ggrothendieck / gsubfn Goto Github PK

options(
  warnPartialMatchArgs = TRUE,
  warnPartialMatchAttr = TRUE,
  warnPartialMatchDollar = TRUE
)
list[a, b] = list(1, 2)

gives

Warning message:
In seq.default(along = args) :
  partial argument match of 'along' to 'along.with'
Calls: [<- -> [<-.result -> seq

In the following a list is output where a plain vector is expected. If engine = "R" is omitted then it runs correctly.

xx <- c("1 year 1 month", "11 \nmonths", "3 years 11 months", "1 year 1 month", 
"8 months", "2 years")
strapply(xx, "(\\d+) +\\W*m", as.numeric, empty = 0, simplify = TRUE, engine = "R")

Installation works fine, but loading package produces a illegal operation.

*** caught illegal operation *** address 0x7fffc435ee97, cause 'illegal opcode'

Installation works fine, but loading package produces a illegal operation.

*** caught illegal operation *** address 0x7fffc435ee97, cause 'illegal opcode'

Hi there!

I have recently switched laptops and am on the version 4.0.4 of R and although install.packages("gsubfn") installs the package when I try to call the package from my library or to call gsubfn via gsubfn::gsubfn() R gets stuck running a process for an insane amount of time (I gave up at 30 mins). Do you have any idea what the problem could be? Is there some dependency that I am unaware of?

Many thanks

Hello,

I am using R 3.5.0 and every time I try to load the package it causes R to freeze. In order to do anything with R again I have to force quit out of RStudio to even start using it again.

I was finally able to work around this by typing:
options(gsubfn.engine = "R")
into the command prompt first before trying to load gsubfn.

best
MB

a <- 10
b <- 5

fn$c("Hello $a + $b should equal `a + b`, but `don't evalute me`, $b")

I would want to be able to escape the second pair of backticks, or dollar signs e.g.:

fn$c("Hello $a + $b should equal `a + b`, but ``don't evalute me``, $$b")
## "Hello 10 + 5 should equal 15, but `don't evalute me`"

I think this can be achieved by changing pattern in:

if (missing(pattern)) pattern <- "[$]([[:alpha:]][[:alnum:].]*)|`([^`]+)`"

To something like (with perl=T):

"(?<!\\$)[$]([[:alpha:]][[:alnum:].]*)|`(?<!`)`([^`]+)`(?!`)`"

This type of escaping is probably simpler than anything involving backslashes given those have special meaning within R and with regexes as well. The only unfortunate thing is this is different to how the formula escaping works, where ~~ is the function.

Additionally, one complication is that we need to make sure that things like:

``don't evaluate me``

are not only not evaluated, but also converted to

`don't evaluate me`