for loops nested in R - r

Question

I have a dataset dt, it stored list dataset names, I need to use them to create some new datasets with select some variables, then I use the dataset I just created, repeat the same process .....
The first row and second row were data available.
Then use data available to create a new data.
Then use data just create to create a new data
The final output was list of datasets
I appreciated any helps or suggestions.
dt <- data.frame(name = c("mtcars","iris", "mtcars_new","mtcars_new_1"),
data_source = c("mtcars","iris", "mtcars","mtcars_new"),
variable = c("","","mpg,cyl,am,hp","mpg,cyl"), stringsAsFactors = FALSE)
> dt
name data_source variable
1 mtcars mtcars
2 iris iris
3 mtcars_new mtcars mpg,cyl,am,hp
4 mtcars_new_1 mtcars_new mpg,cyl
dt_list <- list(mtcars, iris)
names(dt_list ) <- c("mtcars","iris")
# The final list of datasets
final_dt <- list(mtcars, iris, mtcars_new, mtcars_new_1)
So far if I wrote a loop like that, I got only mtcars_new dataset, but I don't know how to return to the list and continue looping to get mtcars_new_1 and so on. I have many datasets, and I don't know how many times I should looping through nested data.
mtcars_new <- data.frame()
for(i in 1:nrow(dt)){
if(dt$data_source[[i]] %in% names(dt_list) && !dt$name[[i]] %in% names(dt_list)){
check <- eval(parse(text = dt$data_source[[i]]))
var <- c(unlist(strsplit(dt$variable[[i]],",")))
mtcars_new <- check[, colnames(check) %in% var]
}
}

Answer 1

This will produce the desired output shown. Since the fourth loop uses the data created in the third loop, you need to have a way to append the results of each loop to a growing list of available data sets. Then within each loop find which one is the right starting data set from the available list.
dt <- data.frame(name = c("mtcars","iris", "mtcars_new","mtcars_new_1"),
data_source = c("mtcars","iris", "mtcars","mtcars_new"),
variable = c("","","mpg,cyl,am,hp","mpg,cyl"), stringsAsFactors = FALSE)
input_data_sets <- list(mtcars, iris)
names(input_data_sets) <- c("mtcars","iris")
final_data_sets <- list()
for(i in 1:nrow(dt)) {
available_data_sets <- c(input_data_sets, final_data_sets) #Grows a list of all available data sets
num_to_use <- which(dt$data_source[[i]] == names(available_data_sets)) #finds the right list member to use
temp <- available_data_sets[num_to_use][[1]]
var <- c(unlist(strsplit(dt$variable[[i]],",")))
temp <- list(subset(temp, select = var)) #keep only the desired variables
names(temp) <- dt$name[i] #assign the name provided
final_data_sets <- c(final_data_sets, temp) #add to list of final data sets which will be the output. Anything listed here will become part of the available list in the next loop
}