txtSQL

What if (i've considered developing this for awhile now...) the engine supported SQL via generation through classes.

Ideal SuperClass:
Query
Subclasses:
SelectQuery,
InsertQuery,
UpdateQuery,
SubQuery (may not be necessary),
DeleteQuery,
CreateQuery,
ExplainQuery (may be synonymous with DescribeQuery)

anyone else interested in this?
basically, all that would be needed then is to create an object of the appropriate query type needed, and bind some fields, choose a table or two, specify where conditions, and the query would be returned via GetQuery();

i dont really get what you mean... your plans are a bit obscure...

the idea behind this is to utilize object oriented programming to generate dynamic SQL queries with little to no knowledge of actual sequel.

The Query class would have a few public methods, which would be overridden by the subclasses SelectQuery, InsertQuery ... and so on.

This way, your queries are always well-formed, optimized, AND injection free.

how is this method any different from the one already implemented in 2.2 final?

maybe I don't really understand what you're saying.

from what i gather about the current sql generation in txtsql, the queries are simply imploded arrays.
while this implementation is KISS, it appears to be prone to syntax and typographical errors.

maybe i need to read the documentation more in depth when i have more time on my hands

current SQL generation in what version? there is none in versions < 2.2 final (4.0 is being worked on)...

when you say the queries are just imploded arrays... where are you pointing to?

http://www.fashion-products.com [url=http://www.fashion-products.com ]is a sell air force one,
air jordan,bape,kicks,nike,Jerseys,Shoes,Clothing,Bag,Glasses,Caps&Hats,sneakers,sunglasses,
watches,Jewelry,Electronics,NBA,NFL,NHL,MLB[/url]
shoes
clothing
Handbags Wallets
Hat
Sunglasses
Belts
Watches
Jewelry

The OQL-SQL mapping generator for HSQLDB generates illegal SQL statements which
refer to aliases that are not declared. This takes place particularly when
table joins are performed.

I've tracked the root of the problem being method addTables in
HsqlQueryExpression class:

private void addTables(StringBuffer buffer, Hashtable tables, boolean first).

This method adds table names to the SQL statement being generated. Although,
unlike its JDBC corresponding code, it never makes use of aliases in the
provided HashMap:

Enumeration enum = tables.elements();
while ( enum.hasMoreElements() )
{ if ( first ) first = false; else buffer.append( JDBCSyntax.TableSeparator ); buffer.append( _factory.quoteName( (String) enum.nextElement() ) ); }

}

I do not know the internals of SQL generation in Castor well enough to
contribute a full patch+test case bundle. Although I have locally replaced the
above method with the related code from JDBC SQL generation class and it passed
all of my project's test cases (these test cases have been designed to work
with MySQL and HSQLDB). Therefore I believe this patch should solve the bug. I
would suggest the person knowledgeable in HSQLDB generation engine takes a look
at the code before it is incorporated in the CVS, however.

Regards,