### basekey.validation.R
###------------------------------------------------------------------------------------------
### What: script to check whether the selected base keys are unique or not.
### Time-stamp: <2018-10-24 12:21:27 assyst>
###-------------------------------------------------------------------------------------------


## Input Parameters
# data file path (csv)
csvpath <- input[[1]]
# list of base key variables
keyVariables <- fromJSON(input[[2]])
# libary path (R library path embedded in MDE)
libPath <- input[[3]]
# directory to load functions
workingDirectory <- input[[4]]

# load custom functions
setwd(workingDirectory)

source("fn.common.utilities.R")

# load packages
load.packages(c('haven', 'readr', 'plyr', 'jsonlite'), libPath)

# read matching variables from CSV
matchingDF <- read.matchingVariables(keyVariables, csvpath, default="c")

# find the duplicate rows
duplicateRows <- which(duplicated(matchingDF))

# find first and second occurance duplicate rows
if(length(duplicateRows) > 0){
  duplicateIndex <- duplicateRows[1]
  duplicateRow <- matchingDF[duplicateIndex,]
  if(class(duplicateRow)!= "data.frame"){
    duplicateRow <- data.frame(duplicateRow, stringsAsFactors=FALSE)
    colnames(duplicateRow) <- keyVariables
  }
  matchingIndex <- row.names(match_df(matchingDF, duplicateRow, on = keyVariables))
  firstIndex <- matchingIndex[1]
  return(c(firstIndex, duplicateIndex))
} else {
  return (0)
}