lei_y601/sqlbot_agent
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

修复枚举对应错误，增加问答示例，优化提示词

Browse Source
This commit is contained in:
yujj128
2025-10-16 16:02:46 +08:00
parent 62fbc0014e
commit 1b4be9a907
3 changed files with 51 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
template.yaml
View File

@@ -56,7 +56,7 @@ template:
</rule>
<rule>
<rule-title>数据查询与排序</rule-title>
<rule-detail>若未明确指定查询字段，涉及人员信息时，默认返回相关性最强的前10个字段。</rule-detail>
<rule-detail>若未明确指定查询字段，涉及人员信息时，一般返回相关性最强的前10个字段。</rule-detail>
<rule-detail>当涉及部门表org_orgs的查询时注意enable启用状态和dr删除标志，且必须限制code字段值包含'CYJ'。</rule-detail>
<rule-detail>若查询字段为 VARCHAR 或 TEXT 类型但需要计算，必须先进行合理的类型转换（如 CAST(... AS NUMERIC)）。</rule-detail>
<rule-detail>若查询包含日期/时间字段：
@@ -67,6 +67,7 @@ template:
<rule>
<rule-title>聚合与计算</rule-title>
<rule-detail>使用了聚合函数（如 COUNT(), SUM(), AVG()）的SQL，必须配置相应的 GROUP BY 子句。</rule-detail>
<rule-detail>禁止使用AGE函数表达式计算年龄</rule-detail>
<rule-detail>使用了函数（如 COUNT(), CAST(), SUM()）的字段，必须为其指定一个英文别名。</rule-detail>
<rule-detail>计算占比或百分比时，结果保留两位小数，并以 '%' 符号结尾。示例：ROUND(COUNT(*) * 100.0 / (SELECT COUNT(*) FROM table), 2) || '%' (PostgreSQL语法)</rule-detail>
<rule-detail>若查询结果包含枚举字段（如 gender=1,2），必须使用 CASE WHEN 语句将其转换为可读的标签。示例: SELECT CASE WHEN "gender" = '1' THEN '男' WHEN "gender" = '2' THEN '女' END AS "gender"</rule-detail>

35
util/q_and_a_dict.py
View File

@@ -386,6 +386,41 @@ question_and_answer = [
AND p."dr" = 0
) AS "女员工数"
'''
},
{
"question": "数信中心今天各个处室在岗人员数量",
"answer": '''
SELECT o.name AS "处室名称", COUNT(p.id) AS "人数"
FROM YJOA_APPSERVICE_DB.t_pr3rl2oj_yj_person_database p
JOIN IUAP_APDOC_BASEDOC.org_orgs o ON p.internal_dept = o.id
WHERE p.internal_dept IN (
SELECT "id" FROM "IUAP_APDOC_BASEDOC"."org_orgs"
START WITH "name"||"shortname" LIKE '%数信中心%' AND "dr"=0 AND "enable"=1 AND "code" LIKE '%CYJ%'
CONNECT BY PRIOR "id" = "parentid"
)
AND p.id IN (SELECT person_id FROM "YJOA_APPSERVICE_DB"."t_yj_person_status" WHERE status = '1001' AND date_value = TO_CHAR(SYSDATE, 'yyyy-MM-dd') AND dr = 0)
AND p.dr = 0
GROUP BY o.name
ORDER BY "人数" DESC
LIMIT 1000
'''
},
{
"question": "数信中心下的综合管理处有多少人",
"answer": '''
SELECT o.name AS "处室名称", COUNT(p.id) AS "人数"
FROM YJOA_APPSERVICE_DB.t_pr3rl2oj_yj_person_database p
JOIN IUAP_APDOC_BASEDOC.org_orgs o ON p.internal_dept = o.id
WHERE p.internal_dept IN (
SELECT "id" FROM "IUAP_APDOC_BASEDOC"."org_orgs"
START WITH "name"||"shortname" LIKE '%数信中心%' AND "dr"=0 AND "enable"=1 AND "code" LIKE '%CYJ%'
CONNECT BY PRIOR "id" = "parentid"
)
AND p.dr = 0 AND o.name LIKE '%综合管理处%'
GROUP BY o.name
ORDER BY "人数" DESC
LIMIT 1000
'''
}

18
util/train_ddl.py
View File

@@ -3,9 +3,15 @@ train_document='''
查询地址，籍贯，公司，单位时，尽量使用like查询;
查询人员信息时,由于数据表字段过多。只需要展示人员关键信息字段，id，工号，姓名，单位以及用户问题中需要查询的字段;
表字段信息以及字段枚举信息在values下，注意相关字段枚举值的转换;
出生日期字段是varchar类型，计算年龄时需转换成合适的格式
查询单位时，通过orgs表查询，且需要基于parentID查询递归查询单位下的子单位；
查询内部单位时，则可以直接查询人员信息表，通过like模糊查询;
没有明确说明查询外部单位，都默认查询通过orgs查询单位;
数信中心和数信部并非同一个部门，注意区分，不要混淆
数信中心下还有多个部门，这些部门里的人员也隶属于数信中心
根据部门名称查询部门时除了全称name，简称shortname也重要信息
数信部是简称
部门表org_orgs中的level字段 {1：公司，2：一级部门，3：二级部门.....}
'''
person_database_ddl = """
@@ -896,8 +902,12 @@ person_ac_area = '''
"name": "area",
"type": "Int",
"comment": "区域位置",
"value": {
"1":"党校",
"2":"凯莱"
}
"role": "dimension",
"tags": ["门禁所属区域"]
"tags": ["门禁所属区域","枚举"]
},
{
"name": "region",
@@ -907,14 +917,14 @@ person_ac_area = '''
"1":"北京",
"2":"成都",
"3":"秭归",
"4":"林芝市区",
"5":"拉萨",
"4":"拉萨",
"5":"林芝",
"6":"米林",
"7":"派镇",
"8":"墨脱",
},
"role": "dimension",
"tags": ["门禁所属地区"]
"tags": ["门禁所属地区","枚举"]
},
],