Re: Transitive Closure with XDb1

From: Hugo Kornelis <hugo_at_pe_NO_rFact.in_SPAM_fo>
Date: Fri, 14 May 2004 23:51:05 +0200
Message-ID: <uhfaa09sh78bu9okijnvjvhe99vjuc802n_at_4ax.com>


On 14 May 2004 12:41:20 -0700, Neo wrote:

>> Paul wrote: "...the transitive closure of a binary relation R on a set X is
>> the smallest transitive relation on X that contains R."
>
>After reading several web pages on transitive closure, I am realizing
>that your definition and output are correct and what I was
>demonstrating is related but different and I am unaware of a name for
>it. I was demonstrating XDb1's ability to retrieve things in a
>hierarchy [given a thing in the hierarchy and a relator [ie
>isPartOf, isParentOf, isBossOf, etc)], where things in the hierarchy
>can be of different type.
>
>When applied to a simple case, man's supposed family tree, RDM's TClose(Tbl_Child)
>returns:
> god adam
> god eve
> god john
> adam john
> eve john
>
>where as XDb1's T_Relative(god, isChildOf) returns:
> adam
> eve
> john

Hi Neo,

And the latter is exactly what would returned if you qualified RDM's TClose(Tbl_Child) with WHERE whole = 'god'.

I suppose XDb1 also has some way to show the equialent of the RDM's TClose with no where clause (as shown above) and a way to get all parents from the same data, like the RDM's TClose(Tbl_Child) WHERE part = 'john', that would return
  god john
  adam john
  eve john

Groetjes, Hugo

-- 

Sorry, vandaag geen grappige sig lines meer.
Received on Fri May 14 2004 - 23:51:05 CEST

Original text of this message