《计算机 软件工程 外文文献 外文翻译 英文文献 数据库开发过程.doc》由会员分享，可在线阅读，更多相关《计算机 软件工程 外文文献 外文翻译 英文文献 数据库开发过程.doc（9页珍藏版）》请在taowenge.com淘文阁网|工程机械CAD图纸|机械工程制图|CAD装配图下载|SolidWorks_CaTia_CAD_UG_PROE_设计图分享下载上搜索。
1、英文原文版出处： A.Ghatak,K.Thyagaraian,Comtemperary Optic,Plenum Press,1978 译文：数据库开发过程 基于信息工程的信息系统规划是数据库开发项目的一个来源。这些开发新数据库的项目通常是为了满足组织的战略需求，例如改善客户支持、提高产品和库存管理或进行更精确的销售预测。然而许多数据库开发项目更多的是以自底向上的方式出现的，例如信息系统的用户需要特定的信息来完成他们的工作，从而请求开始一个项目，又如其他信息系统的专家发现组织需要改进数据管理而开始新的项目。即使在自底向上的情况下，建立企业数据模型也是必须的，以便理解现有的数据库是否可以提供所
2、需的数据，否则，新的数据库、数据实体和属性都应该加到当前的组织数据资源中去。 无论是战略需求还是操作信息的需求，每个数据库开发项目通常集中在一个数据库上。一些数据库项目仅仅集中在定义、设计和实现一个数据库，以作为后续信息系统开发的基础。然而在大多数情况下，数据库及其相关信息处理功能是作为一个完整的信息系统开发项目的一部分而被开发的。一. 系统开发生命周期 指导管理信息系统开发项目的传统过程是系统开发生命周期(SDLC)。系统开发生命周期是指一个组织中由数据库设计人员和程序员组成的信息系统专家小组详细说明、开发、维护和替换信息系统的全部步骤。这个过程比作瀑布是因为每一步都流到相邻的下一步，即信息
3、系统的规格说明是一块一块地开发出来的，每一块的输出是下一块的输入。然而如图所示，这些步骤并不是纯线性的，每个步骤在时间上有所重叠（因此可以并行地管理步骤），而且当需要重新考虑先前的决策时，还可以回滚到前面某些步骤。（因而水可以在瀑布中倒流！）系统开发生命周期的每一阶段都包括与数据库开发相关的活动，所以，数据库管理的问题遍布整个系统开发过程。请注意，系统开发生命周期的阶段和数据库开发步骤之间不存在一一对应的关系，概念数据建模发生在两个系统开发生命周期阶段之间。1.1 企业建模 数据库开发过程从企业建模（系统开发生命周期中项目论证和选择阶段的一部分）开始，设定组织数据库的范围和一般内容。企业建模发
4、生在信息系统规划和其他活动期间，这些活动确定信息系统的哪个部分需要改变和加强并概述出全部组织数据的范围。在这一步中，检查当前数据库和信息系统，分析作为开发项目主体的业务领域的本质，用非常一般的术语描述每个信息系统在开发时所需要的数据。每个项目只有当它达到组织的预期目标时才可以进行下一步。1.2 概念数据建模 对一个已经开始的信息系统项目而言，概念数据建模阶段分析信息系统的全部数据需求。它分为两个阶段。首先，它在项目开始和规划阶段建立一张图。同时建立其他文档来概述不考虑现存数据库的情况下特定开发项目中所需的数据范围。此时仅仅包括高层类别的数据（实体）和主要联系。然后在系统开发生命周期的分析阶段产
5、生确定信息系统必须管理的全部组织数据的详细数据模型，定义所有数据属性，列出全部数据类别，表示数据实体间所有的业务联系，确定描述数据完整性的全部规则。在分析阶段，还要检查概念数据模型（在后面也称作概念模式）与用来解释目标信息系统其他方面的模型类别的一致性，例如处理步骤、处理数据的规则以及时间的时序。然而，即使是这样详细的概念数据模型也只是初步的，因为后续的信息系统生命周期中的活动在设计事务、报表、显示和查询时可能会发现遗漏的元素或错误。因此，经常说到的概念数据建模是以一种自顶向下的方式完成的，它由业务领域的一般理解所驱动，而不是由特定的信息处理活动所驱动。1.3 逻辑数据库设计 逻辑数据库设计从
6、两个角度进行数据库开发。首先，将概念数据模型变换成基于关系数据库理论的标准表示方法关系。然后像设计信息系统的每个计算机程序（包括程序的输入和输出格式）那样，对数据库支持的事务、报表、显示和查询进行详细的检查。在这个所谓的自底向上的分析中，精确地验证数据库中需要维护的数据和在每个事务、报表等等中需要的那些数据的性质。 对于每个单独的报表、事务等等的分析都要考虑一个特定的、有限制的但是完全的数据库视图。当报表、事务等被分析时有可能根据需要而改变概念数据模型。尤其在大型的项目中，不同的分析人员和系统开发者的团队可以独立地工作在不同的程序或程序集中，他们所有工作的细节直到逻辑设计阶段才可能会显示出来。
7、在这种情况下，逻辑数据库设计阶段必须将原始的概念数据模型和这些独立的用户视图合并或集成到一个全面的设计中。在进行逻辑信息系统设计时也可以确定额外的信息处理需求，此时这些新的需求必须集成到前面确定的逻辑数据库设计中。 逻辑数据库设计的最后一步是根据为生成结构良好的数据规格说明而确定的规则，将组合的、协商后的数据规格说明转换成基本的或原子的元素。对当今的大部分数据库而言，这些规则来自关系数据库理论和称作规范化的过程。这一步的结果是产生管理这些数据的、不引用任何数据库管理系统的完整的数据库描述图。在完成逻辑数据库设计后，开始确定详细的计算机程序的逻辑和维护、报告数据库内容所需的查询。1.4 物理数据
8、库设计和定义 物理数据库设计和定义阶段决定计算机存储器（通常是磁盘）中数据库的组织，定义数据库管理系统的物理结构，概述处理事务的程序，产生期望的管理信息和决策支持的报表。本阶段的目标是设计能够有效、安全地管理所有数据处理的数据库，因此物理数据库设计需紧密结合物理信息系统其他方面的设计，包括程序、计算机硬件、操作系统和数据通信网络。1.5 数据库实现 数据库实现阶段编写、测试和安装处理数据库的程序。设计人员可以使用标准的编程语言（如COBOL、C或Visual Basic）、专用的数据库处理语言（如SQL），或专用的非过程化语言来编程，以产生固定格式的报表、显示结果，可能还包括图表。在实现阶段，
9、还要完成所有的数据库文档，培训用户，为信息系统（和数据库）的用户安装程序。最后一步是利用现存的信息源（遗留应用中的文件和数据库以及现在需要的新数据）加载数据。加载数据的第一步经常是将数据从现存的文件和数据库中转到一种中间的格式（如二进制或文本文件），然后再将这些中间数据加载到新的数据库中。最后，运行数据库以及相关的应用以供实际的用户维护和检索数据。在运转期间，定期备份数据库，并当数据库损坏或受到影响时恢复数据库。1.6 数据库维护 数据库在数据库维护期间逐渐发展。在这一步，为了满足变化的业务条件，为了改正数据库设计的错误，或数据库应用的处理速度而增加、删除或改变数据库的结构特征。当一个程序或计
10、算机发生故障而使数据库受到影响或损坏时也可能应该重建数据库。这一步通常是数据库开发过程中最长的一步，因为它持续数据库及相关应用的整个生命周期，每次数据库的发展都可看作一个简略的数据库开发过程，其中会出现概念数据建模、逻辑和物理数据库设计以及数据库实现以处理提出的变化。二. 信息系统开发的其他方法 系统开发生命周期法或其稍作变化的变体经常用于指导信息系统和数据库的开发。信息系统生命周期是一种方法学，它是高度结构化的方法，它包括许多检查和权衡以确保每一步产生精确的结果，而且新的或替代的信息系统与它必须通信的或数据定义需要一致的现存系统保持一致。系统开发生命周期法经常由于产生一个工作系统需要很长的时
11、间而受到批评，因为工作系统仅仅在整个过程结束时才产生。现在组织越来越多的使用快速应用开发法，它是一个包含分析、设计和实现步骤的快速重复的迭代过程，直到汇聚到用户所需的系统为止。快速应用开发法在所需的数据库已经存在、增强系统主要是为了检索数据的应用中适用，而不适用于那些生成和修改数据库的应用。 使用最广泛的快速应用开发法之一是原型法。原型法是一个系统开发的迭代过程，通过分析员和用户的紧密配合，持续地修改系统而最终将所有需求转换成一个工作系统。在此图中我们包含了注释，概略地描述了每个原型法阶段的数据库开发活动。一般来说，当信息系统的问题被确定时，仅仅粗略地尝试概念数据建模。在开发最初的原型时，设计
12、用户想要的显示和报表，同时理解任何新的数据库需求并定义一个用语原型的数据库。这通常是一个新的数据库，它复制现存系统的一部分，还可能增加了一些新的内容。当需要新的内容时，这些内容通常来自外部数据源，如市场研究数据、一般的经济指标或行业标准。 当产生原型新的版本时重复数据库的实现和维护活动。通常仅进行最低限度的安全性和完整性控制，因为此时的重点是尽可能快地产生可以使用的原型版本。而且文档管理也延迟到项目的最后，在交付使用时才进行用户培训。最后，一旦构建了一个可接受的原型，开发者和用户将决定最后的原型和数据库是否能交付使用。如果系统（包括数据库）效率很低，那么系统和数据库将被重新编程和重新组织以达到
13、期望的性能。随着可视化编程工具（如Visual Basic、Java、Visual C+和第四代语言）越来越流行，利用可视化编程工具可以很方便地修改用户与系统间的界面，原型法正成为可供选择的系统开发方法。使用原型法改变用户报表和显示的内容和布局是相当容易的。在这个过程中，新的数据库需求被确定，因此被发展中的应用使用的现存的数据库应该进行修改。甚至有可能为一个需要新的数据库的系统使用原型法，在这种情况下，当系统需求在迭代的开发过程中不断变化时需要获取样本数据以建造或重建数据库原型。外文原文Database Development ProcessBased on information engin
14、eering information systems planning database is a source of development projects. These new database development projects is usually in order to meet the strategic needs of organizations, such as improving customer support, improve product and inventory management, or a more accurate sales forecast.
15、 However, many more database development project is the bottom-up approach emerging, such as information system user needs specific information to complete their work, thus beginning a project request, and as other information systems experts found that organizations need to improve data management
16、and begin new projects. Bottom-up even in the circumstances, to set up an enterprise data model is also necessary to understand the existing database can provide the necessary data, otherwise, the new database, data entities and attributes can be added to the current data resources to the organizati
17、on.Both the strategic needs or operational information needs of each database development projects normally concentrated in a database. Some projects only concentrated in the database definition, design and implementation of a database, as a follow-up to the basis of the development of information s
18、ystems. However, in most cases, the database and associated information processing function as a complete information systems development project was part of the development. one. System Development Life CycleGuide management information system development projects is the traditional process of syst
19、em development life cycle (SDLC). System development life cycle is an organization of the database designers and programmers information system composed of the Panel of Experts detailed description, development, maintenance and replacement of the entire information system steps. This process is beca
20、use Waterfall than for every step into the adjacent the next step, that is, the information system is a specification developed by a piece of land, every piece of the output is under an input. However shown in the figure, these steps are not purely linear, each of the steps overlap in time (and thus
21、 can manage parallel steps), but when the need to reconsider previous decisions, but also to roll back some steps ahead. (And therefore water can be put back in the waterfall!) The system development life cycle including each stage and database development-related activities, therefore, the question
22、 of database management systems throughout the entire development process. In Figure 5 we repeat of the system development life cycle stage of the seven, and outlines the common database at each stage of development activities. Please note that the systems development life cycle stages and database
23、development steps exists between the relationship between the concept of modeling data in both systems development life cycle stages between.1.1 Enterprise Modeling Database development process from the enterprise modeling (system development life cycle stage of the project feasibility studies, and
24、to choose a part)Organizations set the scope and general database content. Enterprise modeling in information systems planning and other activities, these activities determine which part of information systems need to change and strengthen the entire organization and outlines the scope of data. In t
25、his step, check the current database and information systems, development of the project as the main areas of the nature of the business, with a very general description of each term in the development of information systems when needed data. Each item only when it achieved the expected goals of org
26、anizations can be when the next step.1.2 Conceptual Data Modeling One has already begun on the Information System project, the concept of data modeling phase of the information systems needs of all the data. It is divided into two stages. First, it began the project in the planning stage. At the sam
27、e time outlining the establishment of other documents to the existing database without considering the circumstances specific development projects in the scope of the required data. This category only includes high-level data (entities), and main contact. Then in the system development life-cycle an
28、alysis stage must have a management information system set the entire organization Details of the data model definition of all data attributes, listing all data types that all data inter-entity business linkages, defining description of the full data integrity rules. In the analysis phase, but also
29、the concept of inspection data model (also called the concept behind the model) and the goal of information systems used to explain other aspects of the model of consistency categories, such as processing steps, rules and data processing time of timing. However, even if the concept is such detailed
30、data model is only preliminary, because follow-up information system life cycle activities in the design of services, statements, display and inquiries may find that missing element or mistakes. Therefore, the concept of data often said that modeling is a top-down manner, its areas of operation from
31、 the general understanding of the driver, rather than the specific information processing activities by the driver. 1.3 Logical Database Design Logical database design from two perspectives database development. First, the concept of data model transform into relational database theory based on the
32、criteria that means - between. Then, as the design of information systems, every computer procedures (including procedures for the input and output format), database support services, statements, and inquiries revealed that a detailed examination. In this so-called Bottom-up analysis, accurate verif
33、ication of the need to maintain the database and the data in each affairs, statements and so on the needs of those in the nature of the data. For each separate statements, services, and so on the analysis must take into account a specific, limited but complete database view. When statements, service
34、s, and other analysis might be necessary to change the concept of data model. Especially in large-scale projects, the different analytical systems development staff and the team can work independently in different procedures or in a centralized, the details of their work until all the logic design s
35、tage may be displayed. In these circumstances, logic database design stage must be the original concept of data model and user view these independent or merged into a comprehensive design. In logic design information systems also identify additional information processing needs of these new demands
36、at this time must be integrated into the logic of earlier identified in the database design.Logical database design is based on the final step for the formation of good data specifications and determine the rules, the combination, the data after consultation specifications or converted into basic at
37、omic element. Most of todays database, these rules from the relational database theory and the process known as standardization, #. This step is the result of management of these data have not cited any database management system for a complete description of the database map. Logical database desig
38、n completed, we began to identify in detail the logic of the computer program and maintenance, the report contents of the database for inquiries. 1.4 Physical database design and definitionPhysical database design and definition phase decisions computer memory (usually disk) database in the organiza
39、tion, definition of According to the library management system for physical structure, the procedures outlined processing services, produce the desired management information and decision support statements. The objective of this stage is to design an effective and safe management of all data-proces
40、sing database, the physical database design to closely integrate the information systems of other physical aspects of the design, including procedures, computer hardware, operating systems and data communications networks. 1.5 Database Implementation The database prepared by the realization stage, t
41、esting and installation procedures for handling databases. Designers can use the standard programming language (such as COBOL, C or Visual Basic), the dedicated database processing languages (such as SQL), or the process of the non-exclusive language programming in order to produce a statement of th
42、e fixed format, the result will be displayed, and may also include charts. In achieving stage, but also the completion of all the database files, training users for information systems (database) user setup program. The final step is to use existing sources of information (documents legacy applicati
43、ons and databases and now needs new data) loading data. Loading data is often the first step in data from existing files and databases to an intermediate format (such as binary or text files) and then to turn intermediate loading data to a new database. Finally, running databases and related applica
44、tions for the actual user maintenance and retrieval of data. In operation, the regular backup database and the database when damaged or affected resume database.1.6 Database maintenance During the database in the progressive development of database maintenance. In this step, in order to meet changin
45、g business conditions, in order to correct the erroneous database design, database applications or processing speed increase, delete or change the structure of the database. When a procedure or failure of the computer database affect or damage the database may also be reconstruction. This step usual
46、ly is the longest in the database development process step, as it continued to databases and related applications throughout the life cycle, the development of each database can be seen as a brief database development process and data modeling concepts arise, logical and physical database design and
47、 database to achieve dealing with the changes. two. Information System developed by other means System Development Life Cycle minor changes in law or its variant of the often used to guide information systems and database development. Information System is a life-cycle methodology, it is highly stru
48、ctured approach, which includes many checks and balances to ensure that every step of produce accurate results, and new or alternative information system and it must communications or data definitions consistent existing system needs consistency. System development life cycle because of the regular
49、need to have a working system for a long time been criticized because only work in the system until the end of the whole process generated. More and more organizations now use rapid application development method, it is a includes analysis, design and implementation of steps to repeat the rapid iterative process until convergence to users the system so far. Rapid Application Development Act required the database