R读入文件的最新版本

Question

I am currently reading in a file like the example below, is there a way to specify to read the latest version of the file ie if I have files saved as "Abroad v1.csv", "Abroad v2.csv" I would want it to take the latest which would be 2 in this case.

year <- "2015"
species <- "HOM"

root <- "Y:/Pelagic Work/FIN Data"
file <- "Abroad.csv"
ABR <- file.path(root, year, species, file)

Answer 1

If at all reasonable, it would be best to determine the "latest" version of a file by the data in the file.info output ( this post , also suggested by zx8754, is a good answer for that).

If you must do it by filename, be very careful and aware of how your operating system is going to sort characters. Take, for example, this example.

files <- paste0("somepath/directory/filename v", 1:10, ".csv")

basenames <- basename(files)

sort(basenames)

 [1] "filename v1.csv"  "filename v10.csv" "filename v2.csv"  "filename v3.csv"  "filename v4.csv" 
 [6] "filename v5.csv"  "filename v6.csv"  "filename v7.csv"  "filename v8.csv"  "filename v9.csv"

As you can see, filename v10.csv' appears in the second position and would not be picked up by simple methods such at tail(basenames, 1)`. Instead, you need to strip out all of the characters except for those that specify the order of the versions, convert to the correct format, then sort. Here's an example of how to do this with integer versions as you've suggested you have.

Files <- data.frame(path = dirname(files),
                    file = basename(files),
                    stringsAsFactors = FALSE)
Files$v_number <- gsub("(^filename v|[.]csv$)", "", basenames)
Files$v_number <- as.numeric(Files$v_number)

Files <- Files[order(Files$v_number), , drop = FALSE]
tail(Files, 1)

This, however, is clunky and error prone. If at all possible, I'd recommend transitioning to a database, or version control, or both.