.null and .na arguments in glue_sql have no effect

[luke-davis-wp]

I have some code where I'm building up complex conditions to insert into a parameterized query with glue_sql. In a previous glue version, if those conditions were NULL (in R) I would insert nothing into that part of the query using .null = DBI::SQL("") in glue_sql. However, in 1.8.0 it appears that the .null argument no longer has any impact on the output of glue_sql, and the .na argument doesn't seem to do anything either.

Is this the intended behavior of these arguments?

library(glue)
library(DBI)
library(RSQLite)

con <- dbConnect(RSQLite::SQLite(), "test.sqlite")
# In previous glue version these would be equivalent, now first returns empty query
glue_sql("select {cond}", .con = con, .null = DBI::SQL(""), cond = NULL)
#> <SQL>
glue_sql("select {cond}", .con = con, cond = DBI::SQL(""))
#> <SQL> select
# And neither .null nor .na seem to have any effect on output
glue_sql("select {cond}", .con = con, .null = "foo", .na = "foo", cond = NA)
#> <SQL> select NULL
glue_sql("select {cond}", .con = con, .null = "foo", .na = "foo", cond = NULL)
#> <SQL>

^{Created on 2024-11-26 with reprex v2.1.1}

Session info

sessionInfo()
#> R version 4.4.2 (2024-10-31)
#> Platform: x86_64-pc-linux-gnu
#> Running under: Ubuntu 20.04.6 LTS
#> 
#> Matrix products: default
#> BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 
#> LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/liblapack.so.3;  LAPACK version 3.9.0
#> 
#> locale:
#>  [1] LC_CTYPE=C.UTF-8       LC_NUMERIC=C           LC_TIME=C.UTF-8       
#>  [4] LC_COLLATE=C.UTF-8     LC_MONETARY=C.UTF-8    LC_MESSAGES=C.UTF-8   
#>  [7] LC_PAPER=C.UTF-8       LC_NAME=C              LC_ADDRESS=C          
#> [10] LC_TELEPHONE=C         LC_MEASUREMENT=C.UTF-8 LC_IDENTIFICATION=C   
#> 
#> time zone: Etc/UTC
#> tzcode source: system (glibc)
#> 
#> attached base packages:
#> [1] stats     graphics  grDevices utils     datasets  methods   base     
#> 
#> other attached packages:
#> [1] RSQLite_2.3.8 DBI_1.2.3     glue_1.8.0   
#> 
#> loaded via a namespace (and not attached):
#>  [1] digest_0.6.37     fastmap_1.2.0     bit_4.5.0         xfun_0.49        
#>  [5] cachem_1.1.0      blob_1.2.4        knitr_1.49        memoise_2.0.1    
#>  [9] htmltools_0.5.8.1 bit64_4.5.2       rmarkdown_2.29    lifecycle_1.0.4  
#> [13] cli_3.6.3         vctrs_0.6.5       reprex_2.1.1      withr_3.0.2      
#> [17] compiler_4.4.2    rstudioapi_0.17.1 tools_4.4.2       evaluate_1.0.1   
#> [21] yaml_2.3.10       rlang_1.1.4       fs_1.6.5

[jennybc]

All I can say off the top of my head is that there has been a sequence of recent-ish issues and PRs around what glue_sql() does for NULL or length-0 inputs.

#272
which lead to #292
which lead to #318
which lead to 51f889c