Consider adding `tagHasClass()`

[daattali]

Similarly to https://github.com/rstudio/htmltools/issues/389 , parsing out the classes of a tag in order to know if it contains a specific class is something I do often. I've seen in some clients code that they also try to do it, but I've seen it implemented wrongly many times (often people do a simple grep search within the classes string, which is incorrect).

I have the following function:

has_class <- function(tag, class) {
  if (!inherits(tag, "shiny.tag")) {
    stop("has_class: `tag` must be a shiny tag")
  }
  if (!nzchar(class)) {
    stop("has_class: `class` must be a non-empty string")
  }
  if (grepl("\\s", class)) {
    stop("has_class: `class` cannot contain any whitespace")
  }

  classes <- htmltools::tagGetAttribute(tag, "class")
  if (is.null(classes)) {
    return(FALSE)
  }
  classes <- strsplit(classes, "\\s+")[[1]]

Would you accept a PR for this function, or would you prefer not to? I'm also open to changing the behaviour (when to error vs return false / what to do when the class is an empty string / doing less error checking)

[cpsievert]

Now that tagQuery(div(class = "foo"))$hasClass("foo") is a thing, it doesn't seem worth adding (pinging @schloerke to see if he agrees)

[daattali]

That's true, if you think it's redundant then I'd understand. Although I still think it's a bit verbose for doing an operation that is common. Perhaps it's not common for other users though.