修改 R (RStudio) 中的 LexisNexisTools 以重命名文件

Question

I'm trying to rename.txt files in a directory downloaded from Nexis Advance UK. Being unfamiliar with coding, I set about trying to modify LexisNexisTools' code in RStudio.

What I've done is change term.v <- content_v[grep("^Terms: |^Begriffe: ", content_v)] to term.v <- content_v[grep("The Guardian(London)", fixed = T, content_v)] , for instance, and changed the rename function so that it only pastes term.v . However, I'm trying to retain the original OR function so that the code would cycle through a number of strings such "Express Online" or "The Independent (United Kingdom)" and then paste the string found into the file rename function.

Here is what I've tried so far:

1 - Use regular expressions (from what I could gather online on regular expressions with spaces in strings) with fixed = F , such as "^The/sGuardian(London)$|^Express/sOnline$"

2- I've tried using a vector to "house" the different patterns and then paste the vector in the grep command

toMatch.v <- c("Express Online", "The Times (London)", "The Independent (United Kingdom)" 

term.v<- content_v[grep(paste(toMatch, collapse="|"),  content_v)]

The only time the code (as modified) works is when fixed = T and the string is typed as is found in the.txt files.

What am I doing wrong? Thank you so much and I apologize if the terminology isn't accurate.

Extra details:

Originally, the code relies on a set of keywords to find the search term and insert it into the file's name:

    content_v <- readLines(files[i], encoding = encoding, n = 50)
    term.v <- content_v[grep("^Terms: |^Begriffe: ", content_v)]
    # erase everything in the line exept the actual range
    term.v <- gsub("^Terms: |^Begriffe: ", "", term.v)
    # split term into elemets seprated by and or OR
    term.v <- unlist(strsplit(term.v, split = " AND | and | OR ", fixed = FALSE))

I have changed it so that grep begins with the string that I want to append to the filename, as explained above. I have also disabled the gsub line and changed the split argument to "/n" as the string in my text files is separated with line breaks. Here is an example of a sample.txt file.

Answer 1

Assuming that you have a file file1.txt with something like the following content in your working directory:

foo
foo bar Express Online
bar

Then, the following code should rename the file into Express Online.txt .

file1 <- "file1.txt"

text1 <- readLines(file(file1))

# if (any(grepl("The Guardian", text1))) {
#     file.rename(file1, "The Guardian.txt")
# } else if (any(grepl("Express Online", text1))) {
#     file.rename(file1, "Express Online.txt")
# }

newname <- head(
    n = 1,
    na.omit(
        stringr::str_extract(
            text1,
            "(Express Online)|(The Times \\(London\\))|(The Independent \\(United Kingdom\\))")))

file.rename(file1, paste0(newname, ".txt"))

Answer 2

Unfortunatly your file format is quite different from how the files looked when I wrote LexisNexisTools . And so are your requirements. So I would write new code to do the job here. First, let's try for one file:

f <- "/home/johannes/Documents/x.txt"
lines <- readLines(f)

toMatch.v <- c("Express Online", "The Times (London)", "The Independent (United Kingdom)")

# I'm using another function from the package to convenietly look up a several patterns at once
np <- unlist(LexisNexisTools::lnt_lookup(lines, toMatch.v, verbose = FALSE))[1]
new_name <- paste0(dirname(f), "/", np, ".txt")
new_name
#> [1] "/home/johannes/Documents/Express Online.txt"

file.rename(f, new_name)

Once this works for you as intended, you can implement it for a number of files. As in my original function I would suggest you write your new names in a data.frame first so you can check if the new names make sense and if you have duplicates in the new names ( R would write both files into the new name without warning and destroy one file in that case):

files <- list.files("/home/johannes/Documents/", pattern = ".txt$", 
                    ignore.case = TRUE, full.names = TRUE)

make_new_name <- function(old_name) {
  lines <- readLines(old_name)

  np <- unlist(LexisNexisTools::lnt_lookup(lines, toMatch.v, verbose = FALSE))[1]
  paste0(dirname(old_name), "/", np, ".txt")
}  

df <- tibble::tibble(
  old = files,
  new = sapply(files, make_new_name)
)
df               
#> # A tibble: 2 x 2
#>   old                              new                                        
#>   <chr>                            <chr>                                      
#> 1 /home/johannes/Documents//x.txt  /home/johannes/Documents/Express Online.txt
#> 2 /home/johannes/Documents//x2.txt /home/johannes/Documents/Express Online.txt

If the new names make sense to you and if there are no duplicates ( table(duplicated(df$new)) ), you can pull the trigger and rename the files:

file.rename(df$old, df$new)