## symbol

[Emily Greenfest-Allen]

Hi all-

In the query DynSpanAttributes.Bfmv you have the following: ##WDK_ID_SQL##

I am not familiar with this syntax. Could you explain?

Thanks,

Emily

[Cristina Aurrecoechea]

Hi Emily,

WDK susbtitutes this macro with the search cached result.

DynSpan is the new record type called "genomic segment", which is a "span" in the genome, with a start and stop coordinates.
Bfmv is what we call our attribute tables. As you know we pre-compute all attributes for all IDs of a record type, in these attribute tables.
But we cannot do that for "spans" (too many)... so the bfmv query uses the search result to bring the IDs and their attributes into the final result.

Cristina

[Emily Greenfest-Allen]

Cristina-

I appreciate the explanation, that clarifies it a bit.
But I am still a little confused. What is the "cached search result"
that the macro substitutes? I am assuming that it is related to the info
passed to the showRecord action ... is this correct? Does it just
contain the primary key and the project id for the record type? Is this
something that can be used with any record type or is it specific to the
dynaSpan?

Thanks,

Emily

[Jerric Gao]

The ##WDK_ID_SQL## is only used by genomic segments. As Cristina explained, the
genomic segment records are always generated on the fly (from its id queries),
and all the essential information about the records are contained in the id
queries. Therefore, when we need to get the attributes of the genomic segments,
the only source is from the id queries. The macro represents the sql of the ID
query of the set of the genomic segment records. The id sql can be either the
cached sql (if the id query is cache enabled), or the original sql of the id
query (if cache is disabled).

If you are curious about how we handle the genomic segment record page, where
the record is not specifically associated with any id query, here is how we did
it: we replace the ##WDK_ID_SQL## with something like "SELECT source_id FROM
dual", where the source_id is the id of the genomic segment record.

On the other hand, the ##WDK_ID_SQL## can be used by any record types, but we
haven't found any use case of using it for a normal record type.

Jerric

[Emily Greenfest-Allen]

Jerric-

Thanks! Makes sense now.

-E