL) AND :98 = 1) OR "PACKAGE".FID = :99 OR
(("PACKAGE".FID IS NULL) AND :100 = 1) OR "PACKAGE".FID = :101 OR
(("PACKAGE".FID IS NULL) AND :102 = 1) OR "PACKAGE".FID = :103 OR
(("PACKAGE".FID IS NULL) AND :104 = 1) OR "PACKAGE".FID = :105 OR
(("PACKAGE".FID IS NULL) AND :106 = 1) OR "PACKAGE".FID = :107 OR
(("PACKAGE".FID IS NULL) AND :108 = 1) OR "PACKAGE".FID = :109 OR
(("PACKAGE".FID IS NULL) AND :110 = 1) OR "PACKAGE".FID = :111 OR
(("PACKAGE".FID IS NULL) AND :112 = 1) OR "PACKAGE".FID = :113 OR
(("PACKAGE".FID IS NULL) AND :114 = 1) OR "PACKAGE".FID = :115 OR
(("PACKAGE".FID IS NULL) AND :116 = 1) OR "PACKAGE".FID = :117 OR
(("PACKAGE".FID IS NULL) AND :118 = 1) OR "PACKAGE".FID = :119 OR
(("PACKAGE".FID IS NULL) AND :120 = 1) OR "PACKAGE".FID = :121 OR
(("PACKAGE".FID IS NULL) AND :122 = 1) OR "PACKAGE".FID = :123 OR
(("PACKAGE".FID IS NULL) AND :124 = 1) OR "PACKAGE".FID = :125 OR
(("PACKAGE".FID IS NULL) AND :126 = 1) OR "PACKAGE".FID = :127 OR
(("PACKAGE".FID IS NULL) AND :128 = 1) OR "PACKAGE".FID = :129 OR
(("PACKAGE".FID IS NULL) AND :130 = 1) OR "PACKAGE".FID = :131 OR
(("PACKAGE".FID IS NULL) AND :132 = 1) OR "PACKAGE".FID = :133 OR
(("PACKAGE".FID IS NULL) AND :134 = 1) OR "PACKAGE".FID = :135 OR
(("PACKAGE".FID IS NULL) AND :136 = 1) OR "PACKAGE".FID = :137 OR
(("PACKAGE".FID IS NULL) AND :138 = 1) OR "PACKAGE".FID = :139 OR
(("PACKAGE".FID IS NULL) AND :140 = 1) OR "PACKAGE".FID = :141 OR
(("PACKAGE".FID IS NULL) AND :142 = 1) OR "PACKAGE".FID = :143 OR
(("PACKAGE".FID IS NULL) AND :144 = 1) OR "PACKAGE".FID = :145 OR
(("PACKAGE".FID IS NULL) AND :146 = 1) OR "PACKAGE".FID = :147 OR
(("PACKAGE".FID IS NULL) AND :148 = 1) OR "PACKAGE".FID = :149 OR
(("PACKAGE".FID IS NULL) AND :150 = 1)))
ORDER BY "LASTUPDATETIME" ASC,
"NUMBER" ASC;
总之,我认为EAS BOS ７.５ 在SQL语句的生成方式上,相对 Hibernate之流相甚远!!!!完全没有效率可言.特别是创表存储数据,稍后删除的作法,更是让人不可思议.
金蝶BOS 7.5 SQL语句生成分析
声明：以上内容来自用户投稿及互联网公开渠道收集整理发布，本网站不拥有所有权，未作人工编辑处理，也不承担相关法律责任，若内容有误或涉及侵权可进行投诉： 投诉/举报 工作人员会在5个工作日内联系你，一经查实，本站将立刻删除涉嫌侵权内容。