Improve `String#slice!` performance

Instead of searching twice to extract and to delete, extract and
delete the found position at the first search.

This makes faster nearly twice, for regexps and strings.

|              |compare-ruby|built-ruby|
|:-------------|-----------:|---------:|
|regexp-short  |      2.143M|    3.918M|
|regexp-long   |    105.162k|  205.410k|
|string-short  |      3.789M|    7.964M|
|string-long   |      1.301M|    2.457M|

1 files changed, 11 insertions, 0 deletions

diff --git a/benchmark/string_slice.yml b/benchmark/string_slice.yml
new file mode 100644
index 0000000000..fc2393c5d1
--- /dev/null
+++ b/benchmark/string_slice.yml
@@ -0,0 +1,11 @@
+prelude: |
+  long_string = "x"*1000+"-hår"
+benchmark:
+  regexp-short: |
+    "x-hår".slice!(/-(.)(.)(.)/, 3)
+  regexp-long: |
+    long_string.dup.slice!(/-(.)(.)(.)/, 3)
+  string-short: |
+    "x-hår".slice!("r")
+  string-long: |
+    long_string.dup.slice!("r")