lhs field is not being found after := assignement.

[ProfFancyPants]

Is this new behavior intended? I am getting an error when assigning imx in the first example.

DT <- data.table(LETTERS, letters)
DT[, imx := .I]
f_det_blk <- function(x) {!all(is.na(x))}
DT[,   imx := ifelse(apply(.SD, 1, f_det_blk), imx, imx * -1L), .SDc = !'imx'] #error
DT[, ('imx') := ifelse(apply(.SD, 1, f_det_blk), imx, imx * -1L), .SDc = !'imx'] #no error

Is this the new intended behavior?

System info

library(data.table)
data.table 1.10.5 IN DEVELOPMENT built 2017-09-07 14:48:23 UTC; appveyor

data.table::test.data.table()
Running test id 168 Don't know how to automatically pick scale for object of type ITime. Defaulting to continuous.
Don't know how to automatically pick scale for object of type ITime. Defaulting to continuous.
Running test id 1372.2 Tests 1372.3+ not run. If required call library(GenomicRanges) first.
Running test id 1383.3 Tests 1383.4 not run. If required call library(caret) first.
Running test id 1751 Test 1752 not run. If required call library(nanotime) first.
Running test id 1823.3
endian==little, sizeof(long double)==16, sizeof(pointer)==8
All 6010 tests in inst/tests/tests.Rraw completed ok in 00:02:05 on Thu Sep 07 13:04:13 2017

sessionInfo()
R Under development (unstable) (2017-09-06 r73210)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C
[5] LC_TIME=English_United States.1252

attached base packages:
[1] parallel stats graphics grDevices utils datasets methods base

other attached packages:
[1] rmarkdown_1.6 curl_2.8.1 gdata_2.18.0 xts_0.10-0 zoo_1.8-0 nlme_3.1-131
[7] fastmatch_1.1-0 hexbin_1.27.1 testthat_1.0.2 plyr_1.8.4 ggplot2_2.2.1 chron_2.3-50
[13] knitr_1.17 bit64_0.9-7 bit_1.1-12 data.table_1.10.5 reshape2_1.4.2 reshape_0.8.7

loaded via a namespace (and not attached):
[1] Rcpp_0.12.12 highr_0.6 compiler_3.5.0 tools_3.5.0 digest_0.6.12 evaluate_0.10.1
[7] tibble_1.3.4 gtable_0.2.0 lattice_0.20-35 rlang_0.1.2 yaml_2.1.14 stringr_1.2.0.9000
[13] gtools_3.5.0 rprojroot_1.2 grid_3.5.0 R6_2.2.2 magrittr_1.5 scales_0.5.0
[19] backports_1.1.0 htmltools_0.3.6 colorspace_1.3-2 labeling_0.3 stringi_1.1.6 lazyeval_0.2.0
[25] munsell_0.4.3 crayon_1.3.2

rstudioapi::versionInfo()
$mode [1] "desktop"
$version [1] ‘1.1.359’

[MichaelChirico]

Perhaps introduced by #2329 ?

Can you try installing a version of data.table from about a week ago (or even the CRAN version) and confirming the bug is new?

Pinging @renkun-ken in case he sees what's going wrong.

[franknarf1]

@MichaelChirico Fwiw, it does work with CRAN (with this simpler example):

# 1.10.4
data.table(x=1, y=1)[, x := {.SD; x + 1}, .SDcols = "y"][]
#    x y
# 1: 2 1

# current dev (built 2017-09-06 22:02:57 UTC; travis)
data.table(x=1, y=1)[, x := {.SD; x + 1}, .SDcols = "y"][]
# Error in eval(expr, envir, enclos) : object 'x' not found

[renkun-ken]

Yes, it is indeed introduced by #2329 which attempts to address #2326. It seems a bit tricky to resolve both. Need to take a closer look at this.

[MichaelChirico]

Either way, this case seems to have been missed by the test suite.