docx_replace.R

# utils ----
is_scalar_character <- function( x ) {
  is.character(x) && length(x) == 1
}
is_scalar_logical <- function( x ) {
  is.logical(x) && length(x) == 1
}



# functions ----


#' @export
#' @title Replace text at a bookmark location
#' @description Replace text content enclosed in a bookmark
#' with different text. A bookmark will be considered as valid if enclosing words
#' within a paragraph; i.e., a bookmark along two or more paragraphs is invalid,
#' a bookmark set on a whole paragraph is also invalid, but bookmarking few words
#' inside a paragraph is valid.
#' @param x a docx device
#' @param bookmark bookmark id
#' @param value the replacement string, of type character
#' @examples
#' doc <- read_docx()
#' doc <- body_add_par(doc, "a paragraph to replace", style = "centered")
#' doc <- body_bookmark(doc, "text_to_replace")
#' doc <- body_replace_text_at_bkm(doc, "text_to_replace", "new text")
body_replace_text_at_bkm <- function( x, bookmark, value ){
  stopifnot(is_scalar_character(value), is_scalar_character(bookmark))
  xml_replace_text_at_bkm(node = x$doc_obj$get(), bookmark = bookmark, value = value)
  x
}


#' @export
#' @rdname body_replace_text_at_bkm
#' @examples
#'
#'
#' # demo usage of bookmark and images ----
#' template <- system.file(package = "officer", "doc_examples/example.docx")
#'
#' img.file <- file.path( R.home("doc"), "html", "logo.jpg" )
#'
#' doc <- read_docx(path = template)
#' doc <- headers_replace_img_at_bkm(x = doc, bookmark = "bmk_header",
#'                                   value = external_img(src = img.file, width = .53, height = .7))
#' doc <- footers_replace_img_at_bkm(x = doc, bookmark = "bmk_footer",
#'                                   value = external_img(src = img.file, width = .53, height = .7))
#' print(doc, target = tempfile(fileext = ".docx"))
#'
body_replace_img_at_bkm <- function( x, bookmark, value ){
  stopifnot(inherits(x, "rdocx"),
            is_scalar_character(bookmark),
            inherits(value, "external_img"))
  docxpart_replace_img_at_bkm(node = x$doc_obj$get(), bookmark = bookmark, value = value)
  x
}

xml_replace_text_at_bkm <- function(node, bookmark, value){

  text <- enc2utf8(value)
  xpath_ <- sprintf("//w:bookmarkStart[@w:name='%s']", bookmark)
  bm_start <- xml_find_first(node, xpath_)
  if( inherits(bm_start, "xml_missing") ) {
    return(FALSE)
  }

  str_ <- sprintf("//w:bookmarkStart[@w:name='%s']/following-sibling::w:r", bookmark )
  following_start <- sapply( xml_find_all(node, str_), xml_path )
  str_ <- sprintf("//w:bookmarkEnd[@w:id='%s']/preceding-sibling::w:r", xml_attr(bm_start, "id") )
  preceding_end <- sapply( xml_find_all(node, str_), xml_path )

  match_path <- base::intersect(following_start, preceding_end)
  if( length(match_path) < 1 )
    stop("could not find any bookmark ", bookmark, " located INSIDE a single paragraph" )

  run_nodes <- xml_find_all(node, paste0( match_path, collapse = "|" ) )

  for(node in run_nodes[setdiff(seq_along(run_nodes), 1)])
    xml_remove(node)

  xml_text(run_nodes[[1]] ) <- text
  TRUE
}

docxpart_replace_img_at_bkm <- function(node, bookmark, value) {
  stopifnot(is_scalar_character(bookmark))
  stopifnot(inherits(value, "external_img"))

  xpath_ <- sprintf("//w:bookmarkStart[@w:name='%s']", bookmark)
  bm_start <- xml_find_first(node, xpath_)
  if (inherits(bm_start, "xml_missing")) {
    stop("cannot find bookmark ", shQuote(bookmark), call. = FALSE)
  }

  str_ <- sprintf("//w:bookmarkStart[@w:name='%s']/following-sibling::w:r", bookmark)
  following_start <- sapply(xml_find_all(node, str_), xml_path)
  str_ <- sprintf("//w:bookmarkEnd[@w:id='%s']/preceding-sibling::w:r", xml_attr(bm_start, "id"))
  preceding_end <- sapply(xml_find_all(node, str_), xml_path)

  match_path <- base::intersect(following_start, preceding_end)
  if (length(match_path) < 1) {
    stop("could not find any bookmark ", bookmark, " located INSIDE a single paragraph")
  }

  out <- to_wml(value, add_ns = TRUE)

  run_nodes <- xml_find_all(node, paste0(match_path, collapse = "|"))
  for (node in run_nodes[setdiff(seq_along(run_nodes), 1)]) {
    xml_remove(node)
  }
  xml_replace(run_nodes[[1]], as_xml_document(out))
}

#' @export
#' @rdname body_replace_text_at_bkm
headers_replace_text_at_bkm <- function( x, bookmark, value ){
  stopifnot(is_scalar_character(value), is_scalar_character(bookmark))
  for(header in x$headers){
    xml_replace_text_at_bkm(node = header$get(), bookmark = bookmark, value = value)
  }
  x
}

#' @export
#' @rdname body_replace_text_at_bkm
headers_replace_img_at_bkm <- function( x, bookmark, value ){
  for(header in x$headers){
    docxpart_replace_img_at_bkm(node = header$get(), bookmark = bookmark, value = value)
  }
  x
}



#' @export
#' @rdname body_replace_text_at_bkm
footers_replace_text_at_bkm <- function( x, bookmark, value ){
  stopifnot(is_scalar_character(value), is_scalar_character(bookmark))
  for(footer in x$footers){
    xml_replace_text_at_bkm(node = footer$get(), bookmark = bookmark, value = value)
  }
  x
}

#' @export
#' @rdname body_replace_text_at_bkm
footers_replace_img_at_bkm <- function( x, bookmark, value ){
  for(footer in x$footers){
    docxpart_replace_img_at_bkm(node = footer$get(), bookmark = bookmark, value = value)
  }
  x
}

#' @export
#' @title Add plots at bookmark location in a 'Word' document
#' @description
#' Use these functions if you want to replace a paragraph containing
#' a bookmark with a 'ggplot' or a base plot.
#' @param value a ggplot object for body_replace_gg_at_bkm() or a set plot instructions
#' body_replace_plot_at_bkm(), see plot_instr().
#' @param bookmark bookmark id
#' @param keep Should the bookmark be preserved? Defaults to `FALSE`,
#' i.e.the bookmark will be lost as a side effect.
#' @inheritParams body_add_gg
#'
#' @examples
#' if (require("ggplot2")) {
#'   gg_plot <- ggplot(data = iris) +
#'     geom_point(mapping = aes(Sepal.Length, Petal.Length))
#'
#'   doc <- read_docx()
#'   doc <- body_add_par(doc, "insert_plot_here")
#'   doc <- body_bookmark(doc, "plot")
#'   doc <- body_replace_gg_at_bkm(doc, bookmark = "plot", value = gg_plot)
#'   print(doc, target = tempfile(fileext = ".docx"))
#' }
body_replace_gg_at_bkm <- function(x, bookmark, value, width = 6, height = 5,
                                   res = 300, style = "Normal", scale = 1,
                                   keep = FALSE, ...) {
  x <- cursor_bookmark(x, bookmark)
  x <- body_add_gg(
    x = x, value = value, width = width, height = height,
    res = res, style = style, scale = scale, pos = "on", ...
  )
  if (keep) {
    x <- body_bookmark(x, bookmark)
  }

  x
}

#' @export
#' @rdname body_replace_gg_at_bkm
#' @examples
#' doc <- read_docx()
#' doc <- body_add_par(doc, "insert_plot_here")
#' doc <- body_bookmark(doc, "plot")
#' if (capabilities(what = "png")) {
#'   doc <- body_replace_plot_at_bkm(
#'     doc, bookmark = "plot",
#'     value = plot_instr(
#'       code = {
#'         barplot(1:5, col = 2:6)
#'       }
#'     )
#'   )
#' }
#' print(doc, target = tempfile(fileext = ".docx"))
body_replace_plot_at_bkm <- function(x, bookmark, value, width = 6, height = 5,
                                   res = 300, style = "Normal",
                                   keep = FALSE, ...) {
  x <- cursor_bookmark(x, bookmark)
  x <- body_add_plot(
    x = x, value = value, width = width, height = height,
    res = res, style = style, pos = "on", ...
  )
  if (keep) {
    x <- body_bookmark(x, bookmark)
  }

  x
}

#' @export
#' @title Replace text anywhere in the document
#' @description Replace text anywhere in the document, or at a cursor.
#'
#' Replace all occurrences of old_value with new_value. This method
#' uses \code{\link{grepl}}/\code{\link{gsub}} for pattern matching; you may
#' supply arguments as required (and therefore use \code{\link{regex}} features)
#' using the optional \code{...} argument.
#'
#' Note that by default, grepl/gsub will use \code{fixed=FALSE}, which means
#' that \code{old_value} and \code{new_value} will be interepreted as regular
#' expressions.
#'
#' \strong{Chunking of text}
#'
#' Note that the behind-the-scenes representation of text in a Word document is
#' frequently not what you might expect! Sometimes a paragraph of text is broken
#' up (or "chunked") into several "runs," as a result of style changes, pauses
#' in text entry, later revisions and edits, etc. If you have not styled the
#' text, and have entered it in an "all-at-once" fashion, e.g. by pasting it or
#' by outputing it programmatically into your Word document, then this will
#' likely not be a problem. If you are working with a manually-edited document,
#' however, this can lead to unexpected failures to find text.
#'
#' You can use the officer function \code{\link{docx_show_chunk}} to
#' show how the paragraph of text at the current cursor has been chunked into
#' runs, and what text is in each chunk. This can help troubleshoot unexpected
#' failures to find text.
#' @seealso \code{\link{grep}}, \code{\link{regex}}, \code{\link{docx_show_chunk}}
#' @author Frank Hangler, \email{frank@plotandscatter.com}
#' @param x a docx device
#' @param old_value the value to replace
#' @param new_value the value to replace it with
#' @param only_at_cursor if \code{TRUE}, only search-and-replace at the current
#' cursor; if \code{FALSE} (default), search-and-replace in the entire document
#' (this can be slow on large documents!)
#' @param warn warn if \code{old_value} could not be found.
#' @param ... optional arguments to grepl/gsub (e.g. \code{fixed=TRUE})
#' @examples
#' doc <- read_docx()
#' doc <- body_add_par(doc, "Placeholder one")
#' doc <- body_add_par(doc, "Placeholder two")
#'
#' # Show text chunk at cursor
#' docx_show_chunk(doc)  # Output is 'Placeholder two'
#'
#' # Simple search-and-replace at current cursor, with regex turned off
#' doc <- body_replace_all_text(doc, old_value = "Placeholder",
#'   new_value = "new", only_at_cursor = TRUE, fixed = TRUE)
#' docx_show_chunk(doc)  # Output is 'new two'
#'
#' # Do the same, but in the entire document and ignoring case
#' doc <- body_replace_all_text(doc, old_value = "placeholder",
#'   new_value = "new", only_at_cursor=FALSE, ignore.case = TRUE)
#' doc <- cursor_backward(doc)
#' docx_show_chunk(doc) # Output is 'new one'
#'
#' # Use regex : replace all words starting with "n" with the word "example"
#' doc <- body_replace_all_text(doc, "\\bn.*?\\b", "example")
#' docx_show_chunk(doc) # Output is 'example one'
body_replace_all_text <- function( x, old_value, new_value,
                                   only_at_cursor = FALSE,
                                   warn = TRUE, ... ){
  stopifnot(is_scalar_character(old_value),
            is_scalar_character(new_value),
            is_scalar_logical(only_at_cursor))

  oldValue <- enc2utf8(old_value)
  newValue <- enc2utf8(new_value)

  replacement_count <- 0

  base_node <- if (only_at_cursor) {
    docx_current_block_xml(x)
  } else {
    docx_body_xml(x)
  }

  # For each matching text node...
  for (text_node in xml_find_all(base_node, ".//w:t")) {
    # ...if it contains the oldValue...
    if (grepl(oldValue, xml_text(text_node), ...)) {
      replacement_count <- replacement_count + 1
      # Replace the node text with the newValue.
      xml_text(text_node) <- gsub(oldValue, newValue, xml_text(text_node), ...)
    }
  }

  # Alert the user if no replacements were made.
  if (replacement_count == 0 && warn) {
    search_zone_text <- if (only_at_cursor) "at the cursor." else "in the document."
    warning("Found 0 instances of '", oldValue, "' ", search_zone_text)
  }

  x
}

#' @export
#' @title Show underlying text tag structure
#' @description Show the structure of text tags at the current cursor. This is
#' most useful when trying to troubleshoot search-and-replace functionality
#' using \code{\link{body_replace_all_text}}.
#' @seealso \code{\link{body_replace_all_text}}
#' @param x a docx device
#' @examples
#' doc <- read_docx()
#' doc <- body_add_par(doc, "Placeholder one")
#' doc <- body_add_par(doc, "Placeholder two")
#'
#' # Show text chunk at cursor
#' docx_show_chunk(doc)  # Output is 'Placeholder two'
docx_show_chunk <- function( x ){
  cursor_elt <- docx_current_block_xml(x)
  text_nodes <- xml_find_all(cursor_elt, ".//w:t")
  msg <- paste0(length(text_nodes), " text nodes found at this cursor.")
  msg_detail <- ""
  for (text_node in text_nodes) {
    msg_detail <- paste0( msg_detail,
                          paste0("\n  <w:t>: '",
                                 xml_text(text_node), "'") )
  }
  message(paste(msg, msg_detail))
  invisible(x)
}




#' @export
#' @rdname body_replace_all_text
#' @section header_replace_all_text:
#' Replacements will be performed in each header of all sections.
headers_replace_all_text <- function( x, old_value, new_value, only_at_cursor = FALSE, warn = TRUE,  ... ){
  stopifnot(is_scalar_character(old_value),
            is_scalar_character(new_value),
            is_scalar_logical(only_at_cursor))

  oldValue <- enc2utf8(old_value)
  newValue <- enc2utf8(new_value)

  for(header in x$headers){

    replacement_count <- 0

    base_node <- header$get()

    # For each matching text node...
    for (text_node in xml_find_all(base_node, ".//w:t")) {
      # ...if it contains the oldValue...
      if (grepl(oldValue, xml_text(text_node), ...)) {
        replacement_count <- replacement_count + 1
        # Replace the node text with the newValue.
        xml_text(text_node) <- gsub(oldValue, newValue, xml_text(text_node), ...)
      }
    }

    # Alert the user if no replacements were made.
    if (replacement_count == 0 && warn) {
      search_zone_text <- if (only_at_cursor) "at the cursor." else "in the document."
      warning("Found 0 instances of '", oldValue, "' ", search_zone_text)
    }

  }

  x
}
#' @export
#' @rdname body_replace_all_text
#' @section header_replace_all_text:
#' Replacements will be performed in each footer of all sections.
footers_replace_all_text <- function( x, old_value, new_value, only_at_cursor = FALSE, warn = TRUE, ... ){
  stopifnot(is_scalar_character(old_value),
            is_scalar_character(new_value),
            is_scalar_logical(only_at_cursor))

  oldValue <- enc2utf8(old_value)
  newValue <- enc2utf8(new_value)

  for(footer in x$footers){
    replacement_count <- 0

    base_node <- footer$get()

    # For each matching text node...
    for (text_node in xml_find_all(base_node, ".//w:t")) {
      # ...if it contains the oldValue...
      if (grepl(oldValue, xml_text(text_node), ...)) {
        replacement_count <- replacement_count + 1
        # Replace the node text with the newValue.
        xml_text(text_node) <- gsub(oldValue, newValue, xml_text(text_node), ...)
      }
    }

    # Alert the user if no replacements were made.
    if (replacement_count == 0 && warn) {
      search_zone_text <- if (only_at_cursor) "at the cursor." else "in the document."
      warning("Found 0 instances of '", oldValue, "' ", search_zone_text)
    }
  }

  x
}

#' @export
#' @title Remove unused media from a document
#' @description The function will scan the media
#' directory and delete images that are not used
#' anymore. This function is to be used when images
#' have been replaced many times.
#' @param x \code{rdocx} or \code{rpptx} object
#' @keywords internal
sanitize_images <- function(x){

  rel_files <- list.files(x$package_dir, pattern = "\\.xml.rels$", recursive = TRUE, full.names = TRUE)

  image_files <- lapply(rel_files,
                        function(x){
                          zz <- read_xml(x)
                          rels <- xml_children(zz)
                          rels <- rels[xml_attr(rels, "Type") %in% "http://schemas.openxmlformats.org/officeDocument/2006/relationships/image"]
                          xml_attr(rels, "Target")
                        })
  image_files <- unique(unlist(image_files))

  if(inherits(x, "rdocx")){
    base_doc <- file.path(x$package_dir, "word")
    existing_img <- list.files(
      file.path(base_doc, "media"),
      pattern = "\\.(png|jpg|jpeg|eps|emf)$",
      ignore.case = TRUE,
      recursive = TRUE, full.names = TRUE)
    existing_img <- gsub(paste0(base_doc, "/"), "", existing_img, fixed = TRUE)
    unlink(file.path(base_doc,
                     setdiff(image_files, existing_img)
                     ), force = TRUE)
  } else if(inherits(x, "rpptx")){
    base_doc <- file.path(x$package_dir, "ppt")
    existing_img <- list.files(
      file.path(base_doc, "media"),
      pattern = "\\.(png|jpg|jpeg|eps|emf)$",
      ignore.case = TRUE,
      recursive = TRUE, full.names = TRUE)
    existing_img <- gsub(paste0(base_doc, "/"), "",
                         existing_img, fixed = TRUE)
    unlink(file.path(base_doc, setdiff(image_files, existing_img)), force = TRUE)
  }
  x
}