#!/usr/bin/env Rscript
######################################################################
#                  Copyright (c) 2018 Northrop Grumman.
#                          All rights reserved.
######################################################################
#
# Version 1 - January 2018
# Author: Cristel Thomas
#
#

library(flowCore)
library(MetaCyto)

compare_lists <- function(m1, m2) {
  list_check <- T
  if (is.na(all(m1 == m2))) {
    mm1 <- is.na(m1)
    mm2 <- is.na(m2)
    if (all(mm1 == mm2)) {
      if (!all(m1 == m2, na.rm = TRUE)) {
        list_check <- F
      }
    } else {
      list_check <- F
    }
  } else if (!all(m1 == m2)) {
    list_check <- F
  }
  return(list_check)
}


run_batch_processing <- function(sampling_size = 5000, flag_default = T,
                               to_exclude, outdir = "", outfile = "",
                               labels, assays, factors, fcspaths, fcsnames) {
  # Create meta_data object
  fp <- unlist(fcspaths)
  file_counts <- lengths(fcspaths)
  group_names <- rep(labels, times = file_counts)
  group_bs <- rep(factors, times = file_counts)
  group_types <- rep(assays, times = file_counts)

  meta_data <- data.frame(fcs_files = fp, study_id = group_names)

  # excluded_parameters
  default_param <- c("FSC-A", "FSC-H", "FSC-W", "FSC", "SSC-A", "SSC-H",
                     "SSC-W", "SSC", "Time", "Cell_length", "cell_length",
                     "CELL_LENGTH")
  excluded_parameters <- if (flag_default) default_param else to_exclude
  # Run preprocessing.batch
  preprocessing.batch(inputMeta = meta_data,
                      assay = group_types,
                      b = group_bs,
                      fileSampleSize = sampling_size,
                      outpath = outdir,
                      excludeTransformParameters = excluded_parameters)

  # deal with outputs
  # output[2]: a csv file summarizing the pre-processing result.
  ## -> open file to pull info and print out filenames rather than path.
  tmp_csv <- file.path(outdir, "processed_sample_summary.csv")
  tmp <- read.csv(tmp_csv)
  tmp$old_index <- seq(1, length(tmp$fcs_names))

  fn <- unlist(fcsnames)
  df <- data.frame(fcs_files = fp, filenames = fn)

  # merge two data frames by ID
  total <- merge(tmp, df, by = "fcs_files")
  total2 <- total[order(total$old_index), ]
  to_drop <- c("fcs_names", "fcs_files", "old_index")
  newdf <- total2[, !(names(total2) %in% to_drop)]
  write.table(newdf, file = outfile, quote = F, row.names = F, col.names = T, sep = "\t")

  file.remove(tmp_csv)
}

check_fcs <- function(sampling = 5000, flag_default = TRUE, to_exclude,
                     outdir = "", outfile = "", labels, assays, factors,
                     fcspaths, fcsnames) {

  if (length(labels) > length(unique(labels))) {
    # we have repeated group names, all group names need to be different
    print("ERROR: repeated labels among groups, make sure that labels are all different for groups.")
    print("The following labels are repeated")
    table(labels)[table(labels) > 1]
    quit(save = "no", status = 13, runLast = FALSE)
  }

  marker_pb <- FALSE
  for (i in seq_len(length(fcspaths))) {
    for (n in seq_len(length(fcspaths[[i]]))) {
      marker_check <- FALSE
      marker_channel <- FALSE
      tryCatch({
        fcs <- read.FCS(fcspaths[[i]][[n]], transformation = FALSE)
      }, error = function(ex) {
        print(paste("File is not a valid FCS file:", fnames[[i]][[n]], ex))
        quit(save = "no", status = 10, runLast = FALSE)
      })

      if (n == 1) {
        m1 <- as.vector(pData(parameters(fcs))$desc)
        c1 <- colnames(fcs)
      } else {
        m2 <- as.vector(pData(parameters(fcs))$desc)
        c2 <- colnames(fcs)
        marker_check <- compare_lists(m1, m2)
        marker_channel <- compare_lists(c1, c2)
      }
      if (n > 1 && marker_check == F) {
        marker_pb <- TRUE
        print(paste("Marker discrepancy detected in markers -- group", labels[[i]]))
      } else if (n > 1 && marker_channel == F) {
        marker_pb <- TRUE
        print(paste("Marker discrepancy detected in channels -- group", labels[[i]]))
      }
    }
  }

  if (marker_pb) {
    quit(save = "no", status = 12, runLast = FALSE)
  } else {
    run_batch_processing(sampling, flag_default, to_exclude, outdir, outfile,
                       labels, assays, factors, fcspaths, fcsnames)
  }
}

################################################################################
################################################################################
args <- commandArgs(trailingOnly = TRUE)

# Arg 1: sub_sampling number
# Arg 2: output dir for processed FCS files for check_fcs and run_batch_processing
# Arg 3: Main output file (text file)
# Arg 4: excluded params
# Arg 5: Group 1 Name
# Arg 6: Group 1 format
# Arg 7: Group 1 Scaling factor
# Cycle through files in group 1
# Arg  : file path in Galaxy
# Arg  : desired real file name
# Cycle through at at least one additional group
# Arg : 'new_panel' - used as some sort of delimiter
# Arg : Group n+1 Name
# Arg : Group n+1 format
# Arg : Group n+1 Scaling factor
## Cycle through files in that group
## Arg : file path in Galaxy
## Arg : desired real file path

sub_sampling <- NULL
if (as.numeric(args[1]) > 0) {
  sub_sampling <- as.numeric(args[1])
}

# parameters to exclude => args[4]
to_exclude <- vector()
flag_default <- FALSE
i <- 1
if (args[4] == "None" || args[4] == "") {
  flag_default <- TRUE
} else {
  excluded <- unlist(strsplit(args[4], ","))
  for (channel in excluded) {
    stripped_chan <- gsub(" ", "", channel, fixed = TRUE)
    if (!is.na(stripped_chan)) {
      to_exclude[[i]] <- stripped_chan
    }
    i <- i + 1
  }
}

# handle group cycle in arguments to produce iterable panels
tot_args <- length(args)
tmpargs <- paste(args[5:tot_args], collapse = "=%=")
tmppanels <- strsplit(tmpargs, "=%=new_panel=%=")
nb_panel <- length(tmppanels[[1]])

labels <- vector(mode = "character", length = nb_panel)
assay_types <- vector(mode = "character", length = nb_panel)
scaling_factors <- vector(mode = "numeric", length = nb_panel)
filepaths <- list()
filenames <- list()

# iterate over panels (groups of fcs files)
j <- 1
for (pnl in tmppanels[[1]]) {
  tmppanel <- strsplit(pnl, "=%=")
  # number of FCS files
  nb_files <- (length(tmppanel[[1]]) - 3) / 2
  tmplist <- character(nb_files)
  tmpnames <- character(nb_files)
  if (tmppanel[[1]][[1]] == "None" || tmppanel[[1]][[1]] == "") {
    print(paste("ERROR: Empty group name/label for group ", j))
    quit(save = "no", status = 11, runLast = FALSE)
  } else {
    labels[[j]] <- tmppanel[[1]][[1]]
  }
  # assay type
  assay_types[[j]] <- tmppanel[[1]][[2]]

  scaling_factors[[j]] <- 0
  if (as.numeric(tmppanel[[1]][[3]]) > 0) {
    scaling_factors[[j]] <- 1 / as.numeric(tmppanel[[1]][[3]])
  }

  k <- 1
  for (m in 4:length(tmppanel[[1]])) {
    if (!m %% 2) {
      tmplist[[k]] <- tmppanel[[1]][[m]]
      tmpnames[[k]] <- tmppanel[[1]][[m + 1]]
      k <- k + 1
    }
  }
  filepaths[[tmppanel[[1]][1]]] <- tmplist
  filenames[[tmppanel[[1]][1]]] <- tmpnames
  j <- j + 1
}

check_fcs(sub_sampling, flag_default, to_exclude, args[2], args[3], labels,
         assay_types, scaling_factors, filepaths, filenames)

# check_fcs <- function(sampling = 5000, flag_default = TRUE, to_exclude,
#                       outdir = "", outfile = "", labels, assays, factors,
#                       fcspaths, fcsnames)