Extend UP032 (f-string) to support auto-fixing multi-line triple-quoted strings

[harupy]

UP032 currently ignores multi-lines strings:

ruff/crates/ruff/src/rules/pyupgrade/rules/f_strings.rs

Lines 328 to 331 in a4e5e32

	// Avoid refactoring multi-line strings.
	if checker.locator.contains_line_break(template.range()) {
	return;
	}

Examples

# Ignored
a = (
  "{}"
  "{}".format(1, 2)
)

# Ignored
a = """
{}
""".format(1)

# Not ignored
a = "{}".format(1)

Is it possible to support auto-fixing the second case?

[harupy]

I'm testing the following change:

diff --git a/crates/ruff/src/rules/pyupgrade/rules/f_strings.rs b/crates/ruff/src/rules/pyupgrade/rules/f_strings.rs
index 494274b3a..8f6b69892 100644
--- a/crates/ruff/src/rules/pyupgrade/rules/f_strings.rs
+++ b/crates/ruff/src/rules/pyupgrade/rules/f_strings.rs
@@ -325,8 +325,12 @@ pub(crate) fn f_strings(
         return;
     }
 
-    // Avoid refactoring multi-line strings.
-    if checker.locator.contains_line_break(template.range()) {
+    // Avoid refactoring implicitly-concatenated strings.
+    // if checker.locator.contains_line_break(template.range()) {
+    //     return;
+    // }
+    let contents = checker.locator.slice(template.range());
+    if is_implicit_concatenation(contents) {
         return;
     }
 
@@ -338,7 +342,13 @@ pub(crate) fn f_strings(
 
     // Avoid refactors that exceed the line length limit.
     let col_offset = template.start() - checker.locator.line_start(template.start());
-    if col_offset.to_usize() + contents.len() > line_length.get() {
+    // if col_offset.to_usize() + contents.len() > line_length.get() {
+    //     return;
+    // }
+    if contents
+        .split('\n')
+        .any(|line| col_offset.to_usize() + line.len() > line_length.get())
+    {
         return;
     }

Two changes:

• Ignore implicitly-concatenated strings, not all multi-line strings.
• Check the length of each line in the new contents, not the length of the entire new contents.

Result

(Built ruff with the change above, and ran it in https://github.com/mlflow/mlflow)

harupy/mlflow@70eb98c

[harupy]

I'm happy to file a PR if this proposal makes sense.

[zanieb]

Thanks for the issue and patch! It looks reasonable to me, I'd be curious to see the ecosystem report from a pull request :)

[harupy]

@zanieb Thanks for the comment, filed #5862 :)