RE: Want to use Axion as headstart for Excel JDBC driver -- where to begin?

[Waldhoff, Rodney]

Hi Bart.  

If I understand you correctly, then yes, I believe it should be sufficient to create an implementation of Table that is backed by a spreadsheet.  It may help to take a look at MemoryTable as an example, as that's pretty much as pure and simple an implementation you'll find.  Note that all you need to do is create the "row source"--something that will read the rows of spreadsheet into Axion Rows--the rest of the Axion machinery will take care of the filtering/joins etc. for you.

-----Original Message-----
From: Bart Guijt [mailto:bguijt@...]
Sent: Fri 9/9/2005 5:43 AM
To: Axion Dev Mailinglist
Subject:  Want to use Axion as headstart for Excel JDBC driver -- where to begin?
 
Dear Axion devs,

My primary question is: How can I add (or replace) a new persistent
database abstraction (in my case Jakarta POI) to the Axion JDBC
engine?

Objective
---------
In lots of businesses, spreadsheets are used to communicate
information in some kind of user-friendly manner, e.g. reports,
questionaires, software requirements and 'regular' data. Even the
European Central Bank used spreadsheets to distribute timeseries data
across Europe to national banks.
Most of these spreadsheets are structured in a way that these could be
processed by a machine rather than a human. Luckily several frameworks
out there help to do this (e.g. Jakarta POI).

I want to enhance the spreadsheet access further by means of a JDBC
driver. I would primarily build a 'simple' driver, which uses each
worksheet as a table where the first row specifies the column names
(or something like that) -- secondly I would see how to wire complex
spreadsheet data to Table abstractions.

Axion
-----
I searched several OSS JDBC databases to give me a headstart in
creating the driver, and I found out most implementations have
severely complex abstractions of Database, Table, Type etc. that they
seemed unusable.
When I checked out Axion I was surprised by its clean design and
rather clean abstractions of Database and Table so I choose to use
Axion as the JDBC engine for the driver.

Where to start?
---------------
Well, it seems that I just need to wire the Axion engine to POI, so I
need to build some Axion/Poi adapter. I prefer to have a non-obtrusive
adapter which doesn't interfere/change any code in Axion.

I noticed there are three types of Tables:
table.disk=org.axiondb.engine.DiskTableFactory
table.memory=org.axiondb.engine.MemoryTableFactory
table.external=org.axiondb.engine.tables.ExternalTableFactory

This might seem like adding another TableFactory might do the trick.
The question is, is that feasable in the current codebase? Or do I
need to provide another implementation for the DiskTableFactory and
all its helper classes?

Thanks in advance,

--
Bart Guijt
GSM: [phone]

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...



---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Manoj Kr. Sheoran]

Hi Folks,
Idea is very excellent. You guy need this sort of
Implementation. I did something like this , If you
interested to look that then Please download
DaffodilDB and use Browser to explore the stuffs.

Regards,
Manoj Kr. Sheoran
[mksheoran@...]

PS: "I am replying becuase I've subscribed this group
and pay interest on this not for any other means"

--- "Waldhoff, Rodney" <rwaldhof@...> wrote:

> Hi Bart.  
>
> If I understand you correctly, then yes, I believe
> it should be sufficient to create an implementation
> of Table that is backed by a spreadsheet.  It may
> help to take a look at MemoryTable as an example, as
> that's pretty much as pure and simple an
> implementation you'll find.  Note that all you need
> to do is create the "row source"--something that
> will read the rows of spreadsheet into Axion
> Rows--the rest of the Axion machinery will take care
> of the filtering/joins etc. for you.
>
> -----Original Message-----
> From: Bart Guijt [mailto:bguijt@...]
> Sent: Fri 9/9/2005 5:43 AM
> To: Axion Dev Mailinglist
> Subject:  Want to use Axion as headstart for Excel
> JDBC driver -- where to begin?
>  
> Dear Axion devs,
>
> My primary question is: How can I add (or replace) a
> new persistent
> database abstraction (in my case Jakarta POI) to the
> Axion JDBC
> engine?
>
> Objective
> ---------
> In lots of businesses, spreadsheets are used to
> communicate
> information in some kind of user-friendly manner,
> e.g. reports,
> questionaires, software requirements and 'regular'
> data. Even the
> European Central Bank used spreadsheets to
> distribute timeseries data
> across Europe to national banks.
> Most of these spreadsheets are structured in a way
> that these could be
> processed by a machine rather than a human. Luckily
> several frameworks
> out there help to do this (e.g. Jakarta POI).
>
> I want to enhance the spreadsheet access further by
> means of a JDBC
> driver. I would primarily build a 'simple' driver,
> which uses each
> worksheet as a table where the first row specifies
> the column names
> (or something like that) -- secondly I would see how
> to wire complex
> spreadsheet data to Table abstractions.
>
> Axion
> -----
> I searched several OSS JDBC databases to give me a
> headstart in
> creating the driver, and I found out most
> implementations have
> severely complex abstractions of Database, Table,
> Type etc. that they
> seemed unusable.
> When I checked out Axion I was surprised by its
> clean design and
> rather clean abstractions of Database and Table so I
> choose to use
> Axion as the JDBC engine for the driver.
>
> Where to start?
> ---------------
> Well, it seems that I just need to wire the Axion
> engine to POI, so I
> need to build some Axion/Poi adapter. I prefer to
> have a non-obtrusive
> adapter which doesn't interfere/change any code in
> Axion.
>
> I noticed there are three types of Tables:
> table.disk=org.axiondb.engine.DiskTableFactory
> table.memory=org.axiondb.engine.MemoryTableFactory
>

table.external=org.axiondb.engine.tables.ExternalTableFactory

>
> This might seem like adding another TableFactory
> might do the trick.
> The question is, is that feasable in the current
> codebase? Or do I
> need to provide another implementation for the
> DiskTableFactory and
> all its helper classes?
>
> Thanks in advance,
>
> --
> Bart Guijt
> GSM: [phone]
>
>

---------------------------------------------------------------------
> To unsubscribe, e-mail:
> dev-unsubscribe@...
> For additional commands, e-mail:
> dev-help@...
>
>
>
>
---------------------------------------------------------------------
> To unsubscribe, e-mail:
> dev-unsubscribe@...
> For additional commands, e-mail:
> dev-help@...
>
>


Regards,
Manoj Kr. Sheoran
[phone]

PS : "Nothing is impossible even impossible says I M possible"


               
__________________________________
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...