getData

# Usage:
# myData <- getData("pathtofile")

getData <- function(file){
  
  # variables of interest to read in
  variables <- c("PMID","DP  ","JID ", "MH  ")
  
  con <- file(file, "rt")
  test <- substr(scan(con, what = "", sep = "\n", quiet = TRUE), 1, 4)
  close(con)
  
  varLines <- which(test %in% variables) # identifying variables of interest in file
  
  # saving information from variables of interest
  
  #create a skip vector for lines to skip in next scan
  #skip argument in scan skips from the last entry read, hence varlines[i]-varlines[i-1]-1
  #if statement ensures blank lines don't get stored as data 
  
  skip <- vector()
  skip[1] <- 1
  for(i in 2:length(varLines)){
    skip[i] <- (varLines[i]-varLines[i-1]-1)
  }
  
  con <- file(file, "rt")
  entry <- list()
  
  
  for(i in 1:length(varLines)){
    entry[i] <- scan(con, sep="\n", what="", quiet=TRUE, skip = skip[i],blank.lines.skip=FALSE,nlines=1)
    if(entry[i]==""){
      entry[i] <- scan(con,sep="\n",what="",quiet=TRUE,skip=0,nlines=1,blank.lines.skip=FALSE)
    }
  }
  
  close(con)
  
  # creates list varPlusInfo of each variable name and corresponding information as two vectors inside one list element
  varPlusInfo <- list()
  for(i in 1:length(entry)){
    x <- unlist(entry[i])
    varPlusInfo[[i]] <- substr(x,1,4)
    varPlusInfo[[i]][2] <- substr(x,7,nchar(x))
  }
  
  PMIDindex <- vector()
  for(i in 1:length(varPlusInfo)){
    if(varPlusInfo[[i]][1] == "PMID"){
      PMIDindex <- c(PMIDindex,i) 
    }
  }
  
  splitentry <- function(full,ind){
    if(ind==length(PMIDindex)){
      temp <- full[PMIDindex[ind]:length(full)]
    }
    else{
      temp <- full[PMIDindex[ind]:(PMIDindex[ind+1]-1)]
    }
    return(temp)
  }
  
  #create empty dataframe
  d <- data.frame(PMID=vector(),DP=vector(),JID=vector(),MH=I(list()))
  
  for(i in 1:length(PMIDindex)){
    
    temp <- splitentry(varPlusInfo,i)
    
    MH <- vector()
    
    for(j in 1:length(temp)){
      if(temp[[j]][1]=="MH  "){MH[j] <- temp[[j]][2]}
    }
    
    MH <- MH[!is.na(MH)] #remove NAs
    
    for(k in 1:length(temp)){
      if(temp[[k]][1]=="PMID"){
        d[i,1] <- temp[[k]][2]
      }
      if(temp[[k]][1]=="DP  "){
        d[i,2] <- temp[[k]][2]
      }
      if(temp[[k]][1]=="JID "){
        d[i,3] <- temp[[k]][2]
      }
      if(temp[[k]][1]=="MH  "){
        d[[i,4]] <- MH
      }
    }
  }
  return(d)
}