Added compatibility with raw vectors of pdf (this is very useful when reading files for example from aws.s3)

DESCRIPTION

@@ -41,4 +41,4 @@ Suggests:
     testthat
 SystemRequirements: Java (>= 7.0)
 VignetteBuilder: knitr
-RoxygenNote: 6.0.1
+RoxygenNote: 7.1.1

R/extract_metadata.R

@@ -1,6 +1,6 @@
 #' @title extract_metadata
 #' @description Extract metadata from a file
-#' @param file A character string specifying the path or URL to a PDF file.
+#' @param file A character string specifying the path or URL to a PDF file, or raw vector with pdf data.
 #' @param password Optionally, a character string containing a user password to access a secured PDF.
 #' @param copy Specifies whether the original local file(s) should be copied to
 #' \code{tempdir()} before processing. \code{FALSE} by default. The argument is

R/extract_tables.R

@@ -1,6 +1,6 @@
 #' @title extract_tables
 #' @description Extract tables from a file
-#' @param file A character string specifying the path or URL to a PDF file.
+#' @param file A character string specifying the path or URL to a PDF file, or raw vector with pdf data.
 #' @param pages An optional integer vector specifying pages to extract from.
 #' @param area An optional list, of length equal to the number of pages specified, where each entry contains a four-element numeric vector of coordinates (top,left,bottom,right) containing the table for the corresponding page. As a convenience, a list of length 1 can be used to extract the same area from all (specified) pages. Only specify \code{area} xor \code{columns}.
 #' @param columns An optional list, of length equal to the number of pages specified, where each entry contains a numeric vector of horizontal (x) coordinates separating columns of data for the corresponding page. As a convenience, a list of length 1 can be used to specify the same columns for all (specified) pages. Only specify \code{area} xor \code{columns}.

R/extract_text.R

@@ -1,6 +1,6 @@
 #' @title extract_text
 #' @description Extract text from a file
-#' @param file A character string specifying the path or URL to a PDF file.
+#' @param file A character string specifying the path or URL to a PDF file, or raw vector with pdf data.
 #' @param pages An optional integer vector specifying pages to extract from.
 #' @param area An optional list, of length equal to the number of pages specified, where each entry contains a four-element numeric vector of coordinates (top,left,bottom,right) containing the table for the corresponding page. As a convenience, a list of length 1 can be used to extract the same area from all (specified) pages.
 #' @param password Optionally, a character string containing a user password to access a secured PDF.

R/get_page_dims.R

@@ -1,7 +1,7 @@
 #' @rdname get_page_dims
 #' @title Page length and dimensions
 #' @description Get Page Length and Dimensions
-#' @param file A character string specifying the path or URL to a PDF file.
+#' @param file A character string specifying the path or URL to a PDF file, or raw vector with pdf data.
 #' @param pages An optional integer vector specifying pages to extract from.
 #' @param doc Optionally,, in lieu of \code{file}, an rJava reference to a PDDocument Java object.
 #' @param password Optionally, a character string containing a user password to access a secured PDF.

R/utils.R

@@ -19,17 +19,23 @@ localize_file <- function(path, copy = FALSE, quiet = TRUE) {
     path
 }
 
-load_doc <- function(file, password = NULL, copy = FALSE) {
+load_doc <- function(file = NULL, password = NULL, copy = FALSE) {
+  pdfDocument <- new(J("org.apache.pdfbox.pdmodel.PDDocument"))
+  if(typeof(file) != "raw"){
     localfile <- localize_file(path = file, copy = copy)
-    pdfDocument <- new(J("org.apache.pdfbox.pdmodel.PDDocument"))
     fileInputStream <- new(J("java.io.FileInputStream"), name <- localfile)
-    if (is.null(password)) {
-        doc <- pdfDocument$load(input = fileInputStream)
-    } else {
-        doc <- pdfDocument$load(input = fileInputStream, password = password)
-    }
-    pdfDocument$close()
-    doc
+  }
+  else {
+    fileInputStream <- new(J("java.io.ByteArrayInputStream"), buf = rJava::.jbyte(file))
+  }
+
+  if (is.null(password)) {
+    doc <- pdfDocument$load(input = fileInputStream)
+  } else {
+    doc <- pdfDocument$load(input = fileInputStream, password = password)
+  }
+  pdfDocument$close()
+  doc
 }
 
 make_pages <- function(pages, oe) {