bug in emr_extract with list iterator

[nettam]

See following example:

library(tidyverse)

emr_db.connect("/net/mraid14/export/tgdata/db/tgdb/emr/mock")

emr_filter.create('lab_1m', 'lab.101', time.shift=c(-1,0)*month())
emr_filter.create('f_75', 'patients.dob', time.shift=c(-76, -74)*year())
emr_vtrack.create('age', 'patients.dob', time.shift=c(-120,0)*year(), func='dt2.earliest')

x <- emr_extract('age', iterator=list(75*year(), 'patients.dob'), filter='f_75 & !lab_1m', stime=emr_date2time(1,1,2012), etime=emr_date2time(1,1,2014))
y <- emr_extract('age', iterator=x, filter='f_75 & !lab_1m')

x and y should be exactly the same!!!

x %>% anti_join(y)

Joining, by = c("id", "time", "ref", "age")
id time ref age
1 171496 1269701 -1 657000
2 2223455 1270385 -1 657000

#x contains entries that were filtered in y!!!

[aviezerl]

Note that this happens only when filtering on age 75, i.e., when running:

x <- emr_extract('age', iterator=list(75*year(), 'patients.dob'), filter='!lab_1m', stime=emr_date2time(1,1,2012), etime=emr_date2time(1,1,2014))
y <- emr_extract('age', iterator=x, filter='f_75 & !lab_1m')

x and y are the same.