Cmd-Enter to run the current statement stops working mid session

[srhrnkn]

I can't figure out a pattern to this but I'm seeing the command-enter command, which should run the "current" statement if no code is highlighted, stop working in the middle of a session (after working perfectly well for a while). It'll send only the current line to the console rather than multiple lines of piped code (as one example). It seems to be file-specific but I can't figure out what, if anything, in the contents of the file would be breaking the ability to "statements." Is there somewhere this feature is documented where I can read more about how it should work?

[austin3dickey]

Hey @srhrnkn, thanks for the question. Just to confirm, is this in a .R file? And could you please paste the output of the About Positron menu item?

I don't think it's expected that when you use Cmd+Enter at first it displays one behavior, and then changes to a different behavior on the same code later in the same session. Some thoughts:

• Could this be related to a possible syntax error, like in executeCode: Surprising behaviour when evaluating syntactically incorrect code #5023 (comment)?
• If you're willing to share, I'd be interested in the contents of a file that doesn't execute as expected, so we can try to reproduce.
• Next time it happens could you check out the logs? I'm not exactly sure if/where anything would show up, but some possibly-interesting channels are listed in those docs.

[srhrnkn]

About Positron output:
Positron Version: 2025.07.0 (Universal) build 204
Code - OSS Version: 1.100.3
Commit: 03ae7a5
Date: 2025-06-30T16:28:59.945Z
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Darwin arm64 23.6.0

Yes, it's a .R file. I do have a hunch that it's related to syntax somewhere else in the file, though I couldn't find anything wonky yesterday when it was happening. (It stopped again!) It does sound related to that discussion - it's basically sending an incomplete command to the console & then waiting for input. But it's not syntactically incorrect code - highlighting the full statement and running it works.

I will try to capture the code of a non-working file & the logs next time it happens. thanks!

[ASNDataAnalytics]

@srhrnkn, thanks for posting this. I was experiencing the same behavior and "believe" I may have found the issue behind it. I was trying to execute something as simple (and syntactically correct) as
palmerpenguins::penguins |> pull(bill_length_mm)
with Ctrl/Cmd + Enter and it would only send one line at a time (as you described). However, somewhere much later in unrelated code was an error which I'm guessing stops the execution behavior since the entire file isn't error-free. @austin3dickey, I'm guessing this is enforced with a linter? Regardless, it's a relief to (hopefully) find a reason behind the behavior.

Thanks,
Kurtis

[juliasilge]

I do think this is #5023, yep. I just moved that out of our "maybe someday" milestone into the one for this year, so that the R folks on our team can evaluate this for improvement relative to other areas of focus.

Thank you both for these reports!

[srhrnkn]

Yes, cribbing from #5023, this seems to replicate the issue (on its own in a .R file) - with the cursor at the first line, cmd-enter only sends the first line of the statement. If I close the parentheses within the function, it sends both lines.

mtcars |>
select(qsec)

syntactically_bad <- function( ) {
c(
}