Question

我有一个名为sector的字段表，每个扇区通常都是1,2,3,4,5,6,7等。

我想在应用程序中显示可用的扇区，我认为显示所有1,2,3,4,5,6,7都是愚蠢的，所以我应该显示“1到7”。

问题在于，有时扇区跳过一个数字，如1,2,3,5,6,7。所以我想展示1到3,5到7之类的东西。

如何在sql中查询此内容以显示在我的应用程序中？

Answer 1

这在sql中称为“Gaps”。这是一篇详细的文章"Article"

Answer 2

某些DBMS可能具有一些OLAP功能，可以轻松编写此类查询，但IBM Informix Dynamic Server（IDS）尚不具备此类功能。

为了具体起见，我们假设您的表名为'ProductSectors'并且具有如下结构：

CREATE TABLE ProductSectors
(
    ProductID    INTEGER NOT NULL,
    Sector       INTEGER NOT NULL CHECK (Sector > 0),
    Name         VARCHAR(20) NOT NULL,
    PRIMARY KEY (ProductID, Sector)
);

您在特定ProductID中寻找的是Sector的最小和最大连续值列表。当没有小于最小值的值且没有大于最大值的值并且该范围内没有间隙时，范围是连续的。这是一个复杂的查询：

SELECT P1.ProductID, P1.Sector AS Min_Sector, P2.Sector AS Max_Sector
  FROM ProductSectors P1 JOIN ProductSectors P2
    ON P1.ProductID = P2.ProductID
   AND P1.Sector   <= P2.Sector
 WHERE NOT EXISTS (SELECT *     -- no entry one smaller
                     FROM ProductSectors  P6
                    WHERE P1.ProductID  = P6.ProductID
                      AND P1.Sector - 1 = P6.Sector
                  )
   AND NOT EXISTS (SELECT *     -- no entry one larger
                     FROM ProductSectors  P5
                    WHERE P2.ProductID  = P5.ProductID
                      AND P2.Sector + 1 = P5.Sector
                  )
   AND NOT EXISTS (SELECT *     -- no gaps between P1.Sector and P2.Sector
                     FROM ProductSectors P3
                    WHERE P1.ProductID = P3.ProductID
                      AND P1.Sector   <= P3.Sector
                      AND P2.Sector   >  P3.Sector
                      AND NOT EXISTS (SELECT *
                                        FROM ProductSectors P4
                                       WHERE P4.ProductID = P3.ProductID
                                         AND P4.Sector    = P3.Sector + 1
                                     )
                  )
 ORDER BY P1.ProductID, Min_Sector;

以下是使用示例数据的整体查询的跟踪：

CREATE TEMP TABLE productsectors
(
    ProductID   INTEGER NOT NULL,
    Sector      INTEGER NOT NULL CHECK(Sector > 0),
    Name        VARCHAR(20),
    PRIMARY KEY (ProductID, Sector)
);

一些样本数据，有各种差距：

INSERT INTO ProductSectors VALUES(101, 1, "101:1");
INSERT INTO ProductSectors VALUES(101, 2, "101:2");
INSERT INTO ProductSectors VALUES(101, 3, "101:3");
INSERT INTO ProductSectors VALUES(101, 4, "101:4");
INSERT INTO ProductSectors VALUES(101, 5, "101:5");
INSERT INTO ProductSectors VALUES(101, 6, "101:6");
INSERT INTO ProductSectors VALUES(101, 7, "101:7");
INSERT INTO ProductSectors VALUES(102, 1, "102:1");
INSERT INTO ProductSectors VALUES(102, 2, "102:2");
INSERT INTO ProductSectors VALUES(102, 4, "102:4");
INSERT INTO ProductSectors VALUES(102, 5, "102:5");
INSERT INTO ProductSectors VALUES(102, 6, "102:6");
INSERT INTO ProductSectors VALUES(102, 7, "102:7");
INSERT INTO ProductSectors VALUES(103, 1, "103:1");
INSERT INTO ProductSectors VALUES(103, 2, "103:2");
INSERT INTO ProductSectors VALUES(103, 4, "103:4");
INSERT INTO ProductSectors VALUES(103, 6, "103:6");
INSERT INTO ProductSectors VALUES(103, 7, "103:7");
INSERT INTO ProductSectors VALUES(104, 1, "104:1");
INSERT INTO ProductSectors VALUES(104, 2, "104:2");
INSERT INTO ProductSectors VALUES(104, 3, "104:3");
INSERT INTO ProductSectors VALUES(104, 6, "104:6");
INSERT INTO ProductSectors VALUES(104, 7, "104:7");
INSERT INTO ProductSectors VALUES(105, 1, "105:1");
INSERT INTO ProductSectors VALUES(105, 4, "105:4");
INSERT INTO ProductSectors VALUES(105, 5, "105:5");
INSERT INTO ProductSectors VALUES(105, 7, "105:7");
INSERT INTO ProductSectors VALUES(106, 1, "106:1");
INSERT INTO ProductSectors VALUES(106, 2, "106:1");
INSERT INTO ProductSectors VALUES(106, 3, "106:1");
INSERT INTO ProductSectors VALUES(106, 7, "106:7");
INSERT INTO ProductSectors VALUES(107, 7, "107:7");
INSERT INTO ProductSectors VALUES(108, 8, "108:8");
INSERT INTO ProductSectors VALUES(108, 9, "108:9");

必需输出 - 也是实际输出：

101|1|7 
102|1|2 
102|4|7 
103|1|2 
103|4|4 
103|6|7 
104|1|3 
104|6|7 
105|1|1 
105|4|5 
105|7|7 
106|1|3 
106|7|7 
107|7|7 
108|8|9

预期结果为MacOS X 10.6.2，IDS 11.50.FC4W1，SQLCMD 86.04。

Answer 3

好的，我一直在寻找更深层次，找到了this

它有效:)，希望它可以帮助我帮助我。

如何获取sql中的序列号范围？

3 个答案: