dr' Abonyi Jnos

1
Adattárházak és kiaknázásuk
Korszeru adatelérési eszközök (ODBC, OLE, XML)

• dr. Abonyi János
• Veszprémi Egyetem
• abonyij_at_fmt.vein.hu
• www.fmt.vein.hu/softcomp/dw

2
Mirol lesz szó?

• ODBC
• OLE DB
• XML

3
History of data access I.

• In the bad old days, (circa 1984-1991,) if you
  needed a program to access data source specific
  data from a thirdparty data location - whether it
  was an RDBMS, spreadsheet, or legacy system a
  great deal of effort and code was required.
• For Microsoft Excel, it was possible to buy the
  file format specification and write an extensive
  parsing tool, but it was usually easier to export
  it to some text-based format and re-format it
  manually.
• Embedded SQL and clunky preprocessors were the
  tools of choice to access early RDBMS servers.

4
History of data access II.

• Recognizing these problems, the SQL Access Group
  (SAG) was formed in 1989 as an industry
  consortium to promote portability and
  interoperability among DBMS vendors.
• Microsoft representatives were among the members,
  and the company anticipated the eventual release
  of the SAG standard by shipping the Open Database
  Connectivity API ( ODBC) and tool-set in 1992.
• ODBC was a major step forward in data
  connectivity, and played a significant role in
  the legitimization of the PC as a client platform
  for enterprise applications.
• It also fueled the sales of PC-based
  report-writers, rapid application development
  (RAD) environments, and related tools.

5
History of data access III.

• ODBC has its share of problems, however. As a
  product of SAG, ODBC is tightly bound to the SQL
  language, and therefore to the relational model
  that underlies it. This creates a problem when
  attempting to use ODBC to access hierarchical,
  multi-dimensional (OLAP), or freeform
  (spreadsheets, e-mail) data.
• The requirement for SQL processing also poses an
  entry barrier for many of the low-end tabular
  data stores.
• Another problem is the leveling of ODBC
  drivers. The designers of ODBC anticipated that
  not all data-sources would be capable of
  providing the full functionality, so the API
  entries were grouped into Core, Level 1, and
  Level 2 to indicate which features were
  required and which considered optional for
  drivers of various capabilities. Unfortunately,
  since the development of most drivers is outside
  Microsofts control, many use a pickand-choose
  approach to ODBC API support.

6
History of data access IV.

• Similarly, the many flavors of SQL result in each
  driver having to report its degree of conformance
  to various SQL specifications.
• The resulting complexity of these issues,
  exacerbated by the changes contained in various
  versions of ODBC, seriously undermines the
  universal nature of the API. In the worst-case
  scenarios, developers are forced to treat each
  DBMS separately just as they did before ODBC
  was available.
• So, in order to deal with this stack of problems
  (and to address the increasingly important issues
  of Internet integration, multi-threaded
  applications, and other technical issues)
  Microsoft developed the OLE DB specification.

7
What is ODBC ?

• The Open Database Connectivity (ODBC) standard is
  a common application programming interface for
  accessing data files.
• ODBC allows you to easily move data back and
  forth between ODBC compliant applications.
• So, for example, you can easily extract data from
  a Microsoft Excel spreadsheet or Paradox database
  using SAS or SPSS without having to go through
  any tedious translation procedure.
• With ODBC, all you have to do is specify the
  type of file you are accessing and the data you
  want, and the ODBC driver will do the rest of the
  work.

8
Why Use ODBC?

• ODBC can be used as a translation engine,
  facilitating the movement of data between
  different applications with dissimilar file
  structures. While most applications have some
  sort of import file command that performs much
  the same task as ODBC in this kind of situation,
  ODBC typically provides a more versatile and in
  some cases easier to use interface for
  transferring data between programs.

9
ODBC Objectives

• For several years, a working party called the
  SQL-Access Group (SAG), consisting of interested
  hardware and software vendors, worked towards
  defining a common database access method to
  simplify Client-Server computing. Microsoft has
  taken a core set of SAG's output and developed a
  call-level interface called the Open Database
  Connectivity (ODBC) Application Programming
  Interface (API).
• ODBC defines a low-level set of calls which allow
  client applications and servers applications to
  exchange instructions and share data without
  needing to know anything about each other. It
  applies to any Client-Server operation, whether
  or not the client and server applications are
  resident on the same machine, or on different
  PC's, or even if the server is on a remote
  machine running a different operating system.

10
ODBC is based on SQL

• ODBC is based on and closely aligned with The
  Open Group standard Structured Query Language
  (SQL) Call-Level Interface. It allows programs to
  use SQL requests that will access databases
  without having to know the proprietary interfaces
  to the databases.
• ODBC handles the SQL request and converts it into
  a request the individual database system
  understands.

11
History of ODBC

• ODBC was created by the SQL Access Group and
  first released in September, 1992. Although
  Microsoft Windows was the first to provide an
  ODBC product, versions now exist for UNIX, OS/2,
  and Macintosh platforms as well.
• In the newer distributed object architecture
  called Common Object Request Broker Architecture
  (CORBA), the Persistent Object Service (POS) is a
  superset of both the Call-Level Interface and
  ODBC.
• When writing programs in the Java language and
  using the Java Database Connectivity (JDBC)
  application program interface, you can use a
  product that includes a JDBC-ODBC "bridge"
  program to reach ODBC-accessible databases.

12
ODBC Theory of Operation

• The Open Database Connectivity (ODBC) interface
  allows applications to access data from database
  management systems (DBMS).
• The interface permits maximum interoperability -
  a single application can access diverse back-end
  database management systems.
• An application developer can develop, compile,
  and ship an application without targeting a
  specific DBMS product.
• Users can then add modules called database
  drivers that link the application to their choice
  of database management systems.

13
The ODBC interface defines the following

• A library of ODBC function calls that allow an
  application to connect to a DBMS, execute SQL
  statements, and retrieve results.
• A standard way to connect and log on to a DBMS.
• A standardized representation for data types.

14
The Interface is flexible enabling

• An application to construct SQL statements at
  compile time or at run time
• An application to use the same object code to
  access different DBMS products
• An application to connect to multiple instances
  of DBMS products
• An application to send and receive data in a
  format convenient to the application the driver
  converts data values to the format of a specific
  DBMS product.

15
ODBC Architecture I.
Client
DBMS

• Procedural API
• SQL-centric

16
ODBC Architecture II.

• ODBC is composed of four main components the
  client application, ODBC driver manager, ODBC
  drivers, and the ODBC data source.
• The client application in this discussion will
  SAS, and is the computer package that you will be
  using for your data manipulation or analysis.
• The ODBC driver manager is a program used to
  manage the links between the various applications
  and to configure data transfer settings.
• The ODBC drivers serve as the link between the
  client and ODBC data, and must be installed for
  each application. These drivers might also be
  composed of a network component that could be
  used to transfer data from a remote server.
• Finally, an ODBC data source is a user-generated
  configuration of a particular data file created
  using the ODBC driver manager for access by
  client applications.

17
Types of Drivers

• ODBC defines two types of drivers
• Single-tier The driver processes both ODBC calls
  and SQL statements.
• Multiple-tier The driver processes ODBC calls
  and passes SQL statements to the data source.
• One system can contain both types of
  configurations.

18
Single-Tier Configuration

• In a single-tier implementation, the database is
  processed directly by the driver. The driver
  processes SQL statements and retrieves
  information from the database.
• In a single tier driver network environment data
  access software resides on the PC, this implies
  that query resolution intelligence resides on the
  client.

19
Multiple-Tier Configuration I.

• In a multiple-tier configuration, the driver
  sends SQL requests to a server that processes SQL
  requests.
• The application, driver, and Driver Manager
  reside on one system, typically called the
  client. The database and the software that
  controls access to the database typically reside
  on another system, typically called the server.
  This implies that query resolution intelligence
  resides on the server.
• A variant of the multiple-tier configuration is a
  gateway architecture, where the driver passes SQL
  requests to a gateway process. The gateway
  process sends the requests to the data source.
  The gateway in this scenario can be a piece of
  hardware or data access software in the form of a
  low level interface to foreign databases provided
  by the RDBMS vendor.

20
Multiple-Tier Configuration II.
21
Setting Up an ODBC Data Source

• Once the ODBC drivers are installed on your
  workstation you may have to create an ODBC data
  source which serves as a configuration profile
  for the data files.
• To check which data sources have already been
  created, use the ODBC driver manager.

22
ODBC driver installation

• Because ODBC driver installation is independent
  of ODBC data source configuration, we would have
  to click the Drivers... button in the ODBC
  administrator to see a list of the installed
  drivers.

23
Creating an ODBC Data Source

• There are two types of ODBC data sources -- what
  one could refer to as "generic" and
  "file-specific" data sources.
• A generic data source is automatically created by
  some programs during the installation stage.
  These types of data sources provide all of the
  information necessary to access a file using
  ODBC, except for the name and location of the
  file.
• Thus, when using a generic data source, the user
  must specify the drive, path, and filename
  interactively.
• In contrast, a file-specific data source is a
  much more precise configuration that designates a
  particular file in a particular location as an
  ODBC data source. A file-specific ODBC data
  source is usually created in situations where a
  single file is going to be accessed many times
  because it allows you to refer to the file by the
  assigned name rather than having to select the
  file each time you wish to access it.

24
ODBC Overview

• Access to relational data only
• Requires an SQL engine
• Standard SQL defined
• PC and Unix data access (VMS?)
• Mature technology

25
WHAT IS OLE DB?

• OLE DB is not a product at least in the
  traditional sense.
• Instead, it is a specification of how data
  consumers should communicate with data providers.
  
• The goal of OLE DB is to provide an open,
  extensible standardallowing any software that
  conforms to the specification to communicate via
  a set of common interfaces.
• OLE DB is built on some of Microsofts key
  strategic technologies, and builds upon the
  heritage of the previous generation of data
  access software.

26
Overview of COM I.

• Software developers have been talking and writing
  about the software crisis since the late
  1960s. While that is too complex an issue to
  discuss here, in essence it refers to the fact
  that software productivity and quality continues
  to fall further and further behind hardware
  quality and capabilities.
• One partial solution that has been applied to
  this situation is software component reuse in
  the same way that a computer designer creates a
  system from a set of common parts, software
  developers would like to be able to build a
  system from reusable software components.

27
Overview of COM II.

• Object-oriented software languages and design
  have contributed to this effort, and defined some
  useful concepts for managing reuse, such as
  encapsulation, polymorphism, and inheritance.
• Among other benefits, these concepts all serve to
  isolate changes in one piece of software,
  preventing them from affecting other software.
• COM uses these object-oriented concepts, but
  unlike most other approaches, COM attempts to
  promote software component reuse at the binary
  object level instead of at the code level.
• This has several benefits, but most importantly,
  it allows a software component to be upgraded or
  replaced transparently to the applications using
  it.

28
Overview of COM III.

• The primary mechanism in COM for insuring
  consistency across versions is the interface,
• and the first rule of COM programming is that any
  interface, once published, is immutable.
• Specifically, the syntax of all methods and their
  arguments must remain the same.
• In addition, all access to a com object must be
  though the interface methods directly modifying
  an objects data is forbidden (encapsulation.)

29
Overview of COM III.

• Another key aspect of the interface is that it is
  not tied to a specific object, but can be reused
  (inheritance). This allows the calling program to
  treat all objects that provide a specific
  interface identically (polymorphism).
• To further leverage this capability, COM provides
  a mechanism for a program to determine which
  interfaces each object supports at execution
  time, eliminating the need for the consumer to
  limit itself to a lowest-common-denominator
  functionality.
• In addition to addressing the above problems, COM
  was designed with network computing in mind. Such
  programming issues as multi-threaded and parallel
  computing, distributed transaction coordination,
  were addressed and generic solutions devised.

30
HOW DOES OLE DB WORK?

• From a system perspective, OLE DB is still simply
  database middleware like ODBC, in that it serves
  as a translator between a client and a server.
• The calling application obtains a COM Interface
  handles and executes the methods of interest.
• Client and server are sufficiently ambiguous
  terms that OLE DB has introduced the terms
  consumer and provider to clarify the
  relation.
• A consumer is any software that requests an OLE
  DB interface, while a provider is any software
  that implements one or more OLE DB interfaces.

31
OLE DB vs. ODBC I.

• Procedural API
• SQL-centric

ODBC Architecture
Client
DBMS
OLE DB Architecture
Client
DBMS

• COM classes
• Direct table manipulation
• Generic command processing

COM
OLE DB Provider
32
OLE DB vs. ODBC II.

• OLE DB differs from ODBC in that there is no
  equivalent to the ODBC driver manager. OLE DB
  providers are self sufficient in this sense,
  although multiple providers may cooperate to
  access a single data source.
• For example, Microsoft distributes a provider
  enumerator (which is itself an OLE DB provider)
  as part of the OLE DB redistributable files. It
  is used to obtain the list of installed
  providers.
• An important aspect of OLE DB is that instead of
  entirely replacing ODBC, it leverages the
  existing base of ODBC drivers by providing a
  translation provider the OLE DB Provider for
  ODBC Data Sources for those sources.

33
How to develop OLE DB tools?

• The OLE DB API, like all COM APIs, is technically
  language independent but it is most conveniently
  accessed in C.
• To make OLE DB technology more accessible to
  languages such as Visual Basic, Microsoft has
  provided an abstraction layer called Active Data
  Objects (ADO).
• ADO provides most of the power of OLE DB, but
  with less code and a much shallower learning
  curve.

34
ADO Architecture
OLE DB Provider
Client
ADO
ODBC
MSADSQL

• Scriptable bindings
• Simplifies OLE DB model
• Appropriate for e-business applications

35
OLE DB Objects
36
Enumerators

• Enumerators provide lists of information to the
  consumer.
• The most common example, the root provider
  supplied by Microsoft searches for installed OLE
  DB providers, including other enumerators, and
  returns their identifying information in a
  rowset.
• Enumerators are most useful to generic consumers
  those that are designed to connect to arbitrary
  providers selected at run-time.
• In some cases, however, a data provider will
  include additional enumerators such as when an
  RDBMS manages multiple databases residing on a
  single server.

37
Data Sources

• Data source objects perform the work required to
  connect to a data source, such as a tabular file
  or a DBMS.
• They manage connection options make the network
  connection or open the file(s), as appropriate
  and may supply administrative capabilities and
  general information about the data source.
• All consumers must obtain a data source object.
  Further, they must initialize the data source
  object before it can create a session or perform
  other operations of the data source.

38
Sessions

• Sessions provide a framework for transaction
  management, which can be handled automatically,
  or explicitly controlled by the consumer.
• If a data source doesnt support transactions,
  the session object will operate in auto-commit
  mode, where all changes will be saved as they are
  made.
• Sessions are a prerequisite for creating
  transactions, commands, and most rowsets.
• They may also provide schema information and the
  ability to create or modify tables and indexes.

39
Transactions

• Transaction objects are optional, and serve a
  single purpose.
• They allow the consumer to explicitly control the
  commission or abortion of pending database
  options.
• Not all providers support transactions, and those
  that do may handle only simple transactions.
• More sophisticated providers may participate in
  nested or coordinated transaction management.

40
Commands

• Commands are used to submit statements in SQL (or
  other appropriate language, such as the MDX
  extended SQL syntax for OLAP) to the provider.
• Unlike ODBC, command support is optional for OLE
  DB providers.
• Those providers that do have command languages
  may use different dialects, although they must
  report which dialect(s) are supported, and to
  what extent.

41
Rowsets

• Rowsets are the OLE DB objects that contain the
  data
• the actual rows and columns, or observations and
  variables if you prefer.
• rowset is effectively a local buffer for the
  data, although advanced providers support a vast
  array of client and server-side cursors,
  scrolling,bookmarks, etc. through optional
  interfaces.
• This allows buffer management to be tuned
  effectively by each consumer based on its usage
  pattern and the capabilities of the provider.

42
Errors

• Errors in OLE DB are simply an extension of the
  error handling objects built into COM.
• Support for error objects are, sadly, entirely at
  the discretion of the provider often requiring
  the consumer to rely on basic method return codes
  to check for success or failure.
• However, the better providers can and do provide
  extensive diagnostic messages in the error
  objects.

43
OLE ODB Dependencies
44
Providers

• All major RDBMS brands have an available OLE DB
  provider, though some are native while others
  rely on middleware solutions.
• Microsoft SQL Server, Microsoft Access, and
  Oracle providers are available directly from
  Microsoft, while MicroFocus3 sells direct
  providers for Oracle, Sybase, and INFORMIX
• Non-relational providers are also available, and
  are currently available for various flat file
  formats (such as dBase) and email systems
  including Lotus Notes and Microsoft Exchange.
• File system providers are available for Microsoft
  Active Directory Services2 (NT, NetWare, and
  LDAP) and C-ISAM/D-ISAM, with VSAM slated for
  release soon.

45
OLAP providers

• On the OLAP side, providers for the
  multi-dimensional servers from Microsoft, Oracle,
  Red Brick, and Sybase are shipping or in the
  works.
• The OLE DB for OLAP specification is relatively
  recent, so more providers are anticipated in this
  area over the coming year.
• SAS Institute itself is developing four
  providers, three relational and one OLAP
• The SAS/Access Interface to OLE DB has the
  ability to consume data from all of these
  providers, enabling SAS software to serve as a
  central information processing tool for all your
  data, regardless of its origin.

46
SAS Providers
SAS/SHARE Data Provider
SAS/SHARE Server
C ONSUMER
IOM Data Provider
SAS/Integration Technologies Server
SAS Local Data Provider
47
Consumers

• There are two main categories for OLE DB consumer
  applications those that process data from a
  variety of OLE DB providers, and those that are
  designed to interact with a single provider.
• The first group generic consumers largely
  consists of shrink-wrap software.
• The single-source consumers are typically
  proprietary corporate IS applications, and are
  often bound directly to the schema of a
  particular database.
• Of the shrink-wrap applications, few are
  available now, but Microsoft has announced
  sweeping support in the Microsoft Office 2000
  suite. Microsoft Excel 2000, for example, will be
  able to import data from any OLE DB source, and
  will generate pivot tables linked directly to
  OLAP cubes.
• Similarly, Microsoft Access will support import
  and view capabilities for OLE DB providers. One
  application that is available now is Crystal
  Reports 7, from Seagate software, which can
  produce reports using data from OLE DB sources.

48
OLAP Consumers

• OLE DB for OLAP consumers will see some exciting
  developments over the next year. Support
  initiative have been announced by most of the
  leading OLAP companies, including BAAN, Brio,
  Cognos, Hyperion, Knosys, Pilot, and of course
  Microsoft and SAS Institute.
• Among proprietary consumers, the important news
  is not the applications so much as the
  development tools available to leverage OLE DB.
• Microsoft Visual Basic and Visual C offer tools
  and Wizards to simplify the development of OLE
  DB-based applications.
• Sybase has announced OLE DB support in
  PowerBuilder Version 7.
• There are also a number of web technologies such
  as Microsoft ADO/RDS to use OLE DB technology to
  deploy web solutions.

49
OLE DB Overview

• Access to relational/non-relational data
• No SQL required in API
• No standard SQL defined
• PC centric
• Built on top of OLE and COM
• Newer technology will replace ODBC

50
Conclusion

• OLE DB is an open, extensible standard that
  shrewdly leverages existing database access
  tools. It is slated to be a key technology in the
  next generation of Windows based software.
• It is no secret that the industry drive for open,
  collaborative software development is gathering
  momentum driven by distributed, web-based
  technologies.
• The leaders in 21st century information delivery
  will be those who recognize this opportunity and
  seize it.
• The goal of the SAS Nashville project is to put
  SAS Institute at the forefront of this trend, and
  the OLE DB providers and consumer are a critical
  component of that effort.

51
What is XML?

• Web data format standard
• W3C (World Wide Web Consortium)
• Separates data from presentation
• Self describing data format
• ltMasterRecipegt
• ltHeadergt
• ltProductIDgt M101 lt/ProductIDgt
• lt/Headergt
• lt/MasterRecipegt
• ltTagsgt describe data values
• Tags must match
• ltProductIDgt ltproductIDgt ltProd_IDgt

52
XML what does it look like?

• Stuff
• Delimited by lttagsgt ltyougt ltmakegt ltupgt
• Said lttagsgt follow certain rules.
• This is the big deal one can rely on the
  well-formed-ness to build stronger applications

53
XML what does it look like?

• lt?xml version"1.0" encoding"iso-8859-1" ?gt
• ltNHLgt
• ltCONFERENCEgt Eastern
• ltDIVISIONgt Atlantic
• ltTEAM name"New Jersey Devils"
  abbrev"NJ" /gt
• ltTEAM name"New York Islanders"
  abbrev"NYI" /gt
• lt/DIVISIONgt
• ltDIVISIONgt Northeast
• ..
• lt/NHLgt

54
XML an agreed upon vocabulary

• Applications agree to interchange information
  using a common vocabulary
• But way more flexible.
• Easier to evolve.
• Applications ignore vocabulary they dont
  understand.

55
DTDs Schemas

• XML 1.0
• Document Type Definitions (DTDs)
• Inherited from SGML (Standard Generalized Markup
  Language)
• Limitations
• W3C Schema format to replace DTDs
• Expected in 2000
• Overcomes DTD limitations adds new capabilities
• Microsoft developed proprietary schema format
• .xdr schema data reduced
• Expected (hoped?) to use W3C schema format in the
  future

56
XML Schema Usage

• Define valid elements
• Define valid hierarchy organization of elements
• Reduce programming required to validate XML
  documents

XML Processor
XML Processor uses the output of the parser and
displays the XML
XML Schema
XML Document References Schema
Display
or
XML
Data Application
Parser
XML parser validates the XML document using the
referenced schema
XMLDocument
57
XML easier to evolve?
APP 2
APP 1
XML v1
58
XML easier to evolve?
APP 2
APP 1
XML v1 XML v1.1
APP 3
59
External References

• http//www.w3.org
• look down the lefthandside for XML
• http//www.xml.com
• http//www.xml.org
• http//www.simonstl.com
• good stuff in articles/presentations

60
XML for Data Exchange

• Natural fit for structured data
• Mainstream technology
• Overcomes many relational database limitations
• Text files
• Low overhead costs and tools
• Human readable without tools
• Tools enhance readability
• Transportable
• HTTP
• E-mail
• Easy to exchange recipe fragments
• Recipe unit procedure or recipe operation
• Individual formula items

61
Customized Schemas
Industry Standard Schema
Individual Schemas Based upon Industry Standard
Company
Company
A
B
Vendor Y
Vendor X
Corporate
Corporate
Schema
Schema
Schema
Schema
Schema Mapping Conversion Tools
62
E.g. Batch Control Markup Language

• BatchML
• Based on S88.02
• Basis for corporate vendor specific schemas
• Data exchange for
• Master Recipes
• Batch History
• Batch Schedules
• Equipment Definitions

63
E.g. Master Recipe Header
lt?xml version"1.0" encoding"UTF-8"?gt ltMasterReci
pe xmlns"x-schemaMasterRecipeV01.xdr"gt ltHeadergt
ltIDgt PROD100 lt/IDgt ltCreationDategt
2000-09-04T124348 lt/CreationDategt ltVersiongt
4.0 lt/Versiongt ltVersionDategt 2000-10-10T162303
lt/VersionDategt ltVersionAuthorgt Michelle
lt/VersionAuthorgt ltApprovalDategt
2000-10-12T130422 lt/ApprovalDategt ltApprovedBygt
Paul lt/ApprovedBygt ltStatusgt Approvedlt/Statusgt
ltStandardSizegt 100.0lt/StandardSizegt ltMaximumSiz
egt 300.0 lt/MaximumSizegt ltMinimumSizegt 50.0
lt/MinimumSizegt ltEngineeringUnitgt KG
lt/EngineeringUnitgt ltCommentgt PRODUCT X
lt/Commentgt ltDescriptiongt Production of Polymer
X lt/Descriptiongt ltProductNamegt Polymer X
lt/ProductNamegt lt/Headergt lt/MasterRecipegt
64
E.g. Master Recipe XML Schema
lt?xml version "1.0"?gt ltSchema name
"MasterRecipe00-01.xdr" xmlns
"urnschemas-microsoft-comxml-data" xmlnsdt
"urnschemas-microsoft-comdatatypes"gt ltElementT
ype name "MasterRecipe" content "eltOnly"
order "seq"gt ltelement type
"ID"/gt ltelement type "Header minOccurs"0"
maxOccurs"1"/gt ltelement type
"EquipmentRequirement" minOccurs"0"
maxOccurs"1"/gt ltelement type "Formula"
minOccurs"0" maxOccurs"1"/gt ltelement type
"Procedure" minOccurs"0" maxOccurs"1"/gt ltele
ment type "UnitProcedure" minOccurs"0"
maxOccurs""/gt ltelement type
"OtherInformation" minOccurs"0"
maxOccurs""/gt lt/ElementTypegt ltElementType
name "UnitProcedure" content "eltOnly" order
"seq"gt ltelement type "ID"/gt ltelement type
"RPEClass" minOccurs"0" maxOccurs"1"/gt ltel
ement type "InstanceType" minOccurs"0"
maxOccurs"1"/gt ltelement type
"EquipmentProceduralElementID" minOccurs"0
" maxOccurs"1"/gt ltelement type "Header"
minOccurs"0" maxOccurs"1"/gt ltelement type
"EquipmentRequirement" minOccurs"0"
maxOccurs"1"/gt ltelement type "Formula"
minOccurs"0" maxOccurs"1"/gt ltelement type
"Procedure" minOccurs"0" maxOccurs"1"/gt ltele
ment type "Operation" minOccurs"0"
maxOccurs""/gt ltelement type
"OtherInformation" minOccurs"0"
maxOccurs""/gt lt/ElementTypegt
65
XML what's the fuss

• XML might just be the ASCII text file of the
  new millennium. A universal and accepted way to
  transfer information between computer
  applications
• Many big initiatives are embracing XML
• CDISC pharmaceutical submissions
• most B2B exchanges
• most DBMS software
• most ERP software
• XML at the heart of .NET and Hailstorm

66
(No Transcript)
67
Overview

• Using ODBC drivers, OLE DB providers and XML
  extends your data reach from the SAS System.

68
Buzz words I.

• ActiveX A label used to refer to COM-based
  components and technologies, especially those
  related to the internet.
• ADO Active Data Objects an abstraction layer to
  provide OLE DB connectivity
• API Application Programming Interface a
  collection of functions or procedures that
  provide a specific software capability.
• COM The OLE Component Object Model, a specific
  approach to Object-Oriented software development
  developed by Microsoft.
• Consumer Any software component that utilizes an
  OLE DB interface. OLEDB
• Driver In ODBC, a software component providing
  access to a specific DBMS or file format.

69
Buzz words II.

• Encapsulation A protective programming technique
  that prevents external code from modifying the
  contents of data structures internal to an object
  or module.
• Inheritance A programming technique that
  facilitates the re-use of programming constructs.
  In COM, interface specifications are inherited
  rather than method implementation.
• IOM Interface Object Model a SAS Nashville
• initiative to support distributed,
  component-based development.
• ISV Independent software vendor
• Interface An immutable collection of methods.
• MDDB Multi-Dimensional Database
• Method A function implemented by an object.

70
Buzz words III.

• Object In COM, a binary construct that provides a
  set of interfaces.
• ODBC Open Database Connectivity a Microsoft API
  for accessing SQL-based data sources.
• OLAP On-Line Analytical Processingcommonly
  refers to the consolidated storage and
  multi-dimensional analysis of enterprise data.
• OLE Originally an acronym for Object Linking and
  Embedding, it is now used simply as a label to
  describe many COM based technologies. (See also
  ActiveX)
• Polymorphism A programming technique that allows
  disparate object types to be treated
  homogeneously. In COM, inheriting immutable
  interface specifications
• provides this.

71
Buzz words IV.

• Provider Any software component that exposes an
• OLE DB interface. Data providers own or present
  data, while service providers apply an operation
  or transformation
• RDBMS Relational Database Management System
• RIO Remote I/O A SAS Nashville project
  initiative for distributed I/O.
• SQL Originally an acronym for Structured Query
  Language, now a label for the language used to
  retrieve specific data from an RDBMS