Auto merge of #116548 - nnethercote:assert-long-condition, r=matthewj…

…asper

Improve handling of assertion failures with very long conditions

It's not perfectly clear what the best behaviour is here, but I think this is an improvement.

r? `@matthewjasper`
cc `@m-ou-se`

compiler/rustc_builtin_macros/src/assert.rs

@@ -85,7 +85,7 @@ pub fn expand_assert<'cx>(
                 DUMMY_SP,
                 Symbol::intern(&format!(
                     "assertion failed: {}",
-                    pprust::expr_to_string(&cond_expr).escape_debug()
+                    pprust::expr_to_string(&cond_expr)
                 )),
             )],
         );

tests/ui/macros/assert-long-condition.rs

@@ -0,0 +1,9 @@
+// run-fail
+// check-run-results
+// exec-env:RUST_BACKTRACE=0
+// ignore-emscripten no processes
+// ignore-tidy-linelength
+
+fn main() {
+    assert!(1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10 + 11 + 12 + 13 + 14 + 15 + 16 + 17 + 18 + 19 + 20 + 21 + 22 + 23 + 24 + 25 == 0);
+}

tests/ui/macros/assert-long-condition.run.stderr

@@ -0,0 +1,4 @@
+thread 'main' panicked at $DIR/assert-long-condition.rs:8:5:
+assertion failed: 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10 + 11 + 12 + 13 + 14 + 15 + 16 + 17 + 18
+                                + 19 + 20 + 21 + 22 + 23 + 24 + 25 == 0
+note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace