PE> So Paul, how about another database trick?



Abracadabra!



Ok, you need three tables:



File Table

----------



File ID

File name

File size

File desciprition

etc



Area Table

----------

Area ID

Area Name

anything else you can think of like security, retention etc.



File-Area Table

---------------

File ID

Area ID





All ID fields are just internally generated and used number.





As an example:



File Table

----------



File ID    File Name      File Description

-------    -----------    -------------------

1          PD_BBS.ZIP     Public domain BBS

2          SW_BBS.ZIP     Shareware BBS





Area Table

----------



Area ID    Area Description

-------    ----------------

1          Public Domain

2          Shareware

3          BBS stuff

4          Junk DB sent me





File-Area table

---------------



File ID    Area ID

-------    -------

1          1

1          3

1          4

2          2

2          3





This means that PD_BBS.ZIP (file ID 1) is in areas Public domain (area ID
1), BBS stuff (area ID 3) and Junk DB sent me (area ID 4), whereas
SW_BBS.ZIP (file ID 2) is in Shareware (area ID 2) and BBS stuff (area ID
3).



To find what areas a file is in you need to read the file table and get the
file ID. Then read the File-Area table and get all the matching Area IDs
and finally read the area table for the description.



To find which files are in a particular area you start from the area table
and work the other way.



 PE> And while you're at it, how about recommending a database that I can

 PE> use?  I want to be able to write C with embedded SQL, and distribute

 PE> the final executable.  I want OS/2 and DOS support, and grudgingly

 PE> I'm even prepared to buy one, as it is basically everyone's duty to

 PE> support the SQL standard.



I'm not really up on all the SQL products that are available. You'd have to
look through some of the programming magazines to find something.



 PE> Is there some other database-related thing like Btrieve or something that

 PE> I can use instead, even if it means doing some manual translation of CSQL

 PE> into C/Btrieve?  Preferably something Public Domain, or at least easy to

 PE> implement so that I can do it myself.  Crikey, haven't we progressed

 PE> anywhere in databases in the last 40 years?  BFN.



There are plently of things like that around. It really depends on what you
need and how much you are willing to spend. Again, I don't know enough to
be able to recommend one particular product. The only one I own is the MIX
Database Toolchest, which is dirt cheap, pretty good for a single user (not
SQL though) and the C source code is available so you could probably port
it to OS/2.





Paul



--- GoldED/2 2.42.G1114