Adaptive data manager

Abstract

A method for managing a back-end information storage infrastructure and a flexible development environment for data storage using a computer system. The method includes managing system resources including a relational database. Meta data models are created to model processes and to define meta data elements and their relationships by using trees and graphs. The method manages access to the data by authenticating users through a directory describing user rights, while providing management of multi-user access and concurrency. The method includes running the processes that generate instance data, storing the instance data following the meta data model, and transforming the instance data into physical views.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention is an advanced data management method and system, including both an information storage infrastructure and a database development environment, generally relating to relational databases, object oriented databases, and document management systems.

2. Description of the Related Art

The inventive advanced data management method and system are capable of use with relational databases, web site and client server applications, document and knowledge management applications, discrete data storage, and object-oriented databases.

The invention is particularly useful when used with relational databases (RDBMS), such as Oracle8i, for data management and development.

The past three years have seen the emergence of extensible markup language (XML). Extensible markup language describes a class of data objects called XML documents and partially describes the behavior of computer programs which process them. XML is an application profile or restricted form of standard generalized markup language (SGML) (ISO 8879). By construction, XML documents are conforming SGML documents. In other words, XML enables any storage of information into a text file format for example, as a document. XML databases allow discrete storage of such documents.

Lucent Technologies is believed to have patented a database management system using transitive closure solutions using standard SQL, primarily targeted to solve project management issues, i.e., pert charts or directed graph temporal dependencies.

Until now, efforts to create data management systems and software for databases have involved extremely lengthy and costly development cycles for database design, implementation, tuning and maintenance. These problems have been all too typical in the design and specification of schemas for projects.

Similarly, data warehousing projects have involved complex, time-consuming and expensive phases including data cleansing, building a snowflake model or other models, and harvesting reports. The first phase, data cleansing, typically has represented over 80% of the total time required for such data warehousing projects.

What is needed is a data management method and system that will avoid or eliminate the problems of existing systems.

BRIEF SUMMARY OF THE INVENTION

It is an object of the invention to provide a method and system for developing software and managing data, such as for managing patient medical records and other data, while avoiding or minimizing lengthy and costly development cycles for database design, implementation, tuning and maintenance, data cleansing, building of snowflake or other models, and difficulties in harvesting data and generating reports and user screens.

In accordance with the inventive adaptive data management (ADM) method and system, a static database model is provided and is optimized for manipulation of large volumes of data with a standard front-end component interface, thereby simplifying database design and data access, while reducing development cost and development time. Furthermore, the model does not require numerous qualified technical personnel to monitor all database activity, the data to be collected is described, both in format and in relationships, as meta data to the model, and user data (instance data) is then collected and stored using the format defined by the meta data. In tests, the inventive method and system have shown success in managing large numbers of records and in document indexing as useful in such applications as web sites. Due to the nature of ADM data storage, the data requires little of no cleansing before inclusion into a data warehousing database.

ADM provides access protection and tracking, ensuring data security and integrity, through a gateway requiring identity authentication and multi-layered access control. ADM manages multiuser access and concurrency.

The ADM may be used with an Oracle database running on any of several platforms to provide the data storage support, with as many as about 14 or more objects participating to the design. A set of components, developed as Microsoft COM objects, provide access to user front-end applications.

ADM provides both a back-end information storage infrastructure and a flexible development environment for data storage. ADM is based on a meta data model. The organization of the data itself (the meta data) is described to ADM prior to any collection of data. The meta data model encloses definitions of meta data elements as well as the relationships among these meta data elements. Data elements may be organized as trees, i.e., a meta data element has at most one parent data element, or as graphs, i.e., a meta data element may have one or more parent data elements, thus allowing representations of most possible data models.

ADM provides support for multiple development environments, using a simple component interface for complex back-end data storage, thereby simplifying access to instance data. Instance data consists of stored user data patterned after the meta data definition. The development environment includes a COM object, accessed from all applications referencing ADM, and an administration tool for model management. ADM allows transfer of data to and from ADM using XML. The XML document type definition is defined by the meta data definition.

ADM has been developed for access from conventional development environments (Microsoft Visual C++ and Visual Basic, Borland Delphi) as well as web environment tools such as Microsoft Active Server Pages (ASP). This tool is well suited for short transactions characteristic of web environments.

ADM provides additional simplified data access to any user, from the relational database manager standpoint, by allowing view definitions. A view consists of many meta data elements, which may be tightly or loosely connected. As instance data is created, any user can access the instance data represented as views, from any database environment tools, such as Microsoft Access or Microsoft MS Query.

ADM is complemented by Visual ADM. ADM and Visual ADM fit to the object-oriented document-view paradigm, such as: ADM provides the data back end (document layer), while Visual ADM provides user interface(s) to the user (view layer). Visual ADM is a thin-client form based application: forms are defined as scripts, stored into the ADM database, and retrieved at the Visual ADM client location when requested. Visual ADM also provides a robust scripting language, allowing forms to implement any type of business rules.

ADM and Visual ADM can be provided in the form of an InstallShield application, including an ADM COM object, ADM Administration Tool, two Visual ADM executables, several PDF documents (‘ADM User Manual’, ‘ADM Administration Tool User manual’, ‘Visual ADM User Manual’, ‘Visual ADM Reference Manual’), as well as a sample implementation. In one embodiment of the invention, a running instance of Oracle8i is required prior to installation.

The inventive adaptive data management (ADM) method and system provide novel and useful features such as the ability to use both graph and tree structures in an optimized data model stored in a relational database. Also, ADM permits presentation of stored data as conventional tables (data view) for standard reporting. As the data changes and expands, the content of the data views reflects the changes. Any of these data views can be defined by end users and created automatically by ADM back end service. ADM provides a component for simple front-end interface development using Microsoft COM objects while providing access to each aspect of the inventive method and system. Moreover, ADM provides a transactional data access model suitable for web-based and client-server implementation.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic of the inventive adaptive database management component interfaces.

FIG. 2 is a system-wide description of the adaptive data manager.

FIG. 3 is an example of meta data.

FIG. 4 is an example of instance data.

FIG. 5 is a schematic of interactions between ADM and Visual ADM.

FIG. 6 is an overview of the entity relationship diagram for the ADM database.

FIG. 7 is a detailed view of the ADM Directory portion of FIG. 6.

FIG. 8 is a detailed view of the Meta Definitions portion of FIG. 6.

FIG. 9 is a detailed view of the Instance Data Definition portion of FIG. 6.

FIG. 10 is a detailed view of the Reference Data portion of FIG. 6.

FIG. 11 is a screen of the ADM Adminstration Utility Directory.

FIG. 12 is a screen of the ADM Adminstration Utility User Manager.

FIG. 13 is a screen of the ADM Adminstration Utility Facility Manager.

FIG. 14 is a screen of the ADM Adminstration Utility Meta Data Manager.

FIG. 15 is a screen of the ADM Adminstration Utility Meta View Manager.

FIG. 16 is a screen of the ADM Adminstration Utility Instance Data Manager.

FIG. 17 is a screen of the ADM Adminstration Utility Dictionary Manager.

FIG. 18 is a screen of the ADM Adminstration Utility Instance Data Lock Manager.

DETAILED DESCRIPTION OF THE INVENTION

ADM is a data management solution. It is both a back-end information storage infrastructure and a flexible development environment for data storage. ADM is based on a meta data manager concept, namely, the organization of the data itself (the meta data) is described to ADM (prior to any collection of data). The meta data Manager encloses definitions of meta data elements as well as the relationships among these meta data elements. Data elements may be organized as trees (i.e. a meta data element has at most one parent data element) or as graphs (i.e. a meta data element may have one or more parent data elements), thus allowing representations of most possible data models.

Turning to FIG. 1, the inventive adaptive data manager component interfaces are shown. In accordance with the invention as illustrated in FIG. 1 and the other figures, a method is provided for managing data using a computer system, comprising managing system resources, authenticating and selectively providing access to users through a directory describing predetermined user rights, modeling and organizing processes as meta data, running the processes and generating instance data, storing instance data following the meta data model while providing management of multi-user access and concurrency, and transforming the instance data into a report. The meta data may be organized in both trees and graphs. The inventive method is particularly useful for managing data relating to complex processes characteristic of humans and other living organisms.

ADM is a development environment. ADM provides a simple component interface to the complex back-end data storage. This interface simplifies access to instance data. Instance data consists of the stored user data, patterned after the meta data definition. The development environment includes a COM object, accessed from all applications referencing ADM, and an administration tool for model management. ADM allows transfer of data to and from ADM using XML. The XML document type definition is defined by the Meta Data definition.

ADM has been developed for access from conventional development environments (Microsoft Visual C++ and Visual Basic, Borland Delphi) as well as web environment tools such as Microsoft Active Server Pages (ASP). This tool is well suited for short transactions characteristic of web environments.

Turning to FIG. 2, ADM is shown in an implementation context, showing the relationship between the back-end objects including the adaptive data manager and the administration utility vis-à-vis visual ADM in client-server mode or peer to peer mode, web based user interfaces, external interfaces, such as HL7, and XML support for inport and export operations.

Turning to FIG. 5, the interactions between ADM and Visual ADM are shown.

ADM provides additional ‘simplified data access’ to any user from the relational database manager standpoint, by allowing view definitions. A view consists of many meta data elements, tightly or loosely connected. As instance data is created, any user can access the instance data represented as views, from any database environment tools, such as Microsoft Access or Microsoft MS Query.

ADM is provided as an InstallShield. It consists of the ADM COM object, an Administration Tool, an ‘ADM User Manual’ PDF document, an ‘ADM Administration Tool User manual’ PDF document, as well as a sample implementation. A running instance of Oracle8i is required prior to installation.

Turning to FIG. 6, a schematic overview of the database design for ADM and Visual ADM is shown. FIGS. 7-10 provide the detailed views of each of the components of FIG. 6.

COM Technical Description

The ADM COM Object achieves mainly three distinct functions: data access and indexing, data transformation, and administration. These functions are available through several separate interfaces:

ADM.PROFILE

ADM.Profile is the gateway interface to ADM. An instance of the ADM.Profile interface must be created prior to any other ADM interfaces. In order to filly set the ADM.Profile interface, both successful login transaction to the Oracle database and user identification (either using trusted authentication or explicit authentication) are required. The profile interface is then communicated to any other interface for safely communicating user and database information.

ADM.DATAGATE

ADM.Datagate is the interface allowing data access. It provides multiple paths to access, add, modify, and delete any stored data (i.e. instance data). ADM.Datagate is also the interface for processing search requests.

ADM.LOCK

Data elements accessed via the ADM.Datagate interface are read-only unless the user requests exclusive access to portions of the data: during this process, locks will be placed on data elements, guaranteeing exclusive read/write access to the data. The ADM.Lock interface is a purely administrative interface. This interface allows the ADM Administrative Utility to monitor and delete locks.

ADM.DICTIONARY

Relational database model always includes dictionaries or reference tables for standard values. For example, if a study includes a field for ‘Gender’, the values for gender may be coded as 1 being ‘Female’ and 2 being ‘Male’. Only ‘1’ or ‘2’ will be stored into the instance data, thus possibly allowing the description to be changed later without affecting the instance data.

ADM.Dictionary supports this model by providing methods to query codes and descriptions, retrieve a description given a code, etc.

ADM.INDEX

ADM.Index is the interface supporting document indexing under the Adaptive Data Manager paradigm. Indexing is declared at the Meta data level in the form of indexing flags, and any instance data creation will result in indexing requests. The ADM.Index interface is the open interface allowing for any external application to process requests resulting in keyword or concept indexing. This interface will be used, typically, by processes (or services) or applications monitoring newly created data and processing the request for indexing.

ADM.METADATA

The Meta data is the structure formally defining how data is collected. This interface is purely administrative: only the ADM Administrative Utility will make use of this interface. Note that inexperienced use of this interface may be detrimental to existing data elements.

ADM.METAVIEW

The Meta view structure is the data transformation aspect of ADM: instance data (created after the meta data defined pattern) is stored as graph into the database. Such organization does not easily enable users to report on any data elements. The Meta view structure allows users to define views as collections of meta data elements: when processed, these views are physical oracle tables, accessible by any user (with proper connection rights) with any SQL enabled tools.

The ADM.Meta View interface is purely administrative: only the ADM Administrative Utility will make use of this interface. This interface also works in conjunction with several server-side services.

ADM.Script

The ADM.Script interface is the support interface for the Visual ADM suite of tools: Visual ADM's APPLICATION BUILDER is the application managing data using ADM.Script; Visual ADM's FORM RUNNER is the application using this interface for end user processing. Note that this interface is only accessed from Visual ADM tools.

ADM.USER, ADM.FACILITY, ADM.DIRECTORY

These interfaces allow management of information related to user, security and user access rights.

These interfaces are purely administrative: only the ADM Administrative Utility will make use of these. Any information for a currently logged in ADM interface used is available through the required ADM.Profile interface.

ADM.PROFILE

The Adaptive Data Manager (ADM) is a data storage solution, and as such, must enforce and guarantee safety of stored data. ADM achieves this goal with complementary approaches:

By controlling user logins and users data access rights.

By keeping an audit trail of all user transactions (creations, updates and deletions) at the field level. Any user-modified data element is archived and tagged with the user information and date.

The ADM.Profile interface is the interface enforcing security with ADM. In order to access any of the interface services provided by ADM, one must successfully provide accurate information to log to the Oracle-based ADM data store, and also authenticate self as a registered ADM user.

Information required to log on to the Oracle data store consists of connection string, data stroe user name and password, and authentication user name and password information.

User authentication information is either using trusted authentication or explicit authentication.

Trusted authentication: the user has already satisfied corporate security when accessing network resources, and therefore the ADM Profile interface can assume that the user is trustworthy. Network trusted authentication information is gathered by the ADM Profile interface and is checked for authentication. In other words, no input is required from the user.

Explicit authentication: A user name and password is required from the user.

ADM.PROFILE TOOLS AND FUNCTIONS

CONNECTION PROCEDURE

IProfile is the interface granting access to any of the other ADM interfaces. Therefore, in order to access any of the ADM interfaces, one must first successfully create and connect an IProfile interface. For example, using Delphi,

var
pProfile, pInterface: OLEVariant;
begin
try
pProfile := CreateOLEObjects(‘ADM.Profile’);
if pProfile.Open[‘Provider=OraOLEDB.Oracle.1;Persist Security
Info=False;User ID=test;Data Source=tiger’,
‘test’, ‘scott’, ‘host’]
and pProfile.Authenticate[1, ‘’, ‘’] then
try
pInterface:= CreateOLEObject(‘ADM.xx’);
pInterface.SetProfile(pProfile)
[..]
pInterface:= unassigned;
except
[..]
end;
pProfile := unassigned;
except
[..]
end;
end;

In other words, a profile is created, and a connection is established to the Oracle ADM data store; user authentication is immediately established using the Authenticate( ) function. Note that user authentication may be either explicit or trusted. Once the profile object has been successfully initialized, this object can be communicated to other ADM interfaces as a short hand for communicating database and user profiling. Note that a profile interface must be fully defined (both connected to a data store and user authenticated) in order to be communicated to another interface (using the SetProfile( ) function).

REVIEWING AND VERIFYING ACCESS RIGHTS

The IProfile interface describes security attributes to all other ADM interfaces using the HasPermission( ) function. The HasPermission( ) function returns a permission value for some object within a class of object interface. The iDirectoryType value defines the interface; possible interfaces are Facility, User Group, User, Meta Data, Meta View and Script interfaces. The Permission Type value is defined by interface: this value must be supplied in the context of the interface, and must be legal by interface. Using the ViewProfile( ) function is very similar to the HasPermission( ), only differing to that this function returns all rights for a specific interface instance element.

DATABASE CONNECTION AND TRANSACTIONS

The Profile interface performs database transactions for all other ADM interfaces using the Execute( ) function. A SQL statement is passed as a parameter to the Execute( ) function, and if the transaction is successful, Execute( ) returns a record set from which individual field values are retrieved.

Function Reference, including the function name, description and syntax:

Open: Establishes a connection to an Oracle database. This connection must point to an IMO's Adaptive database model. Returns TRUE if connection is established, FALSE otherwise. If connection was not established, check content of LastError to get a description of the connection failure. Note: strConnection string specified OLE DB provider and necessary parameter information for the provider. For correct data retrieval, Oracle Provider for OLE DB should be used. Syntax Function Open(const strConnection, strUserName, strpassword, strhost: String): boolean

Authenticate: Authenticate the user, either using explicit or trusted authentication. Set the first parameter of Authenticate( ) to 0, and provide a user name and password to proceed to explicit authentication. Set the first parameter of Authenticate( ) to 1 to proceed to trusted authentication. In this case, note that no values are necessary for username and password parameters. Returns TRUE if user authentication succeeded. Syntax Function Authenticate(const iAuthentication: longint; sUsername, sPassword: String): boolean

HasPermission: The HasPermission( ) function returns a permission value for some object within a class of object interface. The iDirectoryType value defines the interface; possible interfaces are Facility, User Group, User, Meta Data, Meta View and Script interfaces. The PermissionType permission value is defined by interface: this value must be supplied in the context of the interface, and must be legal by interface. Possible numeric values for iDirectoryType are:

0 Undefined
1 Facility
2 User Group
3 User
4 Meta Data
5 Meta View
6 Script

Possible numeric values for PermissionType, by interface are:

	Interface	Value	Description
	Facility	0	Create
		1	Read
		2	Update
		3	Delete
	User Group	0	Create
		1	Read
		2	Update
		3	Delete
	User	0	Create
		1	Read
		2	Update
		3	Delete
	Meta Data	0	Create
		1	Read
		2	Update
		3	Delete
	Meta View	0	Create
		1	Read
		2	Update
		3	Delete
	Script	0	Create
		1	Read
		2	Update
		3	Delete

sCode is the object database code corresponding to the interface identified by the iDirectoryType value. For example, if iDirectoryType is ‘Meta Data’, then sCode will be a Meta Node Code.

Syntax Function HasPermission(Idirectory: Directory, iDirectoryType: longint, sCode: String, PermissionType: longint): OLEVariant;

ViewProfile: The ViewProfile( ) function returns all permission values for some object within a class of object interface. The iDirectoryType value defines the interface; possible interfaces are Facility, User Group, User, Meta Data, Meta View and Script interfaces. The permission value is defined by interface: this result must be interpreted in the context of the interface.

Possible numeric values for iDirectoryType are:

0 Undefined
1 Facility
2 User Group
3 User
4 Meta Data
5 Meta View
6 Script

sCode is the object database code corresponding to the interface identified by the iDirectoryType value. For example, if iDirectoryType is ‘Meta Data’, then sCode will be a Meta Node Code. Syntax function ViewProfile(Directory: Idirectory; iType longint, sCode: String): OLEVariant;

Properties: The Properties property returns properties related to the user identified as part of the Authentcate( ) method call. Possible values for Item are:

NAME       Returns the user full name
LASTNAME   Returns the user last name
FIRSTNAME  Returns the user first name
CODE       Returns the user code
LEVEL      Returns the user administrative level
USERNAME   Returns the user username
ROLES      Returns the user roles
ACTIVE     Returns TRUE is the user is active
EXTERNALID Returns an additional administrative ID for the
           user

Syntax Property Properties[Item: String]: OLEVariant;

Connected: The Connected property indicates whether IProfile interface is connected to a database. Syntax Property Connected: boolean;

Connection: The Connection property returns an ADO Connection object. This object is defined in the Microsoft ActiveX Data Objects 2.0 Library (see msado20.tlb) and is responsible for all database transactions. Syntax Property Connection:_Connection;

Close: The Close function closes the connection to the database. Syntax Procedure Close;

Execute: The Execute function executes the specified query (SQL statement) and returns resulting RecordSet. Syntax Property Execute(strSQL: string):_RecordSet;

Sequence: The Sequence function returns the next unique number in the specified sequence. Syntax Property Sequence(strSeed: String): integer;

BeginTransaction: The BeginTransaction function begins a new database transaction.

Syntax Procedure BeginTransaction;

Commit: The Commit function saves any changes and ends the current transaction

Syntax Procedure Commit;

Rollback: The Rollback function cancels any changes made during the current transaction and ends the transaction. Syntax Procedure Rollback;

ADM.DATAGATE

The ADM.Datagate COM object interface is a component interface that allows easy access to complex data stored on a relational database server, using the IMO Adaptive Database Manager. ADM.Datagate tremendously simplifies data modeling, design and access to very large amounts of data. The data model can be expanded or reduced at any time. Expanding the data will not affect current applications accessing ADM.Datagate interface services.

The data that is stored through ADM.Datagate is modeled after either trees or graphs. Any real life data can be stored using either of these formal data models into ADM.Datagate. The definition of the data to be captured is defined by the Meta data, and represents the data element hierarchy and associated attributes. The user data, accessed from or sent to the ADM.Datagate COM object, is stored into the database repository as instance data. The instance data is modeled after the Meta data.

DATA STORE MODEL

The database methodology used by the system is a free-form data model (meta data model) that stores all collected (or instance) data in the database as granular information: Many “records” are used to represent a traditional patient demographic record, or an encounter record. For example, the visit date would be granular information about an encounter (encounter class information type) and represents an attribute of the encounter. This model is adaptive to any particular situation; during implementation at a specific location, many more attributes for the “encounter class”, for example, can be defined without having to modify the structural definition of the database.

These structural definitions, or meta data, are organized in the database back-end as a tree: all granular information is stored as nodes, with pointers or edges to define the parent relationships. All user data, or instance data, are themselves patterned after meta data: meta data is the framework for creating new instance data. Therefore, this design schema is data driven, allowing for rapid and dynamic customization of the system for any new data elements that present themselves.

Complementing this design schema is a number of additional tables that store data in a traditional relational format. Security tables allow very specialized access to the system and the data stored therein. A set of archive tables provides a thorough permanent data-auditing trail.

META DATA AND INSTANCE DATA TABLES

As stated earlier, all user data within the system can be represented as a tree. The tables that make up the Meta data section of the database represent how this tree is to be stored. An example is in FIG. ##.

META_NODE

The Meta_Node table will store all of the data definitions and their relationships for the Abstractor+ database. To continue the previous example, a possible iteration (or multiple instances) of the above meta data could be represented as in FIG. ##.

The meta model may have many meta roots; It is possible to define many concurrent meta model trees, and to also have corresponding separate instance data trees in the same database. The previous example meta model tree, instance data would be represented in the database by the following tables:

NODE

The node table encapsulates all of the current data values that were gathered in the system with the relationships between these data elements as defined by the meta data. All data stored in this node table is also marked with an owner. Modifying a row in this table will mark the new data with a new owner and save the previous data value into an archive table. The client software application provides users the ability to review all instance data versions. Deletion of any instance data in the database is strictly prohibited.

META VIEW

Functionality is provided to hide the granularity of the free-form flexible database model of ADM and to enable simple access to the data enclosed into ADM. Any user, with appropriate database rights, can query the ADM data store model database using conventional tools such as Microsoft Access or Microsoft Query.

A mechanism has been created to allow users to define Meta Views constituted of meta data elements. These Meta views are created and maintained by ADM. Any change to the instance data is automatically retrofit into the Meta view structure. Refer to the ‘ADM Administration Utility User Manual’ document for more information.IDENTIFIERS

In order to simplify access to the data, any data transaction is effective using identifiers. An identifier is a string that represents any distinct data element.

Data access is possible using identifiers. An identifier specifies where to access data. For example, a hypothetical Meta data model:

Patient

Last_Name  Child of Patient element
First_Name Child of Patient element
Visits[]   Child of Patient element
Visit_Date Child of Visits Array element
Etc..

Each patient created into this model will be treated as a Patient element. Each Patient element will have a Last_Name, a First_Name and many patient Visits. Patient is the root for all patient related elements.

One instance of the Datagate object only refers to one instance data root at the same time, PATIENT begins the one in our example. Accessing the patient's last name can be done by specifying patient.last_name. The Visit element is an array of elements. Accessing the date of the first visit would be patient.visit[0].visit_date.

IDENTIFIER REFERENCE

An identifier is a reference to a specific meta data layer element. An identifier consists of one or many identifier terms separated by a period (‘.’). Each identifier term in turn is a leaf element of the instance data tree or the root for a specific branch of the instance data tree. For example,

patient.last_name.

Patient is the top root of the tree, and Last_Name is a child element inside the meta data tree of elements. patient.visit[c].admit_date

Patient is the top root of the tree, Visit[ ] is an array element where the c index indicates the current branch, and Visit_Date is a leaf element.

patient.med_rec_no[c].visit[c].diagnosis[0]code

Patient is the top root of the tree, Visit[ ] and Diagnosis[ ] are array elements when the c index indicates the current branch index, Diagnosis[0] is branch 0 of the Diagnosis element array, and code is a leaf element.

Identifier abbreviations can be used, as Datagate will interpret these abbreviations at run time. Using the previous examples, the identifiers can be rewritten as:

patient.last_name

visit[c].admit_date or even admit_date

diagnosis[0].code

Identifier items can be of leaf type (thus string, integer, boolean, date, real, etc.) or array type. When referring to array identifier items, one must also specify the array argument. Array arguments are of several types:

‘c’: The current record of the array. If the array is Visit[ ], then Visit[c] is the current visit. Note that the index of an array can be accessed and set using the indexOf( ) function. The c record index can be offset with an integer value. For example,

• • visit[c].visit_date
  • diagnosis[c+1].code

‘n’: The last record of an array. If this array is empty then an empty value will be returned. The n record index can be offset with an integer value as well. For example,

• • visit[n].admit_date

an integer: All array records can be retrieved individually by an integer. If an array record is requested and that record has not yet been defined, then it will be created, as well as any intermediary records to make up that record index. For example,

• • visit_dx[0].visit_dx_medcode

DATATYPES

Every meta data element has a data type, which specifies a storage format, constraints, and valid range of values. A variety of predefined data types are provided. A scalar type has no internal components. An array type has internal components that can be manipulated individually. The following figure shows the predefined data types.

	Type	Storage Family	Data type
	Scalar	Number	Integer
			Real
		Character	String
			LongString
		Boolean	Boolean
		Date & Time	Date
		Binary large object	Blob
	Array	Array	Array
			Virtual Array

NUMBER TYPES

Number types allow you to store numeric data (integers and floating-point numbers), represent quantities, and perform calculations.

INTEGER

The INTEGER data type is used to store signed integers. Its magnitude range is −2147483647 . . . 2147483647.

FLOAT

The FLOAT data type is used to store fixed or floating-point numbers of virtually any size. Its magnitude range 5.0×10^−324 . . . 1.7×10^308. The maximum precision of a FLOAT value is roughly equivalent to 15 decimal digits.

CHARACTER TYPES

Character types allow you to store alphanumeric data, represent words and text, and manipulate character strings.

STRING

The STRING data type is used to store character data. The maximum amount of data that can be stored into a STRING data type is 255 characters.

LONG STRING

The LONG STRING data type is used to store (nearly) unlimited-size character data. The maximum amount of data that can be stored into a LONG STRING data type is 4294967294 characters.

BOOLEAN TYPES

The following types allow you to store and manipulate logical (true, false) values and date/time.

BOOLEAN

The BOOLEAN data type is used to store the logical values TRUE and FALSE. Only logic operations are allowed on BOOLEAN variables.

DATE

The DATE data type is used to store fixed-length date/time values. DATE values also include the time of day. The integral part of a DATE value is the number of days that have passed since Dec. 30, 1899. The fractional part of a DATE value is the fraction of a 24-hour day that has elapsed.

BINARY OBJECTS

Binary objects can be stored using the BLOB (Binary Large Object) data type.

ARRAY TYPE

The Array data type is the data type allowing for building lists of values. Values can be a single scalar data type, multiple scalar data types thus making a record of scalar and array data types. This is the essential building block for complex business model data storage.

The virtual array data type is the data type allowing to build arrays of references of elements belonging to other arrays. For example:

Patient.Visit[ ].Medication[ ]: Medication[ ] is a an array of medications for the patient for each visit.

Patient.All_Medications[ ]: All_Medications[ ] could be set as a virtual array of all patient medications, regardless of the visits. Virtual Arrays are especially designed for creating summary lists, as a collection of only some array elements of another array Meta data node.

TYPE TO META DATA CODING

Data types are coded and accessed as numbers inside the Meta model. The following table provides the mapping between formal data types and data type number values.

Data type    Number
Undefined    0
String       1
Integer      2
Float        3
Date         4
Long String  5
Boolean      6
Array        7
Array Record 8

Virtual Array

ADM.Datagate Tools and Functions

ADM.Datagate provides a set of functions for accessing and navigating inside atomic data as well as accessing data as a whole. Note that most functions will return data as variants, so that application data typing may be less of an issue.

ACCESSING DATAGATE

IDatagate is a COM object interface. IDatagate relies on a previously declared IProfile interface to establish a connection to the datastore. For example, using Delphi,

var
pDoc, pProfile: OLEVariant;
begin
try
pProfile := CreateOLEObjects(‘ADM.Profile’);
if pProfile.Open[‘Provider=OraOLEDB.Oracle.1;Persist Security
Info=False;User ID=test;Data Source=tiger’,
‘test’, ‘scott’, ‘host’] and
pProfile.Authenticate[1, ‘’, ‘’] then
try
pDoc := CreateOLEObject(‘ADM.DATAGATE’);
pDoc.SetProfile(pProfile)
[..]
pDoc := unassigned;
except
[..]
end;
pProfile := unassigned;
except
[..]
end;
end;

MANAGING INSTANCE DATA

Data elements are retrieved from the database by using the Open( ) function. Closing access to the data elements is set by using the Close( ) function. Saving data element changes is achieved by using the Save( ) function.

For example using Delphi:

• • [..]
  • b :=pDoc.Open[iNodeCode];
  • [..]
  • pDoc.Close;
  • [..]

In this example, the Boolean variable b indicates if loading the pDoc Datagate object with the iNodeCode root code was successful. Note that iNodeCode indicates an instance data root nod code.

ACCESSING AND SETTING ELEMENT VALUES

VALUE, SETLOCK, CLOSELOCK, ISREADONLY

Data elements are accessed and set using identifiers through the value( ) function. For example,

Using Delphi:

• • Var a: olevariant;
  • [..]
  • A :=pDoc.Value[‘patient.first_name’];
  • pDoc.Value[‘patient.first_name’]:=‘Doe’;

Using Active Server Pages:

• • Dim A
  • A=pDoc.Value(“patient.first_name”)
  • pDoc.Value(“patient.first_name”)=‘Doe’

A data element can only be set if it is in Read/Write mode. Use the SetLock( ) function to be able to write to data elements. The SetLock( ) function will return TRUE if the element was effectively locked. To check if a data element is Read Only, use the IsReadOnly( ) function.

INDEXOF, COUNT, CURRENTINDEX, ORDER

Access through the data element is made using identifiers. To set the current data array element, use the IndexOf( ) function. To know the size extent of an array identifier, use the Count( ) function. Note that the count function will return 1 for a non-array identifier. For example,

Using Delphi:

• • pDoc.IdexOf[‘patient.visit[0].DX[ ]’]:=1;

Using Active Server Pages:

• • pDoc.IndexOf(“patient.visit[0].DX[ ]”)=1

After setting the current array subscript using the IndexOf( ) function, it is possible to use a syntax referring to the current subscript. For example:

Using Delphi:

• • pDoc.IndexOf[‘patient.visit[0].DX[ ]’]:=1
  • pDoc.Value[‘patient.visit[0].DX[c]’]:=‘Heart Failure’;

Using Active Server Pages:

• • pDoc.IndexOf(“patient.visit[0].DX[ ]”)=1
  • pDoc.Value(“patient.visit[0].DX[c]”)=‘Heart Failure’

DX[c] is actually making reference to DX[1]. It is possible to set all the current subscripts at once using the CurrentIndex function. For example,

Using Delphi:

• • pDoc.CurrentIndex :=‘patient.visit[2].DX[5]’;

Using Active Server Pages:

• • pDoc.CurrentIndex=“patient.visit[2].DX[5]”

In other words, the current index for User.Visit[ ] is 2, and the current index for User.Visit[2].DX[ ] is 5.

It is possible to change the order of subscripts within an array, using the Order( ) function. For example, one could set that patient visit 2 should come after patient visit 5 by writing:

Using Delphi:

• • pDoc.Order[‘patient.visit[2]]:=5;

Using Active Server Pages:

• • pDoc.Order(“patient.visit[2]”)=5

DELETE

The Delete( ) function allows for element management, by allowing deletion of any element. Note that the parent of the element being deleted must be locked, i.e. into a read-write state. Refer to the SetLock( ) function for more information.

CREATING AND ADDING NEW ELEMENT VALUES

VALUE

A whole new tree can be made by creating a new instance of the ADM.DATAGATE object, establishing a connection, and starting to assign values to the new instance pointer using the Value[ ] property. For example,

Using Delphi:

Var
pDoc, pProfile: OLEVariant;
[..]
pDoc := CreateOLEObject(‘ADM.DATAGATE’);
if pDoc.SetProfile(pProfile) then
[..]
pDoc.Value[‘user.last_name’] := ‘Doe’;
[..]
pDoc.Save;
[..]

Using Active Server Pages:

dim pDoc
dim pProfile
[..]
pDoc = CreateOLEObject(“ADM.DATAGATE”)
if pDoc.SetProfile(pProfile) then
[..]
pDoc.Value(“user.last_name”) = “Doe”
[..]
pDoc.Save
[..]
ADD

New elements are created using the Value[ ] property and the Add( ) function. A new array record is created using the Add( ) function. This presuppose that the array is in read/write mode, i.e. a call to SetLock( ) was previously made. For example,

Using Delphi:

• • Var L: longint;
  • [..]
  • L:=pDoc.Add[‘user.visits[ ]’];

Using Active Server Pages:

• • Dim L
  • L=pDoc.Add(“user.visits[ ]”)

The returned value, L, is the subscript for the new array record. Note that the identifier must be an array, and thus, must have the ‘[ ]’ notation.

SET OPERATIONS

XML

Data elements can be accessed or set, either in parts or in entirety, using the XML property.

SEARCHING THE ADAPTIVE META MANAGER

FIND, FINDKEYWORD, OPTIONS, NODEINFORMATION

ADM.Datagate instance data can be queried using the Find, FindKeyword, Children and NodeInformation functions. Find( ) or FindKeyword( ) are the functions initiating the query, FindFirst( ) returns the first result and FindNext( ) returns all subsequent results. FindFirst( ) may be called again to loop through the result set without calling Find( ) or FindKeyword( ) again. Refer to following respective sections for Children, NodeInformation, FindFirst and FindNext functions.

FIND

The Find( ) function allows for simple element queries, such as ‘user last name is Doe and user first name is John’. This function allows for any combination of Boolean operators and parentheses to support such queries. The syntax for the Find( ) search argument is as follows:

• • statement ->bexpr
  • bexpr ->bterm {‘or’ bterm}
  • bterm ->bfactor {‘and’ bfactor}
  • bfactor ->‘(‘bexpr’)’|MetaID relop factor|MetaID|‘START’ ‘WITH’ number
  • relop ->‘<’|‘<=’|‘<>’|‘>’|‘>=’|‘=’|‘like’
  • factor ->number|String|MetaID

In other words, one can search for all instance data trees that have a Last_Name defined:

• • Find[‘LAST_NAME’]. . or find all records where the last name is “Doe”:
  • Find[‘LAST_NAME=“DOE”’]. . or find all records with a last name where the first name is “John”:
  • Find[‘LAST_NAME and FIRST_NAME=“JOHN”’]

A more complete example could be as such:

Using Delphi:

..// pDoc is an olevariant, has been previously created, and has been
assigned a profile object.
[..]
if pDoc.Find[‘LAST_NAME = “DOE” AND FIRST_NAME=
“JOHN”] > 0 then
begin
if pDoc.FindFirst[i] then
repeat
// the value of ‘i’ is the top root node satisfying the query
until not pDoc.FindNext[i];
end;

Using Active Server Pages:

..// pDoc is an olevariant, has been previously created, and has been
assigned a profile object.
..[..]
if (pDoc.Find(“LAST_NAME = ””DOE”” AND FIRST_NAME=
””JOHN”” ”) > 0)
and—
(pDoc.FindFirst(i)) then
do
// the value of ‘i’ is the top root node satisfying the query
[..]
loop until not pDoc.FindNext(i)
end if

These examples show a query initiated using the Find( ) method. The Find( ) function returns the number of elements matching the criteria set. The first element is retrieved using the FindFirst( ) method. Note that FindFirst( ) returns a Boolean value to indicate if the call to FindFirst( ) was successful. After making use of this value, FindNext( ) is called to retrieve the next element. Note again that FindNext( ) returns a Boolean value to indicate if the call to FindNext( ) was successful. FindNext( ) will return FALSE when the last entry has been retrieved using FindNext( ).

‘START WITH’ EXPRESSION

It is sometimes necessary to perform queries within a range of elements, such as the current data gate element, for example. Such element usually identifies the top of the tree (or sub-graph) where search results are meaningful. This can be achieved using the ‘START WITH’ expression, and specifying a node code being the top of the local sub-tree or sub-graph.

FINDKEYWORD

The FindKeyword( ) function allows for keyword or concept searches for a specified portion of the data store or for the whole data store. For example, if the Adaptive Data Manager data store has data elements about users and media articles, then one could look for all articles making references to ‘Chest’ and ‘Pain’, within the ‘Media[ ]’ branch. Note that the search is case insensitive. FindFirst( ) returns the first result and FindNext( ) returns all subsequent results. FindFirst( ) may be called again to loop through the result set without calling FindKeyword( ) again.

Several options are available to broaden or qualify the search. Refer to the Options[ ] property for more information.

The FindKeyword( ) can be provided with a portion of the data element tree as a starting point and will return all matching elements within that portion of the tree, or can be provided with no specific starting point and thus will query the whole data store.

Specifying a portion of the data store tree is similar to searching across all xxx, of the tree must have been fetched using the open( ) function prior to executing a search using the FindKeyword( ) function A relevant example could be:

Using Delphi:

Var
b: boolean;
i: olevariant;
[..]
if pDoc.Open([..]) then
[..]
if pDoc.FindKeyword[‘Media[]’, ‘CHEST PAIN’] > 0 then
begin
pDoc.FindFirst[i];
repeat
// the value of ‘i’ is the top root node satisfying the query
until not pDoc.FindNext[i];
end;

Using Active Server Pages:

Dim pDoc, i
[..]
if pDoc.Open([..]) then
[..]
if (pDoc.FindKeyword(“Media[]”, “CHEST PAIN”) > 0) and—
(pDoc.FindFirst(i)) then
do
// the value of ‘i’ is the top root node satisfying the query
[..]
loop until not pDoc.FindNext(i)
end if

In both examples, an instance of ADM.Datagate (pDoc) is created and receives a connection to the data store. The pDoc instance is then loaded (open( )) to receive Media[ ] information, and the FindKeyword( ) function is executed to find all media articles that have the words chest and pain as keywords or concepts. Note that FindKeyword( ) returns the number of elements matching the criteria set. The first element is retrieved using the FindFirst( ) method. Note that FindFirst( ) returns a Boolean value to indicate if the call to FindFirst( ) was successful. After making use of this value, FindNext( ) is called to retrieve the next element. Note again that FindNext( ) returns a Boolean value to indicate if the call to FindNext( ) was successful. FindNext( ) will return FALSE when the last entry has been retrieved using FindNext( ).

OPTIONS

Results from the FindKeyword( ) function can be tuned to return a range of results using the Options[ ] property. For example, a search can also include close enough terms using the FUZZY option. All entries similar through Soundex algorithm to the original entry will be returned as well. The Option( ) property is set or accessed as a string.

For example, using Delphi:

pDoc.Options[‘FUZZY’] := ‘TRUE’;
if pDoc.FindKeyword[‘users’, ‘JOHN DOE’] > 0 then
[..]
Active Server Pages:
pDoc.Options(“FUZZY”) = “TRUE”
if pDoc.FindKeyword(“users”, “JOHN DOE”) > 0 then
[..]

CHILDREN

The Children function is an interface function similar to Find( ) and FindKeyword( ): a search operation is scheduled, and results are retrieved using FindFirst, FindNext and optionally FindResult. Given a node code, Children( ) returns all child node codes. This function is useful when navigating any element tree, regardless of the semantics of the data. This function is especially useful in conjunction with FindResult( ): the administration utility is making extensive use of this feature.

For example, using Delphi:

if pDoc.Children[NodeCode] > 0 then
if pDoc.FindFirst[iNode] then
repeat
[Use pDoc.FindResult[] to retrieve ‘METATYPE’,
‘METANODENAME’,
‘ARRAYINDEX’, ‘ARRAYINDEX’, ‘HASCHILDREN’, etc for display]
until not objDatagate.FindNext[iNode];

In this example. iNode contains the node code of the i^th child node of ‘NodeCode’.

NODEINFORMATION

The NodeInformation function is an interface function similar to Children( ): a search operation is scheduled, and results are retrieved using FindResult( ). Given a node code, NodeInformation( ) returns TRUE if such a node exists, FALSE otherwise. In many instances, a great deal more than a node code is required to proceed with the data element. Hence, NodeInformation( ) should be used in conjunction with FindResult( ), as FindResult function provides most of the necessary information. For more details regarding the use of the FindResult function, refer to its section.

For example, using Delphi:

if pDoc.NodeInformation[NodeCode] then
begin
sCode := pDoc.FindResult[‘METANODECODE’];
sName := pDoc.FindResult[‘METANODENAME’];
end;

FINDFIRST, FINDNEXT, FINDRESULT

FindFirst and FindNext function allow for data retrieval after performing a call to Find, FindKeyword or Children functions. All Find, FindKeyword and Children return a count as the result of the query. If this result is 0, then the query operation did not return any result. If this count is greater than 0, then calls to FindFirst and FindNext are required. Typical code example for using FindFirst and FindNext is as follows:

For example, using Delphi:

If pDoc.FindKeyword[‘users’, ‘JOHN DOE’] > 0 then
if pDoc.FindFirst[iNode] then
repeat
// iNode is the code of the element found
[..]
until not objDatagate.FindNext[iNode];
[..]

FindFirst and FindNext return TRUE if the data retrieval operation was successful, FALSE otherwise. In many instances, a great deal more than a node code is required to proceed with the data element. The FindResult( ) function provides most of the necessary information. After a FindFirst or FindNext, one can call the FindResult to find out about the data element value, meta node code and name, array index, data type, and if the data element has children. Possible values for the FindResult unique parameter are:

Parameter    Description
MetaNodeCode Number The meta node code corresponding to the
             element
MetaNodeName String The meta node name corresponding to the
             element
ArrayIndex   Number Variant value - Empty if the element is not
             an array element, or a number otherwise.
Value        String The data value of the element. Contains the
             string value of the element if the data type is not a
             long, otherwise contains ‘<long>’.
MetaType     Number Contains the data type of the element.
             Refer to the DATATYPES section to translate this
             number value to a string.
HasChildren  Number - 0, if not children, greater than 0
             if any child(ren)

The function result is returned as a variant. For example,For example, using Delphi:

if pDoc.FindFirst[iNode] then
repeat
varA := FindResult[‘METANODECODE’];
varB := FindResult[‘ARRAYINDEX’];
varC := FindResult[‘VALUE’];
varD := FindResult[‘METATYPE’];
varE := FindResult[‘METANODENAME’];
varF := FindResult[‘HASCHILDREN’];
[..]
until not objDatagate.FindNext[iNode];
[..]

MISCELLANEOUS

LASTERROR

The LastError( ) function reports the last errors that intervened during the calling application execution. Calling LastError( ) function will reset the error list.

ISA

This function accepts two parameters: 1) the string name of a meta node, and 2) a node code. The IsA( ) function will return true if the second parameter meta node code is either the meta node mentioned as first parameter, or a child meta node of the meta node mentioned as first parameter, otherwise returns false. For example, given the following meta node description:

• • A Top meta node
  • . .B Child of A
    • D Child of B
  • . .C Child of A

For a node n of meta node C, pDoc.IsA(‘A.C’, n) is true, pDoc.IsA(‘A’, n) is true and pDoc.IsA(‘A.B’, n) is false.

For a node n of meta node B, pDoc.IsA(‘A.B.D’, n), pDoc.IsA(‘A.B’, n), pDoc.IsA(‘A’, n) all return true, while pDoc.IsA(‘C.’, n) or pDoc.IsA(‘A.C’, n) return false.

Function Reference

SetProfile
Description Function
            Set a connection between the Profile object and
            current IDatagate new instance.
Syntax      Procedure SetProfile(Profile: IProfile);
Example     DELPHI:
            var
            pProfile, pDoc: OLEVariant;
            begin
            try
            pProfile := CreateOLEObjects(‘ADM.Profile’);
            if pProfile.SetConnection[‘test’, ‘scott’, ‘tiger’,
            ‘host’] and
            pProfile.Authenticate[1, ‘’, ‘’] then
            try
            pDoc:= CreateOLEObject(‘ADM.Datagate’);
            pDoc.SetProfile(pProfile)
            [..]
            pDoc:= unassigned;
            except
            [..]
            end;
            pProfile := unassigned;
            except
            [..]
            end;
            end;
Open
Description Function
            Initializes the COM objects with the data stored into
            the Adaptive Data Manager starting at root element
            intNodeCode.
            Note that not all data is loaded: This object was
            designed using a ‘load on demand’ principle, where
            data is only fetched from the database when
            requested for the first time.
            Note that a connection must have been established
            before calling Open. All loaded data is by default
            Read Only. Some data may have been previously
            locked, and consequently may be set as Read &
            Write.
            Returns TRUE if successful, FALSE otherwise. If
            unsuccessful, possible reasons are failed database
            connection, invalid node code. Check the LastError
            function content for the cause of failure.
Syntax      function Open(intNodeCode: longint): boolean
Example     This example shows all steps to opening a record
            within ADM. The value 123456 is a node code, and
            is a the code for a root node code. Typically this
            value will be returned as part of a search
            (using Find or FindKeyword functions).
            DELPHI:
            var
            pProfile, pDoc: OLEVariant;
            begin
            try
            pProfile := CreateOLEObjects(‘ADM.Profile’);
            if
            pProfile.Open[‘Provider=OraOLEDB.
            Oracle.1;
            Persist Security Info=False;User ID=test;
            Data
            Source=tiger’, ‘test’, ‘scott’, ‘host’] and
            pProfile.Authenticate[1, ‘’, ‘’] then
            try
            pDoc:= CreateOLEObject(‘ADM.Datagate’);
            pDoc.SetProfile(pProfile)
            if pDoc.Open(123456) then
            begin
            [..]
            end;
            pDoc.Close;
            pDoc:= unassigned;
            except
            [..]
            end;
            pProfile := unassigned;
            except
            [..]
            end;
            end;
Close
Description Procedure
            Resets the data content of the IDatagate instantiated
            interface. Note that all locks that may have
            previously set on the data content will not be affected. Refer to CloseLock
            for removing locks. Any elements that may have been modified will
            not be saved to the data store.
            This function always returns TRUE.
Syntax      Procedure Close;
Save
Description Function
            Saves the data content of the IDatagate instantiated
            interface. Only non Read/Only elements (i.e. only
            elements that were successfully locked) will be saved
            to the data store.
            Returns TRUE if successful, FALSE otherwise. If
            unsuccessful, check the LastError function content
            for the cause of failure.
            Note that all locks are not removed after the Save
            operation.
Syntax      function Save: boolean;
Delete
Description Function
            This function allows elements of opened records to be
            deleted, by specifying an identifier; all elements child
            of that identifier will be deleted as well. The identifier
            is any legal identifier. Note that the parent of the
            mentioned identifier must be read/write state: refer to
            the SetLock( ) function.
            This function returns TRUE if successful, FALSE
            otherwise. If unsuccessful, check the LastError
            function content for the cause of failure.
Syntax      function Delete(strIdentifier: String): boolean;
SetLock
Description Function
            Sets a database lock on the database node
            corresponding to the identifier. If successful, the
            identifier becomes Read & Write. Note that all
            identifier child elements are locked as well. The lock
            is based on the owner and session's code, thus
            allowing the same owner (i.e. user) to lock different
            elements for different sessions. Typically, the session
            will either be an IIS web server session ID, or an
            Oracle session ID. The owner and session are
            established when creating the parent IProfile
            interface, that is later communicated to the current
            IDatagate interface..
            This function returns TRUE if successful, FALSE
            otherwise. If unsuccessful, check the LastError
            function content for the cause of failure.
Syntax      function SetLock(strIdentifier: String): boolean;
CloseLock
Description Function
            Closes an identifier lock. The identifier data is reset
            Read Only mode. This function returns TRUE if
            successful, FALSE is the lock could get established.
            Setting a lock will be unsuccessful if another user
            already has a lock the requested data element. Check
            the LastError function content for the cause of failure.
            Check current locks using the ADM Administration
            Utility, Lock Tab, to list existing locks.
Syntax      function CloseLock(strIdentifier: String): boolean;
Find,
FindKeyword,
Description Function
            The Find function allows for queries over the entire
            Adaptive database model. For example, one could
            query all node trees where node of meta node type
            ‘Last Name’ is ‘Doe’, and node of meta node type
            ‘First Name’ is ‘John’. The logic of this function is
            based on sets, and thus allows for set intersections,
            differences and unions.
            The FindKeyword( ) function allows for keyword or
            concept searches for a specified portion of the data
            store or for the whole data store. For example, if the
            Adaptive Data Manager data store has data elements
            about users and media articles, then one could look
            for all articles making references to ‘Chest’ and
            ‘Pain’, within the ‘Media[]’ branch. Note that the
            search is case insensitive
            Find( ) and FindKeyword( ) functions will prepare the
            query and return the number (count) of entries that
            matched the query. The FirstFirst( ) and FindNext( )
            functions will respectively return the first entry and
            the following entries. Note that one can loop through
            all results several times using FindFirst to reset the
            result index without having to re-submit Find.
            Check the Options( ) function for possible behavior
            options.
            Note that only data saved into the data store will
            be queried using this method. In order to have the
            data stored in memory also queried, Save the data
            first before performing a query.
Syntax      function Find(const strLookup: WideString): Integer;
            function FindKeyword(const strIdentifier, strLookup:
            WideString): Integer;
Example     DELPHI:
            Var
            b: boolean;
            i: olevariant;
            [..]
            if pDoc.Find[‘LAST_NAME = “YOUNG” AND
            FIRST_NAME=“ANDRE”’] > 0 then
            begin
            pDoc.FindFirst[i];
            repeat
            // the value of ‘i’ is the top root node tisfying the
            query
            until not pDoc.FindNext[i];
            end;
Children
Description Function
            Return the list of child elements for the ‘node’ parent
            element. The value for node is a node.node_code.
            The Children( ) function will prepare the query and
            return the number (count) of child elements for
            ‘Node’. The FirstFirst( ) and FindNext( ) functions will
            respectively return the first entry and the following
            entries. Note that one can loop through all results
            several times using FindFirst( ) to reset the result
            index without having to re-submit Children( ).
            Check the Options( ) function for possible behavior
            options.
Syntax      Function Children(Node: variant): longint
NodeInformation
Description Read-Only Property
            This function allows returning database information
            about an existing node.
Syntax      property Identifier[intNodeCode: Integer]: String
FindFirst,
FindNext
Description Function
            Allows retrieval of node elements after Find( ),
            FindKeyword( ) or Children( ) functions. FindFirst( )
            and FindNext( ) return TRUE if the operation was
            successful, FALSE otherwise. If unsuccessful, check
            the LastError function content for the cause of failure
Example     DELPHI:
            var
            n, i: longint;
            [..]
            n: = pDoc.Children[‘PATIENT.VISIT[]’];
            if pDoc.FindFirst[I] then
            repeat
            // the value of ‘i’ is the top root node satisfying the
            query
            until not pDoc.FindNext[i];
Syntax      Function FindFirst(out intRootCode: OLEVariant):
            WordBool;
            Function FindNext(out intRootCode: OLEVariant):
            WordBool;
FindResult
Description Function
            The FindResult( ) function is used in conjunction with
            FindFirst( ) and FindNext( ) functions. FindFirst( ) and
            FindNext( ) functions only return a node code; the
            FindResult( ) function allows to get additional
            information to the code returned by FindFirst( ) and
            FindNext( ).
            METANOD  Number The meta node
            METANOD  String The meta node
            ARRAYIN  Number Variant value -
                     number otherwise.
            VALUE    String The data value of
                     the element. Contains the
                     string value of the
                     element if the data type is
                     not a long, otherwise
                     contains ‘<long>’.
            METATYPE Number Contains the data
                     type of the element. Refer
                     to the DATATYPES section
                     to translate this number
                     value to a string.
            HASCHIL  Number - 0, if not
            DREN     children, greater than 0 if
                     any child(ren)
Syntax      Function FindResult(Entry: OLEVariant):
            OLEVariant;
Options
Description Property
            The Options property allows Datagate options to
            be set or accessed. Current Options are:
            ‘FUZZY’: Allows Soundex search mode for
            FindKeyword function.
            ‘MAX CHILDREN’: Limits the maximum
            number of children that can be retrieved using the
            Children( ) function.
Syntax      Property Options(const strValue: WideString):
            OleVariant;
SessionID
Description Read-Only Property
            Returns the SessionID value previously set by the
            Open function. If used previously to any Open
            function call, returns −1.
Syntax      Function SessionID: longint;
Owner
Description Read-Only Property
            Returns the Owner value previously set by the Open
            function. If used previously to any Open function
            call, returns −1.
Syntax      Function Owner: longint;
XML
Description Read & Write Property
            This property allows for all data for the current root
            node to be exported as XML, or for a block of XML
            to be imported either under the current root node or
            under any node element (use the sStart value to
            specify the start node; leave this value to start under
            the root node). Check the LastError content for
            possible failure.
Syntax      Property XML[sStart: String]: String;
Value
Description Read & Write Property
            This property allows setting or getting a value for an
            identifier.
            Note that the transferred value is a variant, and thus
            may cause conversion errors. Check the LastError
            content for possible failure.
Syntax      Property Value[strIdentifier: String]: Variant
Add
Description Read-Only Property
            This property allows for next array records to be
            created into an array. This property returns the index
            of the next created array record. The identifier
            specified into the function must be an array (thus
            have the [ ] notation.)
            This function will return −1 in case of failure; in this
            case, check the content of LastError.
Syntax      Property Add[strIdentifier: String]: integer
Count
Description Read-Only Property
            This property returns the number of child nodes if the
            element specified by strIdentifier is an array. If the
            identifier element has a discrete data type (string,
            integer, real, date, or boolean) then accessing this
            property will cause an exception. Check the LastError
            content for possible failure.
Syntax      Property Count[strIdentifier: String]: longint
IsA
Description Read-Only Property
            This property checks if the instance data referred to
            by intNodeCode is of type meta data strIdentifier.
            This function returns TRUE if the operation was
            successful, FALSE otherwise. Check the LastError
            content for possible failure.
Syntax      function IsA(const strIdentifier: WideString;
            intNodeCode: Integer): WordBool;
IsReadOnly
Description Read-Only Property
            This property returns TRUE if an identifier is Read &
            Write, FALSE if Read Only. Refer to the Lock
            function for setting an identifier as Read & Write.
            Check the LastError content for possible failure.
Syntax      Property IsReadOnly[strIdentifier: String]: boolean
IndexOf
Description Read & Write Property
            This property sets or gets the current index for an
            array node. Check the LastError content for possible
            failure.
Syntax      Property IndexOf[strIdentifier: String]: longint
Datatype
Description Read-Only Property
            This property returns the data type of a node as an
            integer. Possible values are number for String,
            Integer, Real, Date, Long String, Boolean, Array,
            Array Record and Virtual Array. Note that the meta-
            data data type cannot be altered through this
            interface, and can only be managed using the ADM
            Administration Utility.
            Check the LastError content for possible failure.
Syntax      Property Datatype[strIdentifier: String]: longint
LastError
Description Read-Only Property
            This property returns the error log as text. Calling this
            function resets the error log content. All errors met
            until the LastError function call are part of the error
            log.
Syntax      Property LastError: Variant
CurrentIndex
Description Write-Only Property
            This property sets the identifier specified as the
            argument of this function to be the current index.
            Check the LastError content for possible failure. Note
            that this function was designed to work in
            conjunction with the TopNode( ) and Identifier( )
            functions.
Syntax      property CurrentIndex: String
Identifier
Description Read-Only Property
            This function returns a fully defined identifier given
            the node code specified as unique parameter. This
            function is especially useful for indexing and
            searching purposes.
Syntax      property Identifier[intNodeCode: Integer]: String
TopNode
Description Read-Only Property
            For any node element code, this function returns the
            top element node code. This function is usually used
            by passing the node element code obtained from
            FindFirst( ) or FindNext( ). from Check the LastError
            content for possible failure.
Syntax      property TopNode[intNodeCode: Integer]: Integer
Code
Description Read-Only Property
            This function returns the node element associated
            with any identifier. This function will return −1 if the
            element has not been saved to the database yet.
Syntax      Property Code[strIdentifier: String]: OleVariant
Order
Description Read / Write Property
            This function sets and returns the current subscript
            order of an array element. If the new subscript value
            is out of legal array bound, then no action is taken.
Syntax      property Order[strIdentifier: WideString]: Integer
IdentifierItem,
IdentifierItemCount
Description Read Only Property
            These functions are ‘helper’ functions and allow for
            identifier string manipulations. For any legal
            identifier strings, the IdentifierItemCount( ) function
            returns the number of segments within the string, and
            IdentifierItem( ) returns the segment as a string.
Syntax      Function IdentifierItem(strIdentifier: String; Index:
            longint): String;
            Function IdentifierItemCount(strIdentifier: String):
            longint;

ADM.Index

ADM.Index is the interface, supporting document indexing, under the Adaptive Data Manager paradigm. Indexing is declared at the meta data level in the form of indexing flags, and any creation of corresponding instance data will trigger the creation of entries into a ‘request for data processing’ queue. This queue lists all ADM.Datagate elements that are waiting for indexing. Indexing consists of analyzing each pending element and computing or establishing the relevant set of keywords and/or concepts that best describe the element. The resulting keywords will be sent to an index storage that will be accessed in turn by the ADM.Datagate interface.

This interface provides two basic functions:

Access to the data processing queue.

ADM.Index provides all means to query the data processing queue, lock entries for processing and remove from the queue processed items.

Access to the indexing storage.

ADM.Index provides all means to query existing concepts and keywords, add new concepts and keywords, or clear existing entries.

ADM.INDEX TOOLS AND FUNCTIONS

QUEUE MANAGEMENT

QUEUECOUNT, QUEUE, LOCKQUEUE, UNLOCKQUEUE, CLOSEQUEUE,

QUEUEVALUE

The Queue and QueueCount provide the means to browse existing queue entries. A queue entry contains a reference to a particular element from the ADM.Datagate storage, as well as a creation date, and possibly a reference to a user processing this entry. The QueueValue function retrieves the stream of text that is queued to be concept/keyword indexed. Eventually, the LockQueue, UnlockQueue and CloseQueue functions provide queue management: LockQueue provides exclusive access to a queue element, UnlockQueue releases the exclusive access, while CloseQueue removes the element from the queue.

The current state of the queue is buffered internally into the ADM.Index object. Each time an access is made to QueueCount, the buffer is refreshed.

A typical example of use of these functions would be as follows:

For example, using Delphi:

// pProfile Created
[..]
pIndex := CreateOLEObject(‘ADM.INDEX’);
pIndex.SetProfile(pProfile)
for i := 0 to pIndex.QueueCount 1 do
if pIndex.Queue[i, iNode, iUserCode, datCreated, intState] and
pIndex.LockQueue[iNode, kiUSER] then
begin
if pIndex.QueueValue[iNode, ws] then
[..] // Processing happen here for making keywords.
pIndex.CloseQueue[iNode, kiUSER];
end;
pIndex := unassigned;

In other words, this example shows the queue to be opened (SetProfile( ), QueueCount( )) and browsed (Queue( )), each queue element is locked for exclusive use (LockQueue( )), the value of a queue element is accessed (QueueValue( )), and the queue element is removed from the queue (CloseQueue( )).

WORDINDEXCOUNT, WORDINDEX, CLEARWORDINDEX, ADDWORDINDEX

The WordIndex( ) and WordIndexCount( ) functions provide the means to browse existing index entries for a specific ADM.Datagate element. Note that the list of index keywords and concepts for a specific element is buffered internally. This buffer is refreshed each time the WordIndexCount( ) function is accessed. An index entry contains a reference to a particular element from the ADM.Datagate storage, as well as a source and a text. The source identifies what concept domain was used to create the index text entry. For example, valid domains are FREETEXT, ICD, CPT, etc. The ClearWordIndex( ) function removes existing entries for a ADM.Datagate element, while AddWordIndex( ) creates new index entries.

MISCELLANEOUS

LASTERROR

The LastError( ) function reports the last errors that intervened during the calling application execution. Calling LastError( ) function will reset the error list.

SETPROFILE

Instantiating the COM object is the first step to being able to data access. The second step consists in establishing a connection to the data store, i.e. the relational database storage. This is effective using the SetProfile( ) function, by connection this instance of the IIndex interface to a previously created instance of IProfile interface.

Function Reference

QueueCount
Description Read-Write Property
            Returns the current number of data processing
            queue entries.
Syntax      property QueueCount: Integer;
Queue
Syntax      property Queue[intIndex: Integer; out
            intNodeCode: Integer; out intUserCode:
            Integer; out datCreated: OleVariant; out
            intState: OleVariant]: boolean
LockQueue,
UnlockQueue,
CloseQueue
Syntax      property LockQueue[intNodeCode: Integer;
            intUserCode: Integer]: boolean
            property UnlockQueue[intNodeCode Integer;
            intUserCode: Integer]: boolean
            property CloseQueue[intNodeCode: Integer;
            intUserCode: Integer]: boolean
QueueValue
Syntax      property QueueValue[iNodeCode: Integer;
            out wsString: WideString]: boolean
WordIndexCount
Syntax      property WordIndexCount[intNodeCode:
            Integer]: Integer
WordIndex
Description Read-Write Property
            Retrieves any index element for a
            ADM.Datagate element.
Syntax      property WordIndex[intNodeCode: Integer;
            intIndex: Integer; out strSource: OleVariant;
            out strCode: OleVariant; out intWeight:
            Integer]: boolean
ClearWordIndex
Description Read-Write Property
            Clear up any word index for a specific
            ADM.Datagate element.
Syntax      property ClearWordIndex[intNodeCode:
            Integer]: boolean
Add WordIndex
Description Read-Write Property
            Add a new word index for a specific
            ADM.Datagate element.
Syntax      property Add WordIndex[intNodeCode:
            Integer; strSource: OleVariant; strCode:
            OleVariant; intWeight: Integer]: boolean
SetProfile
Description Function
            Set a connection between the Profile object
            and th current IDatagate new instance.
Syntax      Procedure SetProfile(Profile: IProfile);
Example     DELPHI:
            Var
            pProfile, pObj: OLEVariant;
            begin
            //pProfile is created here or has already been
            created
            [..]
            try
            pObj := CreateOLEObject('ADM.INDEX);
            pObj.SetProfile(pProfile);
            [..]
            pObj := Unassigned;
            except
            raise;
            end;
            end;
            ACTIVE SERVER PAGES:
            <%@ Language=VBScript %>
            <%
            Option Explicit
            On Error Resume Next
            Dim pObj
            Set pObj =
            Server.CreateObject(“ADM.INDEX”)
            pObj.SetProfile(pProfile);
            [..]
            Set pObj = nothing
            %>
LastError
Description Read-Only Property
            This property returns the error log as text.
            Calling this function resets the error log
            content. All errors met until the LastError
            function call are part of the error log.
Syntax      Property LastError: Variant

ADM.DICTIONARY

The ADM.Dictionary COM object provides an access to the standard Adaptive Data Manager lookup dictionary. All lookup dictionary entries are classified using a type: for example ‘Gender’, ‘Age Group’, etc. Each dictionary entry is then made a code and a description. This object provides different methods for accessing lists of codes and descriptions for a specific dictionary type, or for retrieving the description for a specific dictionary type and code.

ADM.Dictionary Tools and Functions

ESTABLISHING A CONNECTION TO A DATA STORE

Instantiating the COM object is the first step to being able to data access. The second step consists of establishing a connection to the data store, i.e. the relational database storage. This is effective using the SetProfile( ) function. Examples:

For example, using Delphi:

• • Var
  • pDoc: Idictionary;
  • [..]
  • pDoc:=CreateCOMObject(‘ADM.Dictionary’);
  • pDoc.SetProfile(pProfile);
  • [..]
  • pDoc:=nil;

Active Server Pages:

• • dim pDoc
  • pDoc=CreateOLEObject(“ADM.Dictionary”)
  • pDoc.SetProfile(pProfile)
  • [..]
  • pDoc=nothing

Setting up the IDictionary interface instance with a IProfile interface establishes a connection to the ADM data store. The connection to the data store is closed when the object is de-instantiated.

RETRIEVING LOOKUP CODES AND DESCRIPTIONS BY LOOKUP TYPE

ITEMCOUNT, ITEMCODE, ITEMDESCRIPTION

ADM.Dictionary lookup data is usually retrieved using two distinct methods: either one has a dictionary lookup type and code and wants to obtain the corresponding lookup description, or one wants to retrieve all lookup code & description pairs for a specific lookup type. The ItemCount( ) function returns the number of lookup code & description pairs for a specific lookup type, and the ItemCode( ) and ItemDescription( ) functions respectively return the code and description for a lookup type and collection index. This type of use is typical when populating combo boxes, list boxes, etc.

For example, using Delphi:

Var
i: longint;
[..]
for i := 0 to pDoc.ItemCount[‘DATA_TYPE’] 1 do
showMessage(pDoc.ItemCode[‘DATA_TYPE’, i] + ‘ ‘ +
pDoc.ItemDescription[‘DATA_TYPE’, i]);

Using Active Server Pages:

Dim i
[..]
for i = 0 to pDoc.ItemCount(‘DATA_TYPE’) 1
response.write pDoc.ItemCode[‘DATA_TYPE’, i] & “ “ &
pDoc.ItemDescription[‘DATA_TYPE’, i] & “<br>”
next

In both examples, the i subscript is the index for getting all code/description pairs.

ITEMDESCRIPTIONBYCODE

The second retrieval method consists of looking for a description corresponding to a lookup type and code pair. The ItemDescriptionByCode( ) function will return such a value.

MISCELLANEOUS

LASTERROR

The LastError( ) function reports the last errors that intervened during the calling application execution. Calling LastError( ) function will reset the error list.

Function Reference

SetProfile
Description Function
            Set a connection between the Profile
            object and the current IDictionary new
            instance.
Syntax      Procedure SetProfile(Profile: IProfile);
Example     DELPHI:
            Var
            pProfile, pObj: OLEVariant;
            begin
            // pProfile is created here or has already
            been created
            [..]
            try
            pObj :=
            CreateOLEObject(‘ADM.DICTIONARY)
            ;
            pObj.SetProfile(pProfile);
            [..]
            pObj := Unassigned;
            except
            raise;
            end;
            end;
            ACTIVE SERVER PAGES:
            <%@ Language=VBScript %>
            <%
            Option Explicit
            On Error Resume Next
            Dim pObj
            Set pObj =
            Server.CreateObject(“ADM.DICTIONAR
            Y”)
            pObj.SetProfile(pProfile);
            [..]
            Set pObj = nothing
            %>
ItemCount
Description Read-Only Property
            This property returns the count of lookup code &
            description pairs for a specific lookup type.
Syntax      property ItemCount[strType: String]: Variant
ItemCode
Description Read-Only Property
            This property returns the intIndexth lookup code
            for a specific lookup type. Note that intIndex
            ranges from 0.. count − 1. Count is issued from
            ItemCount( ).
Syntax      property ItemCode[strType: String; intIndex:
            Integer]: Variant
ItemDescription
Description Read-Only Property
            This property returns the intIndexth lookup
            description for a specific lookup type. Note that
            intIndex ranges from 0.. count − 1. Count is issued
            from ItemCount( ).
Syntax      Property ItemDescription[strType: String;
            intIndex: Integer]: Variant
ItemDescriptio
nByCode
Description Read-Only Property
            This property returns the lookup description for a
            specific lookup type and code.
Syntax      property ItemDescriptionByCode[strType: String;
            strCode: String]: Variant
LastError
Description Read-Only Property
            This property returns the error log as text. Calling
            this function resets the error log content. All errors
            met until the LastError function call are part of the
            error log.
Syntax      Function LastError: String;

Claims

1. A method for managing a back-end information storage infrastructure and a flexible development environment for data storage using a computer system, comprising: managing system resources including a relational database; authenticating and selectively providing access to users through a directory describing predetermined user rights; modeling processes to create a meta data model that encloses definitions of meta data elements and relationships among the elements using trees and graphs; running the processes and generating instance data; storing the instance data following the meta data model while providing management of multi-user access and concurrency; and transforming the instance data into physical views.

2. A method in accordance with claim 1, further comprising optimizing the meta data model by optimally structuring the trees and graphs in the relational database.

3. A method in accordance with claim 1, wherein the relationships among the data elements are characteristic of the relationships between living organisms.

4. A method in accordance with claim 1, further comprising expanding the meta data model, as meta data expands, without effect on current applications.

5. A method in accordance with claim 2, wherein a plurality of concurrent meta data model trees with corresponding separate instance data trees are used in the same database.

6. A method in accordance with claim 1, further comprising auditing and tracking database changes with user and date information.

7. A method in accordance with claim 1, further comprising providing for import and export of both meta data model and instance data using extensible mark-up language.

8. A method in accordance with claim 1, wherein the database is an Oracle database and the transforming of instance data into a view comprises: defining of the view by the user; processing the view using traditional relational Oracle tables; and allowing access of the view with any SQL enabled tool.