Delete all rows in a file that fit between certain headers? - r

Question

I would like to delete all of the rows that sit between certain headers in this example text file.
fileConn <- file("sample.txt")
one <- "*Keyword"
two <- "*Node"
three <- "$ Node,X,Y,Z"
four <- "1,639982.78040607,4733827.5104821,0"
five <- "2,639757.59709573,4733830.43494066,0"
six <- "3,639738.81268144,4733834.3619618,0"
seven <- "*End"
writeLines (c(one, two, three, four, five, six, seven), fileConn)
close(fileConn)
sample <- readLines("sample.txt")
What I am looking to do is delete all of the rows/lines between "*Node" and "*End". Since I am dealing with files with different lengths of rows between these headers, the deletion method needs to be based on headers only. I have no idea how to do this since I've only deleted rows in dataframes referenced by row numbers previously. Any clues?
Expected output is:
*Keyword
*Node
*End

Answer 1

readLines returns a vector, not a data frame, so we can create the sample input more simply:
sample = c("*Keyword",
"*Node",
"$ Node,X,Y,Z",
"1,639982.78040607,4733827.5104821,0",
"2,639757.59709573,4733830.43494066,0",
"3,639738.81268144,4733834.3619618,0",
"*End")
Find the starting and ending headers, and remove the elements in between with negative indexing:
node = which(sample == "*Node")
end = which(sample == "*End")
result = sample[-seq(from = node + 1, to = end - 1)]
result
# [1] "*Keyword" "*Node" "*End"
This assumes there is a single *Node and a single *End line. It also assumes that there is at least one line to delete. You may want to create a more robust solution with some handling for those special cases, e.g.,
delete_between = function(input, start, end) {
start_index = which(sample == start)
end_index = which(sample == end)
if (length(start_index) == 0 | length(end_index) == 0) {
warning("No start or end found, returning input as-is")
return(input)
}
if (length(start_index) > 1 | length(end_index) > 1) {
stop("Multiple starts or ends found.")
}
if (start_index == end_index - 1) {
return(input)
}
return(input[-seq(from = start_index + 1, to = end_index - 1)])
}