R/backend.R

In roblen001/document_similarity_checker: ONDRI Proposal Similarity Checker ShinyApp

# backend.R
#
# Purpose: The back end of the application
#
# Author: Roberto Lentini (rlentini@research.baycrest.org)
#
# Date: 2021-05-20
#
# ========================================================================================

backend <- function(input, output, session){
  # creating empty state
  full_proposal_path <- shiny::reactiveVal()
  fileName <- shiny::reactiveVal()
  commonKeyWords <- shiny::reactiveVal('')
  selectedFilePath <- shiny::reactiveVal('')
  proposalTitle <- shiny::reactiveVal('')
  commonKeyWords_OtherProposals <- shiny::reactiveVal('')
  amount_of_commonWords <- shiny::reactiveVal(0)
  amount_of_commonWords_publishedPapers <- shiny::reactiveVal(0)
  proposalDF <- shiny::reactiveVal()
  similar_articles_df <- shiny::reactiveVal()
  DataframeProposalFile <- shiny::reactive({input$DataframeProposalFile})
  DownloadLocationReport <- shiny::reactive({input$DownloadLocationReport})

  # getting working directory of report.rmd file
  report_path_src <- system.file("www", "ONDRI_full-logo_web.png", package = "ProposalSimilarityChecker")
  report_path <- shiny::reactive({report_path_src})
  # Computer volumes depend on OS: Windows, Mac, or Linux.
  volumes <- c(Home = fs::path_home(), shinyFiles::getVolumes()())

  shinyFiles::shinyFileChoose(input, id = "proposalFile", roots = volumes, filetypes = c("pdf"))
  output$proposalFileOutput <- shiny::renderText("No proposal file selected.")

  shinyFiles::shinyFileChoose(input, id = "DataframeProposalFile", roots = volumes, filetypes = c("csv"))
  output$DataframeProposalFileOutput <- shiny::renderText("No proposal dataframe file selected.")

  shinyFiles::shinyDirChoose(input, id = "DownloadLocationReport", roots = volumes)
  output$DownloadLocationReportFile <- shiny::renderText("No location selected.")

  #  FOR OTHER PROPOSAL SELECTION INPUT TYPE
  # 1) Select file containing proposal dataframe
  observeEvent(
    input$DataframeProposalFile, {
      proposalDFPath <- shinyFiles::parseDirPath(volumes, input$DataframeProposalFile)

      # Reset output when selection of directory is canceled.
      output$DataframeProposalFileOutput <- shiny::renderText("No file selected.")


      # Otherwise if directory has been selected, print path of directory to screen.
      if (length(proposalDFPath) > 0){
        # Update output by printing new directory path.
        output$DataframeProposalFileOutput <- shiny::renderPrint(proposalDFPath)


      }
    }
  )

  observeEvent(
    input$DownloadLocationReport, {
      downloadPath <- shinyFiles::parseDirPath(volumes, input$DownloadLocationReport)

      # Reset output when selection of directory is canceled.

      output$DownloadLocationReportFile <- shiny::renderText("No location to download report selected.")

      # Otherwise if directory has been selected, print path of directory to screen.
      if (length(downloadPath) > 0){
        # Update output by printing new directory path.

        output$DownloadLocationReportFile <- shiny::renderPrint(downloadPath)

      }
    }
  )

  # detects when the select input is changed
  observeEvent(input$checkUsing, {
    shinyjs::hide("main_content")
  })

  # submit button
  observeEvent(input$BeginCheck, {
      proposal_df_path =  shinyFiles::parseFilePaths(volumes, input$DataframeProposalFile)
      # check that a csv file has been selected
      if (identical(proposal_df_path$datapath, character(0))){
        # warning pop up
        shinyalert::shinyalert("Oops!", "Please make sure you have selected a proposal dataframe prior to starting the check.", type = "error")
      }else{
        #  checking what the user has selected for CHECK USING
        if (input$checkUsing == 'Keywords'){
          # make sure the input keyword list is not empty
          input_word_list <- as.list(strsplit(input$keywordsList, ","))
          if (identical(input_word_list[[1]], character(0))){
            shinyalert::shinyalert("Oops!", "Please make sure you have at least one word in the keyword input.", type = "error")
          } else {
          # keep track of words to make sure there are no duplicates
          input_word_list <- as.list(strsplit(input$keywordsList, ","))
          words <- c()
          for (word in input_word_list) {
            # separating input into list of words
            word <- trimws(word, which = "both")
            word <- tolower(word)
            words <- c(words, word)
            # TODO: make this automatic instead of having a pop up
            duplication <- duplicated(words)
            if (TRUE %in% duplication){
              shinyalert::shinyalert("Oops!", "Please make sure you do not have any duplicates in the keyword input.", type = "error")
            } else{
              # show loading screen
              shinyjs::show("loading_page")
              shinyjs::hide("main_content")

              # getting selected file name
              similar_proposal <- checkWithKeywords(filePath = proposal_df_path$datapath, input = input$keywordsList)
              proposalDF(similar_proposal)

              load_data()

            }
          }
          }

        }else if (input$checkUsing == 'ProposalId'){
          corpus_raw <- readr::read_csv(proposal_df_path$datapath)
          # the inputed dataframe should always have the same ordering ID, Author, Title, Background information
          colnames(corpus_raw) <- c('id', 'author', 'proposal_title', 'background', 'hypothesis', 'variables', 'analysis', 'significance',
                                    'Keyword 1', 'Keyword 2', 'Keyword 3', 'Keyword 4', 'Keyword 5', 'status')
          # check if the id exists in the dataframe
          if (input$proposalID %in% corpus_raw$id){
            # show loading screen
            shinyjs::show("loading_page")
            shinyjs::hide("main_content")

            # getting selected file name
            similar_proposal <- getSimilarProposalsFromCSV(proposalDataFile = proposal_df_path$datapath,
                                                           idForFileBeingChecked=input$proposalID)
            commonKeyWords_OtherProposals(unlist(similar_proposal[4]))
            amount_of_commonWords(similar_proposal$common_words_weighted)
            # getting the title of the most similar proposal
            corpus_raw <- read.csv(proposal_df_path$datapath)
            colnames(corpus_raw) <- c('id', 'author', 'proposal_title', 'text')
            title <- corpus_raw[corpus_raw$id == similar_proposal$most_similar_proposal ,]$proposal_title
            proposalTitle(title)

            load_data()

            }else{
            shinyalert::shinyalert("Oops!", "Please make sure the id exists in the proposal dataframe.", type = "error")
          }
        } else if (input$checkUsing == "BackgroundInformation") {
          # show loading screen
          shinyjs::show("loading_page")
          shinyjs::hide("main_content")

          # getting selected file name
          similar_proposal <- getSimilarProposalsFromCSV(proposalDataFile = proposal_df_path$datapath,
                                                         idForFileBeingChecked='TEMPORARYID16352',
                                                         background_info=input$backgroundInformation)
          commonKeyWords_OtherProposals(unlist(similar_proposal[4]))
          amount_of_commonWords(similar_proposal$common_words_weighted)
          # getting the title of the most similar proposal
          corpus_raw <- read.csv(proposal_df_path$datapath)
          colnames(corpus_raw) <- c('id', 'author', 'proposal_title', 'text')
          title <- corpus_raw[corpus_raw$id == similar_proposal$most_similar_proposal ,]$proposal_title
          proposalTitle(title)

          load_data()
        } else if (input$checkUsing == "SimilarityReport"){
          # show loading screen
          shinyjs::show("loading_page")
          shinyjs::hide("main_content")

          proposal_df_path =  shinyFiles::parseFilePaths(volumes, input$DataframeProposalFile)
          similar_articles_df(getSimilarProposalsFromCSV(proposalDataFile = proposal_df_path$datapath,
                                                         type = 'SimilarityReport'))
          load_data()

        }
      }
  })

  output$downloadReport <- downloadHandler(
    filename = 'similarity-report.pdf',
    content = function(file) {
      # print(getwd())
      # print(dirname(sys.frame(1)$ofile))
      src <- normalizePath(system.file("rmd", "report.Rmd", package = "ProposalSimilarityChecker"))
      # # # temporarily switch to the temp dir, in case you do not have write
      # # # permission to the current working directory
      owd <- setwd(tempdir())
      on.exit(setwd(owd))
      # file.copy(paste0(tempfile(), ".pdf"), file, overwrite = TRUE)
      file.copy(src, 'report.Rmd', overwrite = TRUE)

      out <- rmarkdown::render(src,
                               rmarkdown::pdf_document())
      file.rename(out, file)
    }
  )

  # renders common keywords
  output$KeywordsOtherProposal <- renderUI ({
    HTML(getHTML_Keywords_OtherProposals(commonKeyWords_OtherProposals()))
  })

  # renders proposal title dynamically
  output$proposalTitle <- renderUI({
    HTML(getHTML_ProposalTitle(proposalTitle()))
  })

  # renders proposal list dynamically
  output$proposalList <- renderUI({
    # check if there were any common words found
    if (proposalDF() == "NO WORDS IN COMMON"){
      HTML('<p>NO WORDS IN COMMON</p>')
    }else{
      HTML(getHTML_proposalList(proposalDF()))
    }
  })

  # renders similarity report dynamically
  output$similarityReportPreview <- renderUI({
    HTML(getHTML_similarityReportPreview(similar_articles_df()))
  })

  # renders similarity level dynamically
  output$similarityLevel <- renderUI({
    HTML(getHTML_Similarity_indicator_otherproposal(amount_of_commonWords()))
  })

}