International Business Machines Corporation

Michael James Carey

Gerald George Kiernan

Serge Philippe Rielau

Bennet Vance

Mir Hamid Pirahesh

Eugene Jon Shekita

George Lapis

Jayavel Shanmugasundaram

Philip L Boyer

Donald Alan Kossmann

Narayana Iyer

A method for publishing relational data as XML by translating XML queries into queries against an relational database. Conversion of the relational database into an XML database is not required. Each relational table is mapped to a virtual XML document, and XML queries are issued over these virtual documents. An XML query is parsed and transformed into a language-neutral intermediate representation, which is a sequence of operations describing how the output document is derived from the underlying relational tables. The intermediate representation is then translated into an SQL query over the underlying relational tables and into instructions for a tagger. The SQL query is executed, and the SQL query results are then fed into the tagger, which follows tagger instructions to generate the marked up output.

Using an XML query language to publish relational data as XML

A computer system identifies a predicate in a computer language containing constant expressions as vacuous. The system identifies distinct variables contained in the predicate and represents the predicate by a set of bounding rectangles. The bounding rectangles are represented in a space having a number of dimensions equal to the number of variables. There are finite limits on dimensions of a bounding rectangle which represent the relationship between the variables in the predicate and the constant expressions in the predicate. The predicate is identified as vacuously FALSE where the set of bounding rectangles is empty.

Identification of vacuous predicates in computer programs

A method, apparatus, and article of manufacture for a computer-implemented technique for query optimization with deferred updates and autonomous sources. An object-oriented query is executed to retrieve data from a database. The database is stored on a data storage device connected to a computer. The object-oriented query is transformed into subqueries, wherein at least one subquery is directed against a database, and wherein one subquery is directed against an object cache. Each subquery that is directed against a database is executed to retrieve data from the database into the object cache. The subquery that is directed against the object cache is executed to retrieve data for the query, wherein the data incorporates updates to the object cache and updates to the database.

Query optimization with deferred updates and autonomous sources

The system, method, and program of this invention enables an object language application to issue a query over a view and to receive back, as query results, handles to application type objects which can be further manipulated by the application. A view is defined herein as a collection of a view type, and a view type is defined as a class or type. Object language class definitions are created based upon the view type of the view referenced by the query. Upon receipt of the query referencing a view type, a query engine generates a query plan that builds mock application type objects in memory based upon the view types and having a form consistent with the class definition for a type of object returned as a result.

System, method, and program for object building in queries over object views

A method, apparatus, and article of manufacture for a computer-implemented technique for query optimization using a multi-layered object cache. An object-oriented query is executed to retrieve data from a database. The database is stored on a data storage device connected to a computer and queries are run against lower cache layers which have better performance characteristics than the external or upper layers. The multi-layered cache has an application objects (AOs) cache and a data access objects (DAOs) cache, and the application objects include methods deriving data from DAOs. The method includes a step of rewriting the query into a query directed against the DAOs cache, applying a pushdown transformation algorithm to the rewritten query directed against the DAOs cache, and executing the transformed query against the DAOs cache. The method also allows optimizing queries with path expressions by transforming path expression into joins. Another performance optimization includes using a limited data set only including the inserted and updated data, which allows the method to evaluate queries against a subset of a collection of cached objects instead of an entire collection of cached objects.

Query optimization using a multi-layered object cache

A method for typed view hierarchies for an object-relational database management (OPDBMS). The views in the hierarchy are merged into a unified hierarchy view through case expressions for mapping and wrapping of overloaded columns. During the merge, the underlying logical tables are eliminated which results in a simplified query graph. The transformation is a non-loss operation which condenses the type information within the case expressions. This allows update, delete and insert operations to be resolved on the query graph on a local level. Using query rewrite rules, the case expressions may be simplified further.

Efficient implementation of typed view hierarchies for ORDBMS

Query optimization with deferred update and autonomous sources

An object language application (e.g., C&plus;&plus;, JAVA, etc.,) issues a query over a view and receives back, as query results, handles to application type objects which can be further manipulated by the application. A view is defined herein as a collection of a view type, and a view type is defined as a class or type. A tool is used by a programmer writing the application to create object language class definitions that are based upon the view type of the view referenced by the query. Upon receipt of the query referencing a view type, a query engine generates a query plan that builds mock (i.e., proxy) application type objects in memory based upon the view types. The application objects have a form that is consistent with the class definition for a type of object returned as a result. The application can run methods on the application type objects or point to other application type objects from the handles, to the application objects, that are returned to the application; and these manipulations will be understood by the query engine. In a preferred embodiment, query rewrite optimizations are applied to the queries over views requiring object building in order to optimize the evaluation of the query and the building of view objects as query results. Query rewrite optimizations can also be applied so that parts of the query are pushed down to the DBMS to minimize the number of objects that need to be built.

System, method, and program, for object building in queries over object views

The system, method, and program of this invention enables an object language application (e.g., C++, JAVA, etc.,) to issue a query over a view and to receive back, as query results, handles to application type objects which can be further manipulated by the application. A view is defined herein as a collection of a view type, and a view type is defined as a class or type. Upon receipt of the query referencing a view type, a query engine generates a query plan that builds mock (i.e., proxy) application type objects in memory based upon the view types. The application can run methods on the application type objects or point to other application type objects from the handles, to the application objects, that are returned to the application; and these manipulations will be understood by the query engine. In a preferred embodiment, query rewrite optimizations are applied to the queries over views requiring object building in order to optimize the evaluation of the query and the building of view objects as query results. For example, when a query over a view is analyzed and it is determined that the query is not requesting a handle, and is not referencing a method, but only asks for values, no objects are built. Also, if a query traverses a reference type attribute, but the query can be transformed into a join or outer join operation between relational tables, then no object building is required. In these above described situations, the rewritten query can be pushed down to the database management system of the data source for resolution. If the query does request a handle or references a method, then some object building is required. However, query rewrite techniques can still be applied so that parts of the query are pushed down to the DBMS to minimize the number of objects that are built.

System, method, and program for applying query rewrite technology to

The system, method, and program of this invention enables an object language application to issue a query over a view and to receive back, as query results, handles to application type objects which can be further manipulated by the application. A view is defined herein as a collection of a view type, and a view type is defined as a class or type. Upon receipt of the query referencing a view type, a query engine generates a query plan that builds mock (i.e., proxy) application type objects in memory based upon the view types. The application objects have a form that is consistent with the class definition for a type of object returned as a result. The application can run methods on the application type objects or point to other application type objects from the handles, to the application objects, that are returned to the application; and these manipulations will be understood by the query engine. In a preferred embodiment, query rewrite optimizations are applied to the queries over views requiring object building in order to optimize the evaluation of the query and the building of view objects as query results.

System, method, and program for object building in queries over object

The system, method, and program of this invention, referred to herein as nest elimination, performs query rewrite transformations, within a database management system (DBMS), for a certain class of object queries over views that involve derived nested sets. The nest elimination algorithm uses query rewriting to avoid computing these nested sets in certain common cases. For each quantifier in a query (if the quantifier is defined over a nested set that is a part of a view or if the quantifier is implemented by a subquery that projects a NEST aggregate) the following functions are performed: a) resetting an element to which a quantifier is bound in the database query to an argument of a nest aggregate function in a nest subquery of a view; b) adding a null testing predicate to the rewritten database query; c) adding quantifiers defined within the nest subquery to the rewritten database query; d) adding predicates from the nest subquery of the view to the rewritten database query; and e) rewriting each expression involving a quantifier over an instance into a path expression. The resulting rewritten query is then further processed by the DBMS such as through a query optimizer, etc. By avoiding nested set computations for a certain class of queries, the rewritten query can result in improved processing performance.

Database management system, method, and program for providing query

A STOP AFTER clause for a relational database management system (RDBMS) structured query language (SQL) causes the RDBMS, in response to a query, to return a query result having at most a predetermined cardinality specified in the query. A query with a STOP AFTER clause is processed by the RDBMS by provision of one or more STOP operators that are inserted into a query execution plan generated by the RDBMS to execute the query. In a conservative policy, STOP operators are inserted in the query execution plan such that no tuples that might be required in the query result are discarded. In contrast, an aggressive policy inserts a STOP operator in the query execution plan wherever it is able to provide a beneficial cardinality reduction. A RESTART operator is inserted into aggressive policy query execution plans to ensure that at least the predetermined number of tuples are returned in the query result, and a final STOP operator is added at or near the root of the plan to ensure that at most the specified number of tuples are produced.

Method and system for limiting the cardinality of an SQL query result

A method, apparatus, and article of manufacture for handling NULL values in SQL queries over object oriented data. A two-phase method is used to enable a query evaluator in a two-valued logic environment to properly handle occurrences of NULL values for predicates that involve subqueries, i.e., basic subquery predicates and/or quantified subquery predicates. For basic subquery predicates, negation reduction is performed by applying logical equivalence rules and inverting basic comparators (e.g., transforming &lt; to &gt;=) to eliminate NOTs. Then, transformations are employed for the resulting positive predicates to include NULL value testing, i.e., NULL protection. For quantified subquery predicates, in addition to performing negation reduction, quantified subqueries are converted into existential subqueries. In most cases, this yields a predicate that can be handled using NULL protection transformations for positive predicates. The exception (i.e., a 'NOT EXISTS' resulting from the conversion step of a universally quantified subquery) is handled using NULL protection transformations for negative predicates. The evaluation of the NULL tested positive and negative predicates ensures that if a predicate has a NULL value, it is not included in the query result.

Handling null values in SQL queries over object-oriented data

A method of, and system for, interactively accessing information in response to a user command having a predefined operator and specifying one of a plurality of collections of information. The collections of information are stored in an object-oriented database in a hierarchical arrangement of data members. The hierarchical arrangement can include one level of data members and in which one of the data members is composed of a next level of data members. Each data member is stored according to one of a system-specified and a user-specified storage definition. A set of user-specified storage definitions is defined from the plurality of storage definitions. Upon detecting the user command, the specified collection of information is analyzed to determine which data members of the one level are stored according to one of the storage definitions of the set. Each data member not in the set is added to a projection list; each data member in the set is expanded into a next level of data members composing the data member in the set. Expanding can include analyzing each data member of the next level to determine whether it is in the set and if so again expanding up until a specified level of expanding. The projection list is processed to return to the user the information corresponding to the data members in the projection list. The invention may be implemented in a Parser layer of a known layered architecture for database systems.

Growing community of inventors

San Jose, CA, United States of America

Michael James Carey