You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
* refactor cedta rules into helper
* implement for bug
* test, NEWS
* About 100 packages define this now, no sense keeping track
* trailing ws
* rearrange: .datatable.aware to the top
* push isNamespace check into helper
* Fix: look up another level, use %iscall%
* nocov very-difficult-to-test lines
* more nocov
* apply data.table coding style
exchange <- with = for assignment
---------
Co-authored-by: Benjamin Schwendinger <[email protected]>
isTRUE(ns$.datatable.aware) ||# As of Sep 2018: RCAS, caretEnsemble, dtplyr, rstanarm, rbokeh, CEMiTool, rqdatatable, RImmPort, BPRMeth, rlist
74
-
tryCatch("data.table" %chin% get(".Depends",paste("package",nsname,sep=":"),inherits=FALSE),error=function(e)FALSE) # both ns$.Depends and get(.Depends,ns) are not sufficient
91
+
ans= .cedta_impl_(ns, n+1L)
75
92
if (!ans&& getOption("datatable.verbose")) {
76
93
# nocov start
77
-
catf("cedta decided '%s' wasn't data.table aware. Here is call stack with [[1L]] applied:\n", nsname)
94
+
catf("cedta decided '%s' wasn't data.table aware. Here is call stack with [[1L]] applied:\n", getNamespaceName(ns))
78
95
print(sapply(sys.calls(), `[[`, 1L))
79
96
# nocov end
80
97
# so we can trace the namespace name that may need to be added (very unusually)
0 commit comments