users-prolog
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

DES 2.0 Released


From: Fernando Sáenz Pérez
Subject: DES 2.0 Released
Date: Tue, 31 Aug 2010 13:21:37 +0200
User-agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax)

[Apologies for multiple receptions]

Hello.
This is a mail just for informing that the new release

 Datalog Educational System version 2.0
 http://des.sourceforge.net

has been launched on August, 31st, 2010, and ported to

 GNU Prolog 1.3.1

The main novelty of the current release is the support of ODBC
connections to relational database management systems (RDBMSs)
to provide data sources for relations. This means that a relation
defined in a RDBMS as a view or table is allowed as any other
relation defined with a predicate in the deductive database. Then,
computing a query can involve computations both in the deductive
inference engine and in the  external RDBMS SQL engine. In
addition, duplicates are now supported and can be enabled upon
request via a command. Also, naive tracers for Datalog and SQL has
been included, which are intended to catch buggy predicates and
views, respectively. A new port to SICStus 4.x is provided, which
replaces the old port to SICStus 3.x.

A complete list of addtions, changes and fixed bugs is attached to
the end of this message.
Please, see http://des.sourceforge.net for details.

 Best regards.


==============================================================
Fernando Saenz Perez
Profesor Titular de Universidad / Associate Professor
Home Page: http://www.fdi.ucm.es/profesor/fernan
Tel: + 34 913947642. Fax: + 34 913947547
Despacho / Office: 435 (4ª planta / 4th floor)
Dept. Ingenieria del Software e Inteligencia Artificial /
Department of Software Engineering and Artificial Intelligence
Universidad Complutense de Madrid
Facultad de Informatica
C/Profesor Jose Garcia Santesmases, s/n
E - 28040 Madrid. Spain
==============================================================



Version 2.0 of DES (released on August, 31st, 2010)

* Enhancements:
o Connection to RDBs via ODBC connections (DSN providers as MySQL, MS Access, Oracle, ...) RDB tables and views can be queried both from SQL and Datalog
 o Duplicates are allowed in answers, both for Datalog and SQL
 o Datalog and SQL tracers
 o New commands:
- /open_db Name [Options] Open and set the current ODBC connection to Name, where Options=[user(Username)] [password(Password)]. This connection must be already defined at the OS layer
   - /close_db Close the current ODBC connection
   - /current_db Display the current ODBC connection name and DSN provider
   - /duplicates Display whether duplicates are enabled
   - /duplicates Switch Enable or disable duplicates (on or off, resp.)
- /trace_sql View [Order] Trace a SQL view in the given order (postorder or the default preorder) - /trace_datalog Goal [Order] Trace a Datalog basic goal in the given order (postorder or the default preorder)
   - /output Switch Enable or disable display output (on or off, resp.)
   - /save_ddb Filename Save the current Datalog database to a file
- /restore_ddb Filename Restore the Datalog database in the given file (same as consult) o Results from SELECT SQL statements (those sent to an ODBC connection) can contain duplicates
 o Added UPDATE SQL statement
 o Added varchar2 Oracle SQL datatype
 o Remarks can now start with '--', as in Oracle SQL
 o Both EXCEPT and MINUS are allowed to express SQL set difference
o SQL user identifiers can be enclosed between quotation marks (either double quotes "", or square brackets [], or backquotes ``)
 o Closing the opened log file, if any, on quitting
o Added timing information to SQL query processing, including listings which may include view processing from RDBs
 o Some dead code removal

* Changes:
 o New port to SICStus 4.x, which replaces the old port to SICStus 3.x
 o The command debug is renamed as debug_datalog
 o Executables have been built with SWI-Prolog, instead of SICStus Prolog

* Fixed bugs:
o Asserting rules with a number as atom/string changed the type to number, as in /assert t('1'), which asserted t(1) instead of t('1') o Disjunctions in aggregate goals might lead to missing answers, as in group_by((p(X,Y),(Y=a;Y=b)),[X],C=count) o Some infix builtins were accepted without delimiting blanks, as Xis1, posed as a goal, and interpreted as X is 1

* Caveats and limitations:
 o See Section 10 of the user manual

* Known bugs:
 o The projection list of a natural outer join is not correct in all cases
o Disjunctions in having conditions in the group_by clause may display errors which are not o Operator precedence in SQL conditions are not correctly handled. Use parentheses to ensure correct operator applications




reply via email to

[Prev in Thread] Current Thread [Next in Thread]