首先介紹Oracle 9i新增加的一個系統自帶的排序函數
1、按首字母排序
在oracle9i中新增了按照拼音、部首、筆畫排序功能。設置NLS_SORT值
SCHINESE_RADICAL_M 按照部首（第一順序）、筆劃（第二順序）排序
SCHINESE_STROKE_M 按照筆劃（第一順序）、部首（第二順序）排序
SCHINESE_PINYIN_M 按照拼音排序
oracle9i中新增了按照拼音、部首、筆畫排序功能
用法示例:

1. 拼音
2. SELECT * FROM TEAM ORDER BY NLSSORT(排序字段,‘NLS_SORT = SCHINESE_PINYIN_M‘)
3. 筆劃
4. SELECT * FROM TEAM ORDER BY NLSSORT(排序字段,‘NLS_SORT = SCHINESE_STROKE_M‘)
5. 部首
6. SELECT * FROM TEAM ORDER BY NLSSORT(排序字段,‘NLS_SORT = SCHINESE_RADICAL_M‘)

2、按首字母拼音查詢
首先我們要創建一個可以查詢漢字首字母的函數，函數代碼如下：

1. CREATE OR REPLACE FUNCTION F_TRANS_PINYIN_CAPITAL(P_NAME IN VARCHAR2) RETURN VARCHAR2 AS
2. V_COMPARE VARCHAR2(100);
3. V_RETURN VARCHAR2(4000);
5. FUNCTION F_NLSSORT(P_WORD IN VARCHAR2) RETURN VARCHAR2 AS
6. BEGIN
7. RETURN NLSSORT(P_WORD, ‘NLS_SORT=SCHINESE_PINYIN_M‘);
8. END;
9. BEGIN
10. FOR I IN 1..LENGTH(P_NAME) LOOP
11. V_COMPARE := F_NLSSORT(SUBSTR(P_NAME, I, 1));
12. IF V_COMPARE >= F_NLSSORT(‘ 吖 ‘) AND V_COMPARE <= F_NLSSORT(‘驁 ‘) THEN
13. V_RETURN := V_RETURN || ‘a‘;
14. ELSIF V_COMPARE >= F_NLSSORT(‘八 ‘) AND V_COMPARE <= F_NLSSORT(‘簿 ‘) THEN
15. V_RETURN := V_RETURN || ‘b‘;
16. ELSIF V_COMPARE >= F_NLSSORT(‘嚓 ‘) AND V_COMPARE <= F_NLSSORT(‘錯 ‘) THEN
17. V_RETURN := V_RETURN || ‘c‘;
18. ELSIF V_COMPARE >= F_NLSSORT(‘咑 ‘) AND V_COMPARE <= F_NLSSORT(‘鵽 ‘) THEN
19. V_RETURN := V_RETURN || ‘d‘;
20. ELSIF V_COMPARE >= F_NLSSORT(‘妸 ‘) AND V_COMPARE <= F_NLSSORT(‘樲 ‘) THEN
21. V_RETURN := V_RETURN || ‘e‘;
22. ELSIF V_COMPARE >= F_NLSSORT(‘發 ‘) AND V_COMPARE <= F_NLSSORT(‘猤 ‘) THEN
23. V_RETURN := V_RETURN || ‘f‘;
24. ELSIF V_COMPARE >= F_NLSSORT(‘旮 ‘) AND V_COMPARE <= F_NLSSORT(‘腂 ‘) THEN
25. V_RETURN := V_RETURN || ‘g‘;
26. ELSIF V_COMPARE >= F_NLSSORT(‘妎 ‘) AND V_COMPARE <= F_NLSSORT(‘夻 ‘) THEN
27. V_RETURN := V_RETURN || ‘h‘;
28. ELSIF V_COMPARE >= F_NLSSORT(‘丌 ‘) AND V_COMPARE <= F_NLSSORT(‘攈 ‘) THEN
29. V_RETURN := V_RETURN || ‘j‘;
30. ELSIF V_COMPARE >= F_NLSSORT(‘哢 ‘) AND V_COMPARE <= F_NLSSORT(‘穒 ‘) THEN
31. V_RETURN := V_RETURN || ‘k‘;
32. ELSIF V_COMPARE >= F_NLSSORT(‘垃 ‘) AND V_COMPARE <= F_NLSSORT(‘擽 ‘) THEN
33. V_RETURN := V_RETURN || ‘l‘;
34. ELSIF V_COMPARE >= F_NLSSORT(‘嘸 ‘) AND V_COMPARE <= F_NLSSORT(‘椧 ‘) THEN
35. V_RETURN := V_RETURN || ‘m‘;
36. ELSIF V_COMPARE >= F_NLSSORT(‘拏 ‘) AND V_COMPARE <= F_NLSSORT(‘瘧 ‘) THEN
37. V_RETURN := V_RETURN || ‘n‘;
38. ELSIF V_COMPARE >= F_NLSSORT(‘筽 ‘) AND V_COMPARE <= F_NLSSORT(‘漚 ‘) THEN
39. V_RETURN := V_RETURN || ‘o‘;
40. ELSIF V_COMPARE >= F_NLSSORT(‘妑 ‘) AND V_COMPARE <= F_NLSSORT(‘曝 ‘) THEN
41. V_RETURN := V_RETURN || ‘p‘;
42. ELSIF V_COMPARE >= F_NLSSORT(‘七 ‘) AND V_COMPARE <= F_NLSSORT(‘裠 ‘) THEN
43. V_RETURN := V_RETURN || ‘q‘;
44. ELSIF V_COMPARE >= F_NLSSORT(‘亽 ‘) AND V_COMPARE <= F_NLSSORT(‘鶸 ‘) THEN
45. V_RETURN := V_RETURN || ‘r‘;
46. ELSIF V_COMPARE >= F_NLSSORT(‘仨 ‘) AND V_COMPARE <= F_NLSSORT(‘蜶 ‘) THEN
47. V_RETURN := V_RETURN || ‘s‘;
48. ELSIF V_COMPARE >= F_NLSSORT(‘侤 ‘) AND V_COMPARE <= F_NLSSORT(‘籜 ‘) THEN
49. V_RETURN := V_RETURN || ‘t‘;
50. ELSIF V_COMPARE >= F_NLSSORT(‘屲 ‘) AND V_COMPARE <= F_NLSSORT(‘鶩 ‘) THEN
51. V_RETURN := V_RETURN || ‘w‘;
52. ELSIF V_COMPARE >= F_NLSSORT(‘夕 ‘) AND V_COMPARE <= F_NLSSORT(‘鑂 ‘) THEN
53. V_RETURN := V_RETURN || ‘x‘;
54. ELSIF V_COMPARE >= F_NLSSORT(‘丫 ‘) AND V_COMPARE <= F_NLSSORT(‘韻 ‘) THEN
55. V_RETURN := V_RETURN || ‘y‘;
56. ELSIF V_COMPARE >= F_NLSSORT(‘帀 ‘) AND V_COMPARE <= F_NLSSORT(‘咗 ‘) THEN
57. V_RETURN := V_RETURN || ‘z‘;
58. END IF;
59. END LOOP;
60. RETURN V_RETURN;
61. END;

隨後，我們可以先取出姓名，然後截取第一個漢字，最後取漢字的首字母，即可用來作為查詢條件

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

最後展示一個將兩個結合起來運用的例子：

1. select e.expert_id,e.eperson_id,e.fullname,
2. sum(b2i.browse_num) browsecount,
3. count(o.object_id),wmsys.wm_concat(distinct d.name)
4. from expert e
5. left join expert2disciplinetype e2d on e2d.expert_id = e.expert_id
6. left join disciplinetype d on d.discipline_type_id = e2d.discipline_type_id
7. and d.upid = ‘0‘
8. left join community2expert c2e on c2e.expert_id = e.expert_id
9. left join item i on i.item_type_id = ‘AcademicRes‘ and i.withdrawn = ‘N‘
10. join metadatavalue m on m.item_id = i.item_id and m.metadata_field_id = ‘64‘
11. join objectpublishinfo o on o.object_id = i.item_id
12. and o.object_type = ‘item‘ and o.viewobjecttype = ‘eperson‘
13. and o.viewobjectid = e.eperson_id and o.state = ‘1‘
14. and o.publishstate_id = ‘3‘
15. left join browse2item b2i on b2i.item_id = o.object_id
16. where e.state = ‘1‘ and f_trans_pinyin_capital(substr(e.fullname,1,1)) = ‘z‘
17. group by e.expert_id,e.eperson_id,e.fullname
18. order by nlssort(e.fullname,‘NLS_SORT=SCHINESE_PINYIN_M‘)

關於java中實現在oracle數據庫中實現對中文首字母進行排序的解決方案