深度剖析Duwamish 7.0 (1--數(shù)據(jù)結(jié)構(gòu))
2024-07-21 02:25:27
供稿：網(wǎng)友
不好意思啊，讓大家久等了，第一次寫這種文章，可能沒有開心那么專業(yè)，廢話少說，我們開始：
1．結(jié)構(gòu)概述
duwamish 7.0 結(jié)構(gòu)分為四個(gè)邏輯層：
web 層
web 層為客戶端提供對(duì)應(yīng)用程序的訪問。這一層是作為 duwamish.sln 解決方案文件中的 web 項(xiàng)目實(shí)現(xiàn)的。web 層由 asp.net web 窗體和代碼隱藏文件組成。web 窗體只是用 html 提供用戶操作，而代碼隱藏文件實(shí)現(xiàn)各種控件的事件處理。
業(yè)務(wù)外觀層
業(yè)務(wù)外觀層為 web 層提供處理帳戶、類別瀏覽和購書的界面。這一層是作為 duwamish.sln 解決方案文件中的 businessfacade 項(xiàng)目實(shí)現(xiàn)的。業(yè)務(wù)外觀層用作隔離層，它將用戶界面與各種業(yè)務(wù)功能的實(shí)現(xiàn)隔離開來。除了低級(jí)系統(tǒng)和支持功能之外，對(duì)數(shù)據(jù)庫服務(wù)器的所有調(diào)用都是通過此程序集進(jìn)行的。
業(yè)務(wù)規(guī)則層
業(yè)務(wù)規(guī)則層是作為 duwamish.sln 解決方案文件中的 businessrules 項(xiàng)目實(shí)現(xiàn)的，它包含各種業(yè)務(wù)規(guī)則和邏輯的實(shí)現(xiàn)。業(yè)務(wù)規(guī)則完成如客戶帳戶和書籍訂單的驗(yàn)證這樣的任務(wù)。
數(shù)據(jù)訪問層
數(shù)據(jù)訪問層為業(yè)務(wù)規(guī)則層提供數(shù)據(jù)服務(wù)。這一層是作為 duwamish.sln 解決方案文件中的 dataaccess 項(xiàng)目實(shí)現(xiàn)的。
在這里，對(duì)于duwamish 7.0的分布式結(jié)構(gòu)我就不再羅嗦了，msdn寫的絕對(duì)比我好..
下面就從數(shù)據(jù)訪問層開始解剖,我單獨(dú)提出customer這一段進(jìn)行程序講解：
1.customerdata.cs（數(shù)據(jù)層）
code:
//----------------------------------------------------------------
// copyright (c) 2000-2001 microsoft corporation
// all rights reserved.
//
// this source code is intended only as a supplement to microsoft
// development tools and/or on-line documentation. see these other
// materials for detailed information regarding microsoft code samples.
//
// this code and information is provided "as is" without warranty
// of any kind, either expressed or implied, including but not
// limited to the implied warranties of merchantability and/or
// fitness for a particular purpose.
//----------------------------------------------------------------
namespace duwamish7.common.data
{
using system;
using system.data;
using system.runtime.serialization;
/// <summary>
/// a custom serializable dataset containing customer information.
/// <remarks>
/// this class is used to define the shape of customerdata.
/// </remarks>
/// <remarks>
/// the serializale constructor allows objects of type customerdata to be remoted.
/// </remarks>
/// </summary>
[serializableattribute]
public class customerdata : dataset
{
//
//customer constants
//
/// <value>the constant used for customers table. </value>
public const string customers_table = "customers";
/// <value>the constant used for email field in the customers table. </value>
public const string email_field = "email";
/// <value>the constant used for name field in the customers table. </value>
public const string name_field = "name";
/// <value>the constant used for address field in the customers table. </value>
public const string address_field = "address";
/// <value>the constant used for country field in the customers table. </value>
public const string country_field = "country";
/// <value>the constant used for password field in the customers table. </value>
public const string password_field = "password";
/// <value>the constant used for phonenumber field in the customers table. </value>
public const string phone_field = "phonenumber";
/// <value>the constant used for fax field in the customers table. </value>
public const string fax_field = "fax";
/// <value>the constant used for pkid field in the customers table. </value>
public const string pkid_field = "pkid";
//
// error messages
//
/// <value>the constant used for row error when 'email not unique' field in customerdata. </value>
public const string email_field_not_unique = "email not unique";
/// <value>the constant used for row error when 'email invalid format' field in customerdata. </value>
public const string email_field_invalid_format = "email invalid format";
/// <value>the constant used for row error when there is an 'invalid field' in customerdata. </value>
public const string invalid_field = "invalid field";
/// <value>the constant used for error when 'invalid fields' exist in customerdata. </value>
public const string invalid_fields = "invalid fields";
/// <summary>
/// constructor to support serialization.
/// <remarks>constructor that supports serialization.</remarks>
/// <param name="info">the serializationinfo object to read from.</param>
/// <param name="context">information on who is calling this method.</param>
/// </summary>
public customerdata(serializationinfo info, streamingcontext context) : base(info, context)
{
}
/// <summary>
/// constructor for customerdata.
/// <remarks>initialize a customerdata instance by building the table schema.</remarks>
/// </summary>
public customerdata()
{
//
// create the tables in the dataset
//
builddatatables();
}
//----------------------------------------------------------------
// sub builddatatables:
// creates the following datatables: customers
//----------------------------------------------------------------
private void builddatatables()
{
//
// create the customers table
//
datatable table = new datatable(customers_table);
datacolumncollection columns = table.columns;
datacolumn column = columns.add(pkid_field, typeof(system.int32));
column.allowdbnull = false;
column.autoincrement = true;
columns.add(email_field, typeof(system.string));
columns.add(password_field, typeof(system.string));
columns.add(name_field, typeof(system.string));
columns.add(address_field, typeof(system.string)).allowdbnull= false;
columns.add(country_field, typeof(system.string)).allowdbnull= false;
columns.add(phone_field, typeof(system.string)).allowdbnull= false;
columns.add(fax_field, typeof(system.string));
this.tables.add(table);
}
} //class customerdata
} //namespace duwamish7.common.data
大家可以看到，這里的customerdata類繼承與dataset，詳細(xì)定義字段名稱以及有可能產(chǎn)生的出錯(cuò)信息，
并同時(shí)繪制出一張空表格。
ok,這里的代碼非常簡(jiǎn)單，大家一看應(yīng)該就可以明白。代碼的技巧性非常強(qiáng)，以后的數(shù)據(jù)訪問、交換、傳遞全部通過這里進(jìn)行。
私底下認(rèn)為，采用這種方法有利于團(tuán)隊(duì)協(xié)作，代碼編輯也很簡(jiǎn)單。
但是，微軟自己也說過dataset的效率沒有datareader高，而dataset的靈活性遠(yuǎn)遠(yuǎn)超過于datareader，真搞不懂ms在企業(yè)事例中為什么采用dataset，也許只是讓我們學(xué)習(xí)其中的思想吧。
過兩天再和大家談?wù)劻硪环N方法，采用datareader，效率高點(diǎn)，但是代碼太難看，難懂：（
