R/IMPORT-nmr-guess.R

In jmstrat/NMR.Utils: Importing and processing Bruker NMR data

NMR_TEXT_FILE_PATTERN <- "*.(txt|dat)"

read.nmr.guess <- function(path, ...) {
  if (dir.exists(path)) {
    return(read.nmr.guess.directory(path, ...))
  } else if (file.exists(path)) {
    # We're working with a text file
    return(read.nmr.guess.file(path, ...))
  } else {
    stop(path, " was not found.")
  }
}

read.nmr.guess.directory <- function(path, ...) {
  jms.classes::log.debug("Treating %s as a directory", path)
  # File is a directory, determine whether it is a bruker folder or a normal folder

  # First look for numbered subfolders
  numbered_contents <- dir(path, pattern="^[[:digit:]]+$", full.names=T)
  # Then check if any of them are directories
  if (any(dir.exists(numbered_contents))) {
    # We have a brucker experiment folder
    return(read.nmr.directory.bruker.experiment(path, ...))
  }

  # Next check for a pdata folder
  if (dir.exists(paste0(path, "/pdata"))) {
    # We have a bruker process directory
    return(read.nmr.directory.bruker.process(path, ...))
  }
  jms.classes::log.debug("Treating %s as a non-bruker directory", path)
  # We now assume we're working with a folder of text files
  # Two options:
  # 1) In-situ files saved as individual scans (e.g. if carrier frequency is changed during scan)
  # 2) 1D spectrum with multiple carrier frequencies

  # In order to determine which is which we have to inspect the ppm scale:
  # We assume case 1 if the ppm scale is identical for each scan
  # Otherwise we choose case 2

  # Firstly we find all of the files
  files <- jms.classes::list.files.sorted(path, pattern=NMR_TEXT_FILE_PATTERN, full.names=TRUE)
  jms.classes::log.debug("%s contains %s files", path, length(files))
  # We check that we have some files
  if (length(files) == 0) {
    stop("No files were found.")
  }

  # Then we inspect the first file
  type_skip <- determine_1d_file_type(files[[1]])

  if (type_skip[[1]] == 1) {
    return(read.nmr.directory.text.1d.1col(files))
  } else if (type_skip[[1]] == 2) {
    return(read.nmr.directory.text.1d.2col(files, type_skip[[2]]), type_skip[[3]])
  } else {
    stop("File type not supported")
  }
}

read.nmr.guess.file <- function(path, imaginary_file=NA, ...) {
  jms.classes::log.debug("Treating %s as a file", path)
  dim <- determine_dimensions_from_text_file(path)
  if (dim == 1) {
    return(read.nmr.text.1d(path))
  } else if (dim == 2) {
    return(read.nmr.text.2d.1col(path, imaginary_file=imaginary_file))
  } else {
    stop("Input file not supported.")
  }
}

#' Determine NMR file type
#'
#' This function attempts to determine the type of a bruker text file
#' @param file File to read
#' @return An integer representing the file type
#' @examples
#' determine_1d_file_type(file)
#' @keywords internal
determine_1d_file_type <- function(file) {
  tryCatch({
    # Extract info from preamble
    # Open file for reading
    con <- file(file, open="r")
    # Read first line
    lines <- readLines(con, n=2)
    # Close file
    close(con)
    # Extract lines
    l1 <- lines[[1]]

    if (grepl("^# File created", l1)) {
      return(c(1, NA, NA))
    } else if (grepl("^ppm scale", l1)) {
      return(c(2, 3, ""))
    } else if (grepl("^[-+eE.0-9]* [-+eE.0-9]*$", l1)) {
      # 2 columns of data, no header
      return(c(2, 0, ""))
    } else {
      n <- tryCatch({
        ncol(read.table(text=lines[[2]]))
      }, error=function(e) {
        0
      })
      if (n == 4) {
        # ascii-spec
        return(c(2, 1, ","))
      }

      stop("Input file not supported.")
    }
  }, error=function(e) {
    jms.classes::log.error(e)
    stop("Input file not supported.")
  })
}

#' Read NMR Data Dimensions
#'
#' This function attempts to determine the dimension of a bruker text file by examining the preamble.
#' @param file Path to txt file saved using topspin's File->Save->Save visible region as text file (with or without imaginary)
#' @return A number indicating the dimensions
#' @examples
#' determine_dimensions_from_text_file("/path/to/file.txt")
#' @keywords internal
determine_dimensions_from_text_file <- function(file) {
  tryCatch({
    # Extract info from preamble
    # Open file for reading
    con <- file(file, open="r")
    # Read first 5 lines
    lines <- readLines(con, n=5)
    # Close file
    close(con)
    # Extract lines
    l4 <- lines[[4]]
    l5 <- lines[[5]]

    if (grepl("^# LEFT", l4)) {
      return(1)
    } else if (grepl("^# F2LEFT", l5)) {
      return(2)
    } else {
      n <- tryCatch({
        ncol(read.table(text=lines[[2]]))
      }, error=function(e) {
        0
      })
      if (n == 4) {
        # ascii-spec
        return(1)
      }
      stop("Input file not supported.")
    }
  }, error=function(e) {
    jms.classes::log.error(e)
    stop("Input file not supported.")
  })
}