script to fix an out-of-order .srt caption file

srt_sort.R

setwd('~/Documents/personal/r_stuff/bcc/srt_sort/')

# processing function. input is lines of the file as char vector
fix_srt <- function(srt_file) {

  # caption chunks are delimited by a blank line ''. 
  # so... add a blank line to the start
  srt_file <- c('', srt_file)
  
  # ...now the cumulative sum of blanks so far is a vector we can split on
  srt_chunks <- split(srt_file, cumsum(srt_file == ''))
  
  # in each chunk the 3nd element = time stamp, so we can sort on that
  correct_order <- order(sapply(srt_chunks, function(x) x[3]))
  srt_chunks <- srt_chunks[correct_order]
  
  # reassign index numbers to reflect new order (2nd element of each chunk)
  fixed_chunks <- lapply(seq_along(srt_chunks), function(i) {
    temp_chunk <- srt_chunks[[i]]
    temp_chunk[2] <- i
    temp_chunk
  })
  
  # now we have the new file
  fixed_srt <- unlist(fixed_chunks)
  
  # return, less first blank line
  return(fixed_srt[-1])
}

# get list of .srt files with full paths
file_list <- list.files('original_srt', pattern = 'srt$',
                        recursive = F, full.names = T)
outdir <- 'fixed_srt/'

# process files
for(f in file_list) {
  cat('here goes', basename(f))
  
  srt_file <- suppressWarnings(readLines(f))
  fixed <- fix_srt(srt_file)
  
  out_name <- paste0(outdir, basename(f))
  cat(' =>', out_name, '\n')
  writeLines(fixed, out_name)
}