Re: [mdb-dev] table.c

[Brian B. <ca...@um...>]

The main reason is that I wanted to play around with exporting to Sybase
:-)

Most of the code at this point is just working out the structure of the
mdb files.  I don't have my heart set on any of it...in fact I've been
mulling over it, trying to decide on better ways to do things.  I'm not
real comfortable with the GArrays for instance, but on the other hand
reading the catalog once to get a count and then allocating it, and
rereading seems wasteful. Also, the KKD properties (design view stuff) its
highly variable, so a hash table seemed appropriate.

But, some of it strikes me as not The Right Way(tm) so I've been a bit
hesitant to cast much in stone yet.  Perhaps, I'm worrying over nothing!

As far as exporting schema, my current plan is to have a utility such as
'export-schema' and take switches like:

$ ./export-schema --target=sybase -f mydb.mdb -o mydb.sql

so you can specify different target engines.  Maybe we could even do it in
a plugin style using dlopen() and friends...that way different database
backends could be added apart from the mainline code.

Brian

On Wed, 8 Mar 2000, Carl Seutter wrote:

> I finally got a chance to start looking at some of the code today.
> Looking good.  However, while I was reading the comments in table.c the
> question was raised about how to make the column_types generic.
> Therefore, why not print out the Access datatype.
> 
> Then, have different modules for the target database.  Each module would
> then be responsible for the necessary tasks specific to that target.
> 
> 
> _______________________________________________
> mdbtools-dev mailing list
> mdb...@li...
> http://lists.sourceforge.net/mailman/listinfo/mdbtools-dev
>