关于java中实现在oracle数据库中实现对中文首字母进行排序的解决方案

首先介绍Oracle 9i新增加的一个系统自带的排序函数
1、按首字母排序
在oracle9i中新增了按照拼音、部首、笔画排序功能。设置NLS_SORT值
SCHINESE_RADICAL_M 按照部首（第一顺序）、笔划（第二顺序）排序
SCHINESE_STROKE_M 按照笔划（第一顺序）、部首（第二顺序）排序
SCHINESE_PINYIN_M 按照拼音排序
oracle9i中新增了按照拼音、部首、笔画排序功能
用法示例:

Java代码


拼音

SELECT * FROM TEAM ORDER BY NLSSORT(排序字段,'NLS_SORT = SCHINESE_PINYIN_M')

笔划

SELECT * FROM TEAM ORDER BY NLSSORT(排序字段,'NLS_SORT = SCHINESE_STROKE_M')

部首

SELECT * FROM TEAM ORDER BY NLSSORT(排序字段,'NLS_SORT = SCHINESE_RADICAL_M')

2、按首字母拼音查询
首先我们要创建一个可以查询汉字首字母的函数，函数代码如下：

Java代码


CREATE OR REPLACE FUNCTION F_TRANS_PINYIN_CAPITAL(P_NAME IN VARCHAR2) RETURN VARCHAR2 AS

V_COMPARE VARCHAR2(100);

V_RETURN VARCHAR2(4000);

FUNCTION F_NLSSORT(P_WORD IN VARCHAR2) RETURN VARCHAR2 AS

BEGIN

RETURN NLSSORT(P_WORD, 'NLS_SORT=SCHINESE_PINYIN_M');

END;

BEGIN

FOR I IN 1..LENGTH(P_NAME) LOOP

V_COMPARE := F_NLSSORT(SUBSTR(P_NAME, I, 1));

IF V_COMPARE >= F_NLSSORT(' 吖 ') AND V_COMPARE <= F_NLSSORT('驁 ') THEN

V_RETURN := V_RETURN || 'a';

ELSIF V_COMPARE >= F_NLSSORT('八 ') AND V_COMPARE <= F_NLSSORT('簿 ') THEN

V_RETURN := V_RETURN || 'b';

ELSIF V_COMPARE >= F_NLSSORT('嚓 ') AND V_COMPARE <= F_NLSSORT('錯 ') THEN

V_RETURN := V_RETURN || 'c';

ELSIF V_COMPARE >= F_NLSSORT('咑 ') AND V_COMPARE <= F_NLSSORT('鵽 ') THEN

V_RETURN := V_RETURN || 'd';

ELSIF V_COMPARE >= F_NLSSORT('妸 ') AND V_COMPARE <= F_NLSSORT('樲 ') THEN

V_RETURN := V_RETURN || 'e';

ELSIF V_COMPARE >= F_NLSSORT('发 ') AND V_COMPARE <= F_NLSSORT('猤 ') THEN

V_RETURN := V_RETURN || 'f';

ELSIF V_COMPARE >= F_NLSSORT('旮 ') AND V_COMPARE <= F_NLSSORT('腂 ') THEN

V_RETURN := V_RETURN || 'g';

ELSIF V_COMPARE >= F_NLSSORT('妎 ') AND V_COMPARE <= F_NLSSORT('夻 ') THEN

V_RETURN := V_RETURN || 'h';

ELSIF V_COMPARE >= F_NLSSORT('丌 ') AND V_COMPARE <= F_NLSSORT('攈 ') THEN

V_RETURN := V_RETURN || 'j';

ELSIF V_COMPARE >= F_NLSSORT('咔 ') AND V_COMPARE <= F_NLSSORT('穒 ') THEN

V_RETURN := V_RETURN || 'k';

ELSIF V_COMPARE >= F_NLSSORT('垃 ') AND V_COMPARE <= F_NLSSORT('擽 ') THEN

V_RETURN := V_RETURN || 'l';

ELSIF V_COMPARE >= F_NLSSORT('嘸 ') AND V_COMPARE <= F_NLSSORT('椧 ') THEN

V_RETURN := V_RETURN || 'm';

ELSIF V_COMPARE >= F_NLSSORT('拏 ') AND V_COMPARE <= F_NLSSORT('瘧 ') THEN

V_RETURN := V_RETURN || 'n';

ELSIF V_COMPARE >= F_NLSSORT('筽 ') AND V_COMPARE <= F_NLSSORT('漚 ') THEN

V_RETURN := V_RETURN || 'o';

ELSIF V_COMPARE >= F_NLSSORT('妑 ') AND V_COMPARE <= F_NLSSORT('曝 ') THEN

V_RETURN := V_RETURN || 'p';

ELSIF V_COMPARE >= F_NLSSORT('七 ') AND V_COMPARE <= F_NLSSORT('裠 ') THEN

V_RETURN := V_RETURN || 'q';

ELSIF V_COMPARE >= F_NLSSORT('亽 ') AND V_COMPARE <= F_NLSSORT('鶸 ') THEN

V_RETURN := V_RETURN || 'r';

ELSIF V_COMPARE >= F_NLSSORT('仨 ') AND V_COMPARE <= F_NLSSORT('蜶 ') THEN

V_RETURN := V_RETURN || 's';

ELSIF V_COMPARE >= F_NLSSORT('侤 ') AND V_COMPARE <= F_NLSSORT('籜 ') THEN

V_RETURN := V_RETURN || 't';

ELSIF V_COMPARE >= F_NLSSORT('屲 ') AND V_COMPARE <= F_NLSSORT('鶩 ') THEN

V_RETURN := V_RETURN || 'w';

ELSIF V_COMPARE >= F_NLSSORT('夕 ') AND V_COMPARE <= F_NLSSORT('鑂 ') THEN

V_RETURN := V_RETURN || 'x';

ELSIF V_COMPARE >= F_NLSSORT('丫 ') AND V_COMPARE <= F_NLSSORT('韻 ') THEN

V_RETURN := V_RETURN || 'y';

ELSIF V_COMPARE >= F_NLSSORT('帀 ') AND V_COMPARE <= F_NLSSORT('咗 ') THEN

V_RETURN := V_RETURN || 'z';

END IF;

END LOOP;

RETURN V_RETURN;

END;

随后，我们可以先取出姓名，然后截取第一个汉字，最后取汉字的首字母，即可用来作为查询条件

Java代码


select e.fullname,e.expert_id from expert e where f_trans_pinyin_capital(substr(e.fullname,1,1)) = 'z'

最后展示一个将两个结合起来运用的例子：

Java代码


select e.expert_id,e.eperson_id,e.fullname,

sum(b2i.browse_num) browsecount,

count(o.object_id),wmsys.wm_concat(distinct d.name)

from expert e

left join expert2disciplinetype e2d on e2d.expert_id = e.expert_id

left join disciplinetype d on d.discipline_type_id = e2d.discipline_type_id

and d.upid = '0'

left join community2expert c2e on c2e.expert_id = e.expert_id

left join item i on i.item_type_id = 'AcademicRes' and i.withdrawn = 'N'

join metadatavalue m on m.item_id = i.item_id and m.metadata_field_id = '64'

join objectpublishinfo o on o.object_id = i.item_id

and o.object_type = 'item' and o.viewobjecttype = 'eperson'

and o.viewobjectid = e.eperson_id and o.state = '1'

and o.publishstate_id = '3'

left join browse2item b2i on b2i.item_id = o.object_id

where e.state = '1' and f_trans_pinyin_capital(substr(e.fullname,1,1)) = 'z'

group by e.expert_id,e.eperson_id,e.fullname

order by nlssort(e.fullname,'NLS_SORT=SCHINESE_PINYIN_M')