在 PL-SQL 存储过程中为 XML 中的每个段生成唯一 ID
Posted
技术标签:
【中文标题】在 PL-SQL 存储过程中为 XML 中的每个段生成唯一 ID【英文标题】：Unique ID generation in PL-SQL stored procedure for each segment in XML 【发布时间】：2017-09-14 09:22:52 【问题描述】：我编写了一个存储过程来将数据从 xml 存储到表中。以下是 DDL
CREATE TABLE STUDENT (
CONTROL_ID NUMBER(*,0),
BATCHS VARCHAR2(255),
FNAME VARCHAR2(255),
ROLL VARCHAR2(255),
CITY VARCHAR2(255),
CONSTRAINT PK_STUDENT PRIMARY KEY (CONTROL_ID, BATCHS)
);
CREATE TABLE MARKS_CHAR (
CONTROL_ID NUMBER(*,0),
BATCHS VARCHAR2(255),
SUBJECT VARCHAR2(255),
MARK_NUMBER NUMBER(*,0),
MARK_GRADE VARCHAR2(255),
UNIT VARCHAR2(255),
FOREIGN KEY (CONTROL_ID, BATCHS) REFERENCES STUDENT (CONTROL_ID, BATCHS)
);
CREATE SEQUENCE SEQ_STUDENT MINVALUE 1000 MAXVALUE 9999999999999999999999999999 INCREMENT BY 3 START WITH 1000 CACHE 10 NOORDER NOCYCLE NOPARTITION ;
&以下分别是程序和示例xml
create or replace PROCEDURE STUDENT_MARKS_UPLOAD(XMLINPUT IN CLOB) AS
controlid INT;
BEGIN
SELECT SEQ_STUDENT.NEXTVAL INTO controlid FROM DUAL;
INSERT INTO STUDENT(CONTROL_ID,BATCHS,FNAME,ROLL,CITY)
SELECT controlid,x.* FROM xmltable (
XMLNAMESPACES('http://some.clg.india/statelevel' as "ns"),
'//ns:student'
PASSING xmltype(XMLINPUT)
COLUMNS
BATCHS VARCHAR2(255) PATH '//ns:header/ns:batch',
FNAME VARCHAR2(255) PATH '//ns:header/ns:name',
ROLL VARCHAR2(255) PATH '//ns:header/ns:roll',
CITY VARCHAR2(255) PATH '//ns:header/ns:city'
)x;
INSERT INTO MARKS_CHAR(CONTROL_ID,UNIT,BATCHS,SUBJECT,MARK_GRADE)
SELECT controlid,'string',x.* FROM xmltable (
XMLNAMESPACES('http://some.clg.india/statelevel' as "ns"),
'//ns:marks[ns:subject/text()!=''Humanities'' and ns:subject/text()!=''Values and Ethics'' and ((string-length(ns:gwrt/text())>0) and (string-length(ns:mnum/text())=0))]'
PASSING xmltype(XMLINPUT)
COLUMNS
BATCHS VARCHAR2(255) PATH './parent::*/ns:header/ns:batch',
SUBJECT VARCHAR2(255) PATH '//ns:subject',
MARK_GRADE VARCHAR2(2000) PATH '//ns:gwrt'
)x;
INSERT INTO MARKS_CHAR(CONTROL_ID,UNIT,BATCHS,SUBJECT,MARK_NUMBER)
SELECT controlid,'number',x.* FROM xmltable (
XMLNAMESPACES('http://some.clg.india/statelevel' as "ns"),
'//ns:marks[ns:subject/text()!=''Humanities'' and ns:subject/text()!=''Values and Ethics'' and ((string-length(ns:mnum/text())>0) and (string-length(ns:gwrt/text())=0))]'
PASSING xmltype(XMLINPUT)
COLUMNS
BATCHS VARCHAR2(255) PATH './parent::*/ns:header/ns:batch',
SUBJECT VARCHAR2(255) PATH '//ns:subject',
MARK_NUMBER INT PATH '//ns:mnum'
)x;
END ANNOUNCEITEMS_UPLOAD;
<class xmlns="http://some.clg.india/statelevel">
<student>
<header>
<name>Topesh</name>
<roll>0002</roll>
<batch>A001</batch>
<address>WB, India</address>
<city>Kolkata</city>
</header>
<marks>
<subject>Math</subject>
<mnum>85</mnum>
</marks>
<marks>
<subject>Language</subject>
<gwrt>A</gwrt>
</marks>
<marks>
<subject>Humanities</subject>
<mnum>89</mnum>
</marks>
<marks>
<subject>Geo</subject>
<gwrt>O</gwrt>
</marks>
<marks>
<subject>History</subject>
<gwrt>A+</gwrt>
</marks>
<marks>
<subject>Physics</subject>
<mnum>90</mnum>
</marks>
<marks>
<subject>Chemistry</subject>
<gwrt>B+</gwrt>
</marks>
<marks>
<subject>Values and Ethics</subject>
<mnum>93</mnum>
</marks>
</student>
<student>
<header>
<name>Subham</name>
<roll>0003</roll>
<batch>B002</batch>
<address>UP, India</address>
<city>Kanpur</city>
</header>
<marks>
<subject>Math</subject>
<mnum>98</mnum>
</marks>
<marks>
<subject>Language</subject>
<gwrt>C+</gwrt>
</marks>
<marks>
<subject>Humanities</subject>
<mnum>67</mnum>
</marks>
<marks>
<subject>Geo</subject>
<gwrt>E</gwrt>
</marks>
<marks>
<subject>History</subject>
<gwrt>A</gwrt>
</marks>
<marks>
<subject>Physics</subject>
<mnum>97</mnum>
</marks>
<marks>
<subject>Chemistry</subject>
<gwrt>O</gwrt>
</marks>
<marks>
<subject>Values and Ethics</subject>
<mnum>58</mnum>
</marks>
</student>
<student>
<header>
<name>Vinay</name>
<roll>0001</roll>
<batch>C003</batch>
<address>Kerakla, India</address>
<city>Bangalore</city>
</header>
<marks>
<subject>Math</subject>
<mnum>99</mnum>
</marks>
<marks>
<subject>Language</subject>
<gwrt>B</gwrt>
</marks>
<marks>
<subject>Humanities</subject>
<mnum>81</mnum>
</marks>
<marks>
<subject>Geo</subject>
<gwrt>E</gwrt>
</marks>
<marks>
<subject>History</subject>
<gwrt>E</gwrt>
</marks>
<marks>
<subject>Physics</subject>
<mnum>92</mnum>
</marks>
<marks>
<subject>Chemistry</subject>
<gwrt>E</gwrt>
</marks>
<marks>
<subject>Values and Ethics</subject>
<mnum>87</mnum>
</marks>
</student>
</class>
输出如下：
对于表 STUDENT
CONTROL_ID BATCHS FNAME ROLL CITY
1000 A001 Topesh 0002 Kolkata
1000 B002 Subham 0003 Kanpur
1000 C003 Vinay 0001 Bangalore
对于表MARKS_CHAR
CONTROL_ID BATCHS SUBJECT MARK_NUMBER MARK_GRADE UNIT
1000 A001 Language A string
1000 A001 Geo O string
1000 A001 History A+ string
1000 A001 Chemistry B+ string
1000 B002 Language C+ string
1000 B002 Geo E string
1000 B002 History A string
1000 B002 Chemistry O string
1000 C003 Language B string
1000 C003 Geo E string
1000 C003 History E string
1000 C003 Chemistry E string
1000 A001 Math 85 number
1000 A001 Physics 90 number
1000 B002 Math 98 number
1000 B002 Physics 97 number
1000 C003 Math 99 number
1000 C003 Physics 92 number
但要求是对于每个student 段，应该有一个唯一的CONTROL_ID，即，所需的输出将是
CONTROL_ID BATCHS FNAME ROLL CITY
1000 A001 Topesh 0002 Kolkata
1003 B002 Subham 0003 Kanpur
1006 C003 Vinay 0001 Bangalore
和
CONTROL_ID BATCHS SUBJECT MARK_NUMBER MARK_GRADE UNIT
1000 A001 Language A string
1000 A001 Geo O string
1000 A001 History A+ string
1000 A001 Chemistry B+ string
1003 B002 Language C+ string
1003 B002 Geo E string
1003 B002 History A string
1003 B002 Chemistry O string
1006 C003 Language B string
1006 C003 Geo E string
1006 C003 History E string
1006 C003 Chemistry E string
1000 A001 Math 85 number
1000 A001 Physics 90 number
1003 B002 Math 98 number
1003 B002 Physics 97 number
1006 C003 Math 99 number
1006 C003 Physics 92 number
请帮助如何在不更改任何架构的情况下在程序中实现这一点。 我正在使用 SQLDeveloper
【问题讨论】：
【参考方案1】：您创建了序列，但在 INSERT 操作后没有增加它。如果您使用的是 Oracle 11 或更早版本的创建触发器。否则使用自增列。
How to create id with AUTO_INCREMENT on Oracle?
【讨论】：
以上是关于在 PL-SQL 存储过程中为 XML 中的每个段生成唯一 ID的主要内容，如果未能解决你的问题，请参考以下文章
获取几个月的最新记录并使用 Oracle PL-SQL 为每个 ID 汇总其值