Get correct line number when doing selection based runs about atom-script HOT 11 CLOSED

That bit of strangeness comes from using python -e.

If we wanted to get around that, we'd have to make a tempfile.

from atom-script.

Ach, weird new lines aren't interpreter as such:

 $ python -c "\n\nfoo" 
  File "<string>", line 1
    \n\nfoo
          ^
SyntaxError: unexpected character after line continuation character

from atom-script.

Color me corrected, I understand now.

from atom-script.

Looking back at the gif, it's because I selected the line above it. Check this out:

I think the demo gif is just a bit haphazard for when I'm selecting (I get nervous recording for the world I think).

from atom-script.

Yeah, line 2 makes sense, I was suggesting it could show line 22.

from atom-script.

ohhh, so sorry. Yeah, we should prepend newlines. Kind of hacky but it would be nice.

I think the reason python -c "\n\nfoo" didn't handle well is that it interpreted as a literal backslash n.

from atom-script.

@rgbkrk @hayd My changes from #107/#108/#109 should help:

"Selection Based":
  command: "python"
  args: (filename, selection) ->
    buffer_row_range = selection.getBufferRowRange()
    line_number = buffer_row_range.reduce (a, b) -> Math.min(a, b)
    line_number += 1
    n = 1
    # ... calculate the number of newlines to add ...
    [Array(n).join("\n"), filename]

from atom-script.

Just realized this happens for at least Ruby, Coffeescript, and Perl too... 😆 I guess the solution is to prepend the whitespace "\n" to all grammars? @rgbkrk What's your opinion?

from atom-script.

As long as the whitespace doesn't affect the interpretation of the code.

from atom-script.

Also, yuck. The only other solution would be if you could specify to an interpreter what line number to treat it as coming from. AFAIK, that's not a feature anywhere. Sadly, I think prepending newlines is what we need to do.

from atom-script.

Yeah. Should we make it an optional/boolean param on the getCode method (prepend set to true/false) and manually enable it per languages vs. polluting all of them with whitespace (some interpreters could be picky about it)?

from atom-script.