cli: Remove duplicated output of RegressOn variant names from the help

This was missed in the recent clap upgrade.

Author: LingMan

src/main.rs

@@ -326,27 +326,27 @@ enum RegressOn {
     /// case for when you want to bisect to see when a bug was fixed.
     Success,
 
-    /// `Ice`: Marks test outcome as `Regressed` if and only if the `rustc`
+    /// Marks test outcome as `Regressed` if and only if the `rustc`
     /// process issues a diagnostic indicating that an internal compiler error
     /// (ICE) occurred. This covers the use case for when you want to bisect to
     /// see when an ICE was introduced pon a codebase that is meant to produce
     /// a clean error.
     Ice,
 
-    /// `NonIce`: Marks test outcome as `Regressed` if and only if the `rustc`
+    /// Marks test outcome as `Regressed` if and only if the `rustc`
     /// process does not issue a diagnostic indicating that an internal
     /// compiler error (ICE) occurred. This covers the use case for when you
     /// want to bisect to see when an ICE was fixed.
     NonIce,
 
-    /// `NonError`: Marks test outcome as `Baseline` if and only if the `rustc`
+    /// Marks test outcome as `Baseline` if and only if the `rustc`
     /// process reports error status and does not issue any diagnostic
     /// indicating that an internal compiler error (ICE) occurred. This is the
     /// use case if the regression is a case where an ill-formed program has
     /// stopped being properly rejected by the compiler.
-    /// (The main difference between this case and `Success` is the handling of
-    /// ICE: `Success` assumes that ICE should be considered baseline;
-    /// `NonError` assumes ICE should be considered a sign of a regression.)
+    /// (The main difference between this case and `success` is the handling of
+    /// ICE: `success` assumes that ICE should be considered baseline;
+    /// `non-error` assumes ICE should be considered a sign of a regression.)
     NonError,
 }
 

tests/cmd/help.stdout

@@ -66,22 +66,22 @@ Options:
             program. This covers the use case for when you want to bisect to see when a bug was
             fixed
           - ice:
-            `Ice`: Marks test outcome as `Regressed` if and only if the `rustc` process issues a
-            diagnostic indicating that an internal compiler error (ICE) occurred. This covers the
-            use case for when you want to bisect to see when an ICE was introduced pon a codebase
-            that is meant to produce a clean error
+            Marks test outcome as `Regressed` if and only if the `rustc` process issues a diagnostic
+            indicating that an internal compiler error (ICE) occurred. This covers the use case for
+            when you want to bisect to see when an ICE was introduced pon a codebase that is meant
+            to produce a clean error
           - non-ice:
-            `NonIce`: Marks test outcome as `Regressed` if and only if the `rustc` process does not
-            issue a diagnostic indicating that an internal compiler error (ICE) occurred. This
-            covers the use case for when you want to bisect to see when an ICE was fixed
+            Marks test outcome as `Regressed` if and only if the `rustc` process does not issue a
+            diagnostic indicating that an internal compiler error (ICE) occurred. This covers the
+            use case for when you want to bisect to see when an ICE was fixed
           - non-error:
-            `NonError`: Marks test outcome as `Baseline` if and only if the `rustc` process reports
-            error status and does not issue any diagnostic indicating that an internal compiler
-            error (ICE) occurred. This is the use case if the regression is a case where an
-            ill-formed program has stopped being properly rejected by the compiler. (The main
-            difference between this case and `Success` is the handling of ICE: `Success` assumes
-            that ICE should be considered baseline; `NonError` assumes ICE should be considered a
-            sign of a regression.)
+            Marks test outcome as `Baseline` if and only if the `rustc` process reports error status
+            and does not issue any diagnostic indicating that an internal compiler error (ICE)
+            occurred. This is the use case if the regression is a case where an ill-formed program
+            has stopped being properly rejected by the compiler. (The main difference between this
+            case and `success` is the handling of ICE: `success` assumes that ICE should be
+            considered baseline; `non-error` assumes ICE should be considered a sign of a
+            regression.)
 
       --script <SCRIPT>
           Script replacement for `cargo build` command