#!/usr/bin/env Rscript

library(PECA)

# Set up R error handling to go to stderr
options(show.error.messages=F, error=function(){cat(geterrmessage(),file=stderr());q("no",1,F)})

# Avoid crashing Galaxy with an UTF8 error on German LC settings
loc <- Sys.setlocale("LC_MESSAGES", "en_US.UTF-8")

main <- function() {
    args <- commandArgs(trailingOnly = TRUE)
        
    # read first argument, which is quant file (proteins in rows, samples in columns)
    quant_file <- read.delim(args[1],
                           na.strings = c("0", "NA"),
                           sep = '\t')

    # read 2nd and 3rd argument, which are path to files with group 1 and 2 column indices 
    group1_ind_file <- as.vector(as.matrix(read.csv(args[2], header = FALSE)[1, ]))
    group2_ind_file <- as.vector(as.matrix(read.csv(args[3], header = FALSE)[1, ]))

    # test type
    ttype <- args[4]

    # read 4th argument, which is output file name
    out_file <- args[5]

    # get vectors of column names
    df_names <- colnames(quant_file)
    group1_names <- df_names[group1_ind_file]
    group2_names <- df_names[group2_ind_file]
 
    # run PECA_df
    # id column is first column (hard-coded)
    results <- PECA_df(quant_file,
                group1_names,
                group2_names,
                type=ttype,
                id = df_names[1])

    # put add protein names column
    results$prot <- row.names(results)
    results <- results[, c("prot", "slr", "t", "score", "n", "p", "p.fdr")]
 
    # write to output
    write.table(results, file=out_file,
               quote=FALSE,
               sep ='\t', 
               row.names = FALSE)
}

main()