Re: [gnome-db] GdaSqlStatementType must be introspectable

[Daniel Espinosa <esodan gmail com>]

Backported to 4.2 branch see at:

http://git.gnome.org/browse/libgda/commit/?h=LIBGDA_4.2&id=3ed8d350ee92e7ae0db09a12201cea4fb0cb6238


Just waiting for comments about to remove and rename functions
recently added in 5.0.


2011/10/13 Daniel Espinosa <esodan gmail com>:
> I've added missing headers to add bindable API to GIR and Vala.
>
> sql-parser/gda-statement-struct-parts.h
> sql-parser/gda-statement-struct-compound.h
>
> are part of GIR introspectable and I added a new function:
>
> void
> gda_sql_builder_compound_add_sub_select_from_builder (GdaSqlBuilder *builder,
> GdaSqlBuilder *subselect);
>
> This function is suppossed to replace gda_sql_builder_compound_add_sub_select,
> because the last requires to create a GdaSqlStatement by hand or by using other
> GdaSqlBuilder and get that struct.
>
> And this function:
>
> GdaSqlBuilderId
> gda_sql_builder_import_expression_from_builder (GdaSqlBuilder *builder,
> GdaSqlBuilder *query, GdaSqlBuilderId *expr_id)
>
> With it you can create an expression created in a Builder and re-use it in
> other.
>
>
> May be the functions:
> gda_sql_builder_compound_add_sub_select
> gda_sql_builder_import_expression_from_builder
>
> must be deprecated on 4.2 and deleted on 5.0. In this case
> gda_sql_builder_compound_add_sub_select_from_builder must be renamed to use
> deleted functions on 5.0.
>
> Se commit:
>
> http://git.gnome.org/browse/libgda/commit/?id=34e459b8274bba796f34e36742a64b3aa50f362c
>
> I will backported this to LIBGDA_4.2 branch.
>
> I'm waiting for additional comments and if that symbols must be deprecated
> and/deleted on 4.2 and 5.0 respectively.
>
> See at Bug #661504.
>
> 2011/10/12 Murray Cumming <murrayc murrayc com>:
>> On Wed, 2011-10-12 at 09:47 -0500, Daniel Espinosa wrote:
>>> I'm including  libgda/sql-parser/gda-sql-statement-struct-parts.h to
>>> create GIR file and adding (skip) annotations to it. This will not
>>> modify GDA headers structure.
>>
>> That does sound better. Maybe you can then gradually remove the skip
>> annotations from other parts, and file bugs for the scanner about parts
>> that it can't handle. Thanks.
>>
>> --
>> murrayc murrayc com
>> www.murrayc.com
>> www.openismus.com
>>
>>
>