Scientific notation formatter misses f-strings

[claydugo]

Sorry for not listing the actual rule.
I tried all the flags in ruff format --help but never got an output that explains which rule is making the change.

I can update the title if pointed to which is doing it. Also is there a command I missed that outputs this? --show-fixes seems to be only a linter flag.

Reproduce:

int_variable = 3

fstring1 = f"This string contains scientific notation. {int_variable * 1E-3}"
fstring2 = f"string containing scientific notation. {int_variable * 1E3:.2f}"

variable_notation1 = 1E3
variable_notation2 = (int_variable * 1E-3)

Output of ruff format --diff

--- example.py
+++ example.py
@@ -3,5 +3,5 @@
 fstring1 = f"This string contains scientific notation. {int_variable * 1E-3}"
 fstring2 = f"string containing scientific notation. {int_variable * 1E3:.2f}"

-variable_notation1 = 1E3
-variable_notation2 = (int_variable * 1E-3)
+variable_notation1 = 1e3
+variable_notation2 = int_variable * 1e-3

1 file would be reformatted

[MichaReiser]

Hy @claydugo

There's no specific rule that performs this change. Instead, the formatter normalizes the scientific notation to consistently use the lowercase `e. So, there's no rule that you need to enable or disable.

However, your observation is correct. The formatter doesn't format expressions inside f-strings yet (similar to black). @dhruvmanila started working on formatting expressions inside of f-strings, which should give you the desired behavior (including splitting expressions to make the f-string fit if possible).

[dhruvmanila]

Hey, thanks for the issue and thanks Micha for providing the context. I'll close this in favor of #7594 which tracks f-string formatting.