Postgres window function with where clause

Question

I have 4 tables in Postgres with the following organization

reads:
id -- primary key
value -- integer
read_datetime -- datetime
patient_id -- integer
reader_id -- integer

readers:
id -- primary key

patient_readers:
id -- primary key
patient_id -- integer
reader_id -- integer
issuance_datetime -- datetime
unassignment_datetime -- datetime
issued_value -- integer

patients:
id -- integer

I want to create a query where I can get the most recent issuance information for a particular read. So far I have the following query

SELECT
value,
read_datetime,
issued_value,
issuance_datetime,
unassignment_datetime
FROM "reads"
INNER JOIN "readers" ON "readers"."id" = "reads"."reader_id"
INNER JOIN "patient_readers" ON "patient_readers"."reader_id" = "readers"."id" AND patient_readers.patient_id = reads.patient_id 
INNER JOIN patientsON patients.id = reads.patient_id
ORDER BY reads.patient_id , read_datetime, issuance_datetime

This produces output where a group of rows are the same for all information other than the issuance information (issuance_datetime, unassignment_datetime, and issued_value) as I would expect. I would like to aggregate the issuance information with respect to readers.id and patients.id, and then I would only like the one set of issuance information where

read_datetime > issuance_datetime AND (
    (unassignment_datetime IS NOT NULL AND read_datetime < unassignment_datetime) OR
    (NEXT(issuance_datetime) IS NOT NULL AND read_datetime < NEXT(issuance_datetime) OR
    (NEXT(issuance_datetime) IS NULL)

Here I'm making up the function NEXT() but basically I want to look ahead one row with respect to the window. I also want to do a WHERE clause in the window (or so it seems). I have just started reading about window functions, so a lot of this is new to me (and I'm not positive that window functions are what I want) so I hope this is solvable within Postgres.

Thanks

Answer 1

Check out the lead() function. That one can look at the "next" row in the window.

http://www.postgresql.org/docs/current/static/functions-window.html