來源：http://bbs.csdn.net/topics/390611005 。
系統採集的JSON格式數據（s.json）如下：
{
“SUCCESS”: [
{
"MESSAGE": "IMEI Service List",
"LIST": {
"MOVISTAR SPAIN": {
"GROUPNAME": "MOVISTAR SPAIN",
"SERVICES": {
"3": {
"SERVICEID": 32,
"SERVICENAME": "MOVISTAR NOKIA INSTANTE",
"CREDIT": 4,
"TIME": "1-30 Minutes",
"INFO": "<p style=\"text-align: center;\">…… </p>",
"Requires.Network": "None",
"Requires.Mobile": "None",
"Requires.Provider": "None",
"Requires.PIN": "None",
"Requires.KBH": "None",
"Requires.MEP": "None",
"Requires.PRD": "None",
"Requires.Type": "None",
"Requires.Locks": "None",
"Requires.Reference": "None"
},
"8": {
"SERVICEID": 77,
"SERVICENAME": "MOVISTAR NOKIA 20 NCK",
"CREDIT": 12,
"TIME": "1-30 Minutes",
"INFO": "<p style=\"text-align: center;\">……</p>",
"Requires.Network": "None",
"Requires.Mobile": "None",
"Requires.Provider": "None",
"Requires.PIN": "None",
"Requires.KBH": "None",
"Requires.MEP": "None",
"Requires.PRD": "None",
"Requires.Type": "None",
"Requires.Locks": "None",
"Requires.Reference": "None"
}
}
},
"VODAFONE SPAIN": {
"GROUPNAME": "VODAFONE SPAIN",
"SERVICES": {
"5": {
"SERVICEID": 50,
"SERVICENAME": "VODAFONE NOKIA BB5 SL3",
"CREDIT": 5,
"TIME": "1-60 Minutes",
"INFO": "<p style=\"text-align: center;\">……</p>",
"Requires.Network": "None",
"Requires.Mobile": "None",
"Requires.Provider": "None",
"Requires.PIN": "None",
"Requires.KBH": "None",
"Requires.MEP": "None",
"Requires.PRD": "None",
"Requires.Type": "None",
"Requires.Locks": "None",
"Requires.Reference": "None"
},
"10": {
"SERVICEID": 95,
"SERVICENAME": "VODAFONE SONY&;SONY ERIC(RAPIDO)",
"CREDIT": 16,
"TIME": "1-24 Hours",
"INFO": "<p style=\"text-align: center;\">……</p>",
"Requires.Network": "None",
"Requires.Mobile": "None",
"Requires.Provider": "None",
"Requires.PIN": "None",
"Requires.KBH": "None",
"Requires.MEP": "None",
"Requires.PRD": "None",
"Requires.Type": "None",
"Requires.Locks": "None",
"Requires.Reference": "None"
}
}
}
}
}
],
“apiversion”: “2.0.0″
}
現需要按照JSON數據內容，將相應節點下的屬性值更新到數據庫的對應表中，需要更新的2張表如下：
Create table [dbo].[Groups]
(
[ID] [int] IDENTITY(1,1) NOT NULL, –id
[Groupname] [nvarchar] (50) not null default(”), –名稱
[groupid] [int] not null default(0),
CONSTRAINT [PK_Groups_id] PRIMARY KEY CLUSTERED
(
[id] ASC
)WITH (IGNORE_DUP_KEY = OFF) ON [PRIMARY]
) ON [PRIMARY]
CREATE TABLE [dbo].[Services](
[id] [int] IDENTITY(1,1) NOT NULL, –id
[Serviceid] [int] not null default(0),
[Servicename] [nvarchar] (50) not null default(”),
[groupid] [int] not null default(0),
[Credit] [decimal] not null default(0.00),
[Time] [nvarchar] (50) not null default(”),
[INFO] [nvarchar] (3000) not null default(”),
[Network] [nvarchar] (100) not null default(‘none’),
[Mobile] [nvarchar] (100) not null default(‘none’),
[Provider] [nvarchar] (100) not null default(‘none’),
[PIN] [nvarchar] (100) not null default(‘none’),
[KBH] [nvarchar] (100) not null default(‘none’),
[MEP] [nvarchar] (100) not null default(‘none’),
[PRD] [nvarchar] (100) not null default(‘none’),
[Type] [nvarchar] (100) not null default(‘none’),
[Locks] [nvarchar] (100) not null default(‘none’),
[Reference] [nvarchar] (100) not null default(‘none’),
[isstatus] [nvarchar] (1) not null default(’0′),
[remark] [nvarchar] (255) not null default(”),
[Pricingid] [int] not null default(0),
CONSTRAINT [PK_Services_id] PRIMARY KEY CLUSTERED
(
[id] ASC
)WITH (IGNORE_DUP_KEY = OFF) ON [PRIMARY]
) ON [PRIMARY]
其中SERVICES下屬性名即爲groupid，如3,5,8,10，剩餘字段與屬性名一一對應。由於LIST和SERVICES下的屬性名不固定，給解析帶來了難度。區別於一般高級語言，集算器支持動態數據結構和集合運算，實現並不困難，腳本如下：
A1：以字符串讀入JSON文件內容，並使用import@j()將字符串轉爲行列式帶有層次的序表。
A2-A3：根據兩個目標表，創建空序表，用於存儲解析結果後一次更新到數據庫。
A4-B4：循環序表，在B4中計算LIST下節點數量。
B5-C6：循環獲取LIST節點內容，並在C6中計算SERVICES下節點數量。
C7-D8：循環獲取SERVICES節點下屬性名和屬性值。
D9-D10：將解析後結果分別寫回到A2A3中的空序表中。
A11：以groupid爲主鍵，將A2內容更新到groups表中。
A12：以Serviceid爲主鍵，將A3內容更新到services表中。
