`add_variables()` doesn't create group with `group`

[TanguyBarthelemy]

The group argument is used to check whether a group with the same name already exists.

If so, the variable will be added to that group:

library("rjd3toolkit")
library("rjd3workspace")

tmp_path <- tempdir()

vars <- list(reg1 = list(x1 = AirPassengers))
my_context <- modelling_context(variables = vars)
my_ws <- jws_new(modelling_context = my_context)

add_variables(
    jws = my_ws, 
    group = "reg1", 
    y = mdeaths, 
    name = "DeathMale"
)

save_workspace(my_ws, file.path(tmp_path, "tmp4.xml"), replace = TRUE)

If not, then a new group is created but with the name of the argument name (not group) :

my_ws <- jws_new()

add_variables(
    jws = my_ws, 
    group = "reg1", 
    y = mdeaths, 
    name = "DeathMale"
)
add_variables(
    jws = my_ws, 
    group = "reg1", 
    y = fdeaths, 
    name = "DeathFemale"
)
add_variables(
    jws = my_ws, 
    group = "DeathMale", 
    y = ldeaths, 
    name = "TotDeath"
)

save_workspace(my_ws, file.path(tmp_path, "tmp5.xml"), replace = TRUE)

readLines(file.path(tmp_path, "tmp5.xml")) |> cat(sep ="\n")
#> <?xml version="1.0" encoding="UTF-8" standalone="yes"?>
#> <ns2:demetraGenericWorkspace xmlns="ec/tss.core" xmlns:ns2="ec/tss.demetra" name="tmp5">
#>     <ns2:items>
#>         <ns2:item readOnly="false">
#>             <ns2:family>Utilities@Variables</ns2:family>
#>             <ns2:name>DeathMale</ns2:name>
#>             <ns2:file>Vars-1</ns2:file>
#>         </ns2:item>
#>         <ns2:item readOnly="false">
#>             <ns2:family>Utilities@Variables</ns2:family>
#>             <ns2:name>DeathFemale</ns2:name>
#>             <ns2:file>Vars-2</ns2:file>
#>         </ns2:item>
#>     </ns2:items>
#> </ns2:demetraGenericWorkspace>

read_variables(normalizePath(file.path(tmp_path, "tmp5/Variables/Vars-1.xml"))) |> str()
#> List of 2
#>  $ DeathMale: Time-Series [1:72] from 1974 to 1980: 2134 1863 1877 1877 1492 ...
#>   ..- attr(*, "name")= chr ""
#>  $ TotDeath : Time-Series [1:72] from 1974 to 1980: 3035 2552 2704 2554 2014 ...
#>   ..- attr(*, "name")= chr ""
read_variables(normalizePath(file.path(tmp_path, "tmp5/Variables/Vars-2.xml"))) |> str()
#> List of 1
#>  $ DeathFemale: Time-Series [1:72] from 1974 to 1980: 901 689 827 677 522 406 441 393 387 582 ...
#>   ..- attr(*, "name")= chr ""

^{Created on 2025-06-03 with reprex v2.1.1}

[AQLT]

I think this is a Java issue.

[annasmyk]

Should be solvable in R according to @palatej

[AQLT]

I might be wrong but my understanding of this line of the Java code is that the Java code uses the name of the series instead of the family to create the new group
https://github.com/jdemetra/jdplus-main/blob/59e793dea767c798054512e75f479c2909a26ff7/jdplus-main-base/jdplus-sa-base-parent/jdplus-sa-base-workspace/src/main/java/jdplus/sa/base/workspace/Ws.java#L189
In that case simply correcting the this line should also solved the problem.

[TanguyBarthelemy]

One way of doing this would be to bypass add_variable() and go directly through the contexts.

[AQLT]

Again, I think that the simplest way would be to simply correct the Java code.