Hi Alfred,

now I see what you mean saying "a very long time". I tried a dataset
with a half a million features. It takes hours and hours to complete.
I tried something different, what seems to work much better, but I am
not sure whether it can be used in your case.

Anyway, here is a solution for a database (I tested it on MySQL).

I took a table mjrrds (major roads in USA) with 450 000 features, made
a copy (updrds), deleted some features from the original table, added
a few, and changed some in the copy (field alt1_name). ID is a unique
identifier. Added four source datasets reading a small simple table,
which has nothing to do with the tables we are interested in. In this
case I used an interesting side effect of SELECT parameter - it can
extract features specified in the query, not in the feature type.

These selects look as following:

DELETED:
select * from mjrrds where id not in (select id from updrds)

ADDED:
select * from updrds where id not in (select id from mjrrds)

UPDATED:
select * from updrds u inner join mjrrds m on u.id = m.id where
u.alt1_name <> m.alt1_name

UNCHANGED:
select * from updrds u inner join mjrrds m on u.id = m.id where
u.alt1_name = m.alt1_name

It takes only a few minutes to run these queries, and after that
Workbench has the same four groups of features as UpdateDetector would
give us after many hours.

Here is an example - 

http://f5.grp.yahoofs.com/v1/oGIxRNidc0gVxJOoqmEYIvKvNAwZY_5t2ywDHeQ3sX4n2bJiTIA2SIAvkGa3R-EXbuWLvN8hoJ-JxHUWOPMlA99BV8w/updateDetectorDATABASE.fmw

Regards,

Dmitri

---------------------------------------------------------------
Dmitri Bagh                               [EMAIL PROTECTED]
Scenario Creation and Testing Analyst   (604) 501-9985 ext.276
Safe Software Inc.
---------------------------------------------------------------


--- In [email protected], "alfred_sawatzky" <[EMAIL PROTECTED]> wrote:
>
> Thanks for this transformer!
> 
> I was wondering if anyone might have a suggestion for this scenario.
> 
> I want to see the updates/adds/deletes that I have made in one ESRI 
> Geodatabase version compared to another version in the same 
> Geodatabase.  I used workbench to create two source datasets (one 
> for each version) and then connected them to the original and 
> revised input ports of the update detector, respectively.  The 
> transformer does correctly identify all the changes but the only 
> problem is that this can take a very long time because FME basically 
> reads each feature twice (once for each version) and then uses 
> transformers to compare the appropriate attributes.
> 
> Is there any way that FME can make use of any of the ArcObject 
> functionality with respect to difference cursors?  Are there other 
> ways to reduce the number of features that are passed into the 
> update detector?
> 
> Regards,
> 
> Alfred
> 
> 







Get the maximum benefit from your FME, FME Objects, or SpatialDirect via our 
Professional Services team.  Visit www.safe.com/services for details. 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/fme/

<*> To unsubscribe from this group, send an email to:
    [EMAIL PROTECTED]

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/
 


Reply via email to