R/seqtype.R

In rOpenGov/bibliographica: Bibliographic Data Analysis

seqtype <- function (z) {

  # series does not have any
  if (length(z)==0) {
    sequence.type <- "empty"
  } else {

    # Determine page count categories 
    # increasing / series / etc.
    sequence.type <- NA

    # Recognize increasing sequence
    increasing <- is.increasing(z)
    if (increasing) {
      # Use if to avoid unnecessary calculations.
      # Increasing series cannot be decreasing at the same time.
      decreasing <- FALSE
    } else {      
      decreasing <- is.decreasing(z)    
    }
    
    # Recognize series (ie. two-number sequences with dashes)
    series <- length(grep("^[0-9]+-[0-9]+$", z))>0 

    # Recognize single number
    single.number <- length(z) == 1 && is.numeric(suppressWarnings(as.numeric(z[[1]])) )

    # sequence has numbers without dashes
    if (!series && !increasing) { sequence.type <- "sequence" }
    if (!series && increasing)  { sequence.type <- "increasing.sequence" }
    if (!series && decreasing)  { sequence.type <- "decreasing.sequence" }    
    if (single.number) { sequence.type <- "sequence"}

    # series has dashes
    if (series && !increasing) { sequence.type <- "series" } 
    if (series && increasing) { sequence.type <- "increasing.series" }
    if (series && decreasing) { sequence.type <- "decreasing.series" }    

    # If there are several arabic elements and at least one dash among them, then use maximum
    multiple <- length(z) > 1
    if (series && multiple) { sequence.type <- "series" }

  }

  sequence.type
}