wcgrads: Williams College Graduating Seniors

pdf2txt <- function(file) {

#' @title PDF to Text
#' @description pdf2txt() converts Williams College Course Catalogs from .pdf to .txt
#' @param file The file, a Williams College Course Catalog, to be converted from pdf
#' @return The output of pdftotxt is a txt version of the Williams College Course Catalog
#' in the same location as the .pdf that it took as input.
#' @usage pdf2txt("file")
#' @export

pdftotext <- system.file("bin/pdftotext", package = "wcgrads", mustWork = TRUE)
pdf.file <- system.file(paste0("extdata/", file, ".pdf"), package = "wcgrads", mustWork = TRUE)
pdf <- file.path(pdf.file)
system(paste("\"", pdftotext, "\" \"", pdf, "\""," -raw", sep=""), wait = TRUE)
text.file <- system.file(paste0("extdata/", file, ".txt"), package = "wcgrads", mustWork = TRUE)

}

findstrings <- function(file){

#' @title Find Strings
#' @description findstrings() cuts the txt version of the Williams College Course Catalog to isolate the section
#'  of the Catalog that is just list of graduating seniors.
#' @param file The file, a Williams College Course Catalog, in .txt form, to be cut down to just the list of graduated seniors.
#' @return findstrings(file) returns a tbl_df() of the section of the Williams College Course Catalog that contains the list of graduated seniors.
#' @usage findstrings("file")
#' @export

x <- read.delim(file, quote = "", stringsAsFactors = FALSE)
colnames(x) <- "strings"
#Find beginning of list of names
y <- as.numeric(which(x$strings == "Bachelor of Arts, Summa Cum Laude")) + 1
#Find end of list of names
z <- as.numeric(which(x$strings == "CONFERRING OF HONORARY DEGREES")) - 1
#Cut out the section of names
w <- tbl_df(x[y:z,])
colnames(w) <- "strings"

return(w)

}

makestrings <- function(file){

#' @title Make Strings
#' @description makestrings() takes the list of names given by findstrings() and combines
#' incomplete entries to form complete entries of graduating seniors
#' @param file The file, a section Williams College Course Catalog, in .txt form. This section may not be complete lines.
#' @return The output of makestrings is a tbl_df() of complete lines from the graduated seniors section of the Williams College Course Catalog.
#' @usage makestrings("file")
#' @import stringr dplyr
#' @export

w <- file
#Examine strings to see if they have various substrings
comma <- str_detect(w$strings, ", with")
contract <- str_detect(w$strings, "Contract")
colon <- str_detect(w$strings, ":")
in. <- str_detect(w$strings, "in ")
inn <- str_detect(w$strings, " in")
inn. <- str_detect(w$strings, " in ")
and. <- str_detect(w$strings, " and ")
and <- str_detect(w$strings, "and ")
studies <- str_detect(w$strings, "Studies")
honors <- str_detect(w$strings, "honors")
others <- str_detect(w$strings, "Sigma|Degrees|Phi Beta|_____|42")
space.count <- str_count(w$strings, " ")
#Assign errors based on substrings
w$error <- ifelse(comma == TRUE & in. == FALSE & inn == TRUE, 1, 0)
w$error <- ifelse(comma == TRUE & (in. + inn) == 0, 2, w$error)
w$error <- ifelse(space.count == 0, 3, w$error)
w$error <- ifelse(comma == TRUE & inn. == FALSE & (is.na(w$error) == TRUE), 4, w$error)
w$error <- ifelse(contract == TRUE, 5, w$error)
w$error <- ifelse(and. == TRUE & honors == FALSE, 6, w$error)
w$error <- ifelse(comma == FALSE & colon == TRUE, 7, w$error)
w$error <- ifelse(others == TRUE, 8, w$error)
w$error <- ifelse(honors == TRUE & inn. == FALSE & w$error == 0, 9, w$error)
w$error <- ifelse(comma == FALSE & honors == TRUE & w$error == 0, 10, w$error)
w$error <- ifelse(comma == FALSE & studies == TRUE & w$error == 0, 11, w$error)
w$error <- ifelse(and == TRUE & space.count == 1, 12, w$error)
#Get errors around each line
w$error.prior <- dplyr::lag(w$error, default = 0)
w$error.post <- dplyr::lead(w$error, default = 0)
#Combine broken strings by error code
w$names <- ifelse(w$error == 1 | w$error == 2 | w$error == 4 | w$error == 9,
                    paste(w$strings, dplyr::lead(w$strings, 1), sep = " "), NA)
w$names <- ifelse(w$error == 3 | w$error == 7 | w$error == 10 | w$error == 11 | w$error == 12,
                    paste(dplyr::lag(w$strings, 1), w$strings, sep = " "), w$names)
w$names <- ifelse(w$error == 7 & (w$error.post == 11 | w$error.post == 6), paste(dplyr::lag(w$strings, 1), w$strings,
                                                             dplyr::lead(w$strings, 1), sep = " "), w$names)
w$names <- ifelse(is.na(w$names) == TRUE, w$strings, w$names)

w <- tbl_df(w)
#Cut out incomplete and/or duplicate strings
v <- w %>%
  filter(error.prior != 2, error.prior != 9,
          error != 5, error != 6, error != 8,
          error.post != 12, error.post != 11, error.post != 10,
          (error == 1 & error.post == 7) == FALSE,
          (error == 11 & error.prior == 7) == FALSE,
          (error == 0 & error.prior == 1) == FALSE,
          (error == 0 & error.post == 3) == FALSE)

v <- tbl_df(unique(v$names))

return(v)

}