It is now possible to paste data as formatted code in RStudio Server/Cloud, and ssh sessions. (Thanks Garrick Aden-Buie and Jon Harmon!)

There is no way to connect your local machine’s clipboard to the R session running on a cloud instance. So datapasta cannot work as per local usage.

What has been developed instead are two kinds of fallback behaviour that capture most of the magic!

## Fallback 1: Text selection

If datapasta detects the clipboard is not available and the editor is RStudio, it will check to see if there is a text selection active (highlighted text).

Given a selection, it will read that text instead of the clipboard and proceed as per the called function, e.g. attempt to parse that text as data and paste as a tribble if tribble_paste was called.

The imagined workflow here is:

1. Copy data from external application
2. Paste as text into an RStudio pane (exposing it to your remote R session)
3. Select text and call datapasta function/addin
4. Text is transformed into code defining data.

## Fallback 2: Pop-up text editor

Again if there is no clipboard available, but in this case there is no text selected in RStudio (or you’re not in RStudio), datapasta will open a temporary file and then call utils::file.edit() on it.

In RStudio server/cloud this will pop-up a modal text editor window where text can be pasted. When the editor is closed, the text will be read by datapasta and then things will proceed as if that data had come from the clipboard. The temporary file is deleted immediately after being read.

If you’re not in RStudio, but on the command line in an ssh session, this will open a command line text editor instead of the modal window.

The workflow in this case is:

1. Copy data from external application
The first time the pop-up text editor is about to be opened in each session, datapasta issues a message in a modal dialogue to explain what is about to happen. This can be disabled with:
options(datapasta.quiet_manual_paste = TRUE)