Fix prompt_start_row reset to 0 when opening a file without newline in Nushell

[nibon7]

Try to fix nushell/nushell#11399

[codecov]

Codecov Report

Attention: 10 lines in your changes are missing coverage. Please review.

Comparison is base (0046da9) 49.17% compared to head (ae68d9b) 49.26%.
Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #688      +/-   ##
==========================================
+ Coverage   49.17%   49.26%   +0.08%     
==========================================
  Files          46       46              
  Lines        7924     7911      -13     
==========================================
  Hits         3897     3897              
+ Misses       4027     4014      -13     

Files	Coverage Δ
src/painting/prompt_lines.rs	0.00% <0.00%> (ø)
src/painting/painter.rs	19.34% <0.00%> (+0.71%)	⬆️

[amtoine]

@nibon7
is it possible to test this patch with one of the Reedline examples?

[nibon7]

It could be reproduced by applying the patch below 😅 , just print something without newline.

When we print something without newline, then the column > 0 is true, self.prompt_start_raw = postition.1 + 1.

reedline/src/painting/painter.rs

Line 109 in a4bfaa5

let new_row = if column > 0 { row + 1 } else { row };

diff --git a/examples/demo.rs b/examples/demo.rs
index 2e7a402..00d3dae 100644
--- a/examples/demo.rs
+++ b/examples/demo.rs
@@ -201,6 +201,7 @@ fn main() -> std::io::Result<()> {
                     continue;
                 }
                 println!("Our buffer: {buffer}");
+                print!(" ");
                 #[cfg(any(feature = "sqlite", feature = "sqlite-dynlib"))]
                 if !buffer.is_empty() {
                     line_editor

[sholderbach]

That change makes sense and fixes your provided repro. Thanks @nibon7

[fdncred]

I think this PR is breaking things. I put some notes about it here nushell/nushell#11406 (comment). There's a gif on discord in that link.

[amtoine]

I think this PR is breaking things. I put some notes about it here nushell/nushell#11406 (comment). There's a gif on discord in that link.

i suspect #675 is the real responsible for these bugs 🤔

[fdncred]

We probably need to revert this one too.

[sholderbach]

@danielsomerfield @nibon7 sounds like we need to figure out a way to make the new prompt position logic from #675 to work without causing either nushell/nushell#11406 (comment) or nushell/nushell#11399

Maybe we need to rethink how initialize_prompt_position interacts with the new cursor position polling (for solving nushell/nushell#11339) to invalidate.