rb_str_resize: Only clear coderange on truncation

byroot · byroot · commit b0b9f7201aca · 2022-08-18T10:09:08.000+02:00
If we are expanding the string or only stripping extra capacity
then coderange won't change, so clearing it is wasteful.
diff --git a/file.c b/file.c
@@ -4090,6 +4090,9 @@ static VALUE
 str_shrink(VALUE str)
 {
     rb_str_resize(str, RSTRING_LEN(str));
+    // expand_path on Windows can sometimes mutate the string
+    // without clearing its coderange
+    ENC_CODERANGE_CLEAR(str);
     return str;
 }
 
diff --git a/sprintf.c b/sprintf.c
@@ -937,6 +937,8 @@ rb_str_format(int argc, const VALUE *argv, VALUE fmt)
         if (RTEST(ruby_verbose)) rb_warn("%s", mesg);
     }
     rb_str_resize(result, blen);
+    // rb_str_format mutates the string without updating coderange
+    ENC_CODERANGE_CLEAR(result);
 
     return result;
 }
@@ -1163,6 +1165,8 @@ ruby_vsprintf0(VALUE result, char *p, const char *fmt, va_list ap)
     buffer.value = 0;
     BSD_vfprintf(&f, fmt, ap);
     RBASIC_SET_CLASS_RAW(result, klass);
+    // vfprintf mutates the string without updating coderange
+    ENC_CODERANGE_CLEAR(result);
     rb_str_resize(result, (char *)f._p - RSTRING_PTR(result));
 #undef f
 }
@@ -1183,7 +1187,6 @@ rb_enc_vsprintf(rb_encoding *enc, const char *fmt, va_list ap)
         rb_enc_associate(result, enc);
     }
     ruby_vsprintf0(result, RSTRING_PTR(result), fmt, ap);
-
     return result;
 }
 
diff --git a/string.c b/string.c
@@ -2498,7 +2498,6 @@ rb_str_modify_expand(VALUE str, long expand)
     else if (expand > 0) {
         RESIZE_CAPA_TERM(str, len + expand, termlen);
     }
-    ENC_CODERANGE_CLEAR(str);
 }
 
 /* As rb_str_modify(), but don't clear coderange */
@@ -3073,16 +3072,16 @@ rb_str_set_len(VALUE str, long len)
 VALUE
 rb_str_resize(VALUE str, long len)
 {
-    long slen;
-    int independent;
-
     if (len < 0) {
         rb_raise(rb_eArgError, "negative string size (or size too big)");
     }
 
-    independent = str_independent(str);
-    ENC_CODERANGE_CLEAR(str);
-    slen = RSTRING_LEN(str);
+    int independent = str_independent(str);
+    long slen = RSTRING_LEN(str);
+
+    if (slen > len && ENC_CODERANGE(str) != ENC_CODERANGE_7BIT) {
+        ENC_CODERANGE_CLEAR(str);
+    }
 
     {
         long capa;
diff --git a/test/ruby/test_sprintf.rb b/test/ruby/test_sprintf.rb
@@ -507,6 +507,16 @@ def test_named_typed_enc
     end
   end
 
+  def test_coderange
+    format_str = "wrong constant name %s"
+    interpolated_str = "\u3042"
+    assert_predicate format_str, :ascii_only?
+    refute_predicate interpolated_str, :ascii_only?
+
+    str = format_str % interpolated_str
+    refute_predicate str, :ascii_only?
+  end
+
   def test_named_default
     h = Hash.new('world')
     assert_equal("hello world", "hello %{location}" % h)

Original file line number	Diff line number	Diff line change
`@@ -4090,6 +4090,9 @@ static VALUE`
`4090`	`4090`	`str_shrink(VALUE str)`
`4091`	`4091`	`{`
`4092`	`4092`	`rb_str_resize(str, RSTRING_LEN(str));`
	`4093`	`+ // expand_path on Windows can sometimes mutate the string`
	`4094`	`+ // without clearing its coderange`
	`4095`	`+ ENC_CODERANGE_CLEAR(str);`
`4093`	`4096`	`return str;`
`4094`	`4097`	`}`
`4095`	`4098`
Original file line number	Diff line number	Diff line change
`@@ -937,6 +937,8 @@ rb_str_format(int argc, const VALUE *argv, VALUE fmt)`
`937`	`937`	`if (RTEST(ruby_verbose)) rb_warn("%s", mesg);`
`938`	`938`	`}`
`939`	`939`	`rb_str_resize(result, blen);`
	`940`	`+ // rb_str_format mutates the string without updating coderange`
	`941`	`+ ENC_CODERANGE_CLEAR(result);`
`940`	`942`
`941`	`943`	`return result;`
`942`	`944`	`}`
`@@ -1163,6 +1165,8 @@ ruby_vsprintf0(VALUE result, char p, const char fmt, va_list ap)`
`1163`	`1165`	`buffer.value = 0;`
`1164`	`1166`	`BSD_vfprintf(&f, fmt, ap);`
`1165`	`1167`	`RBASIC_SET_CLASS_RAW(result, klass);`
	`1168`	`+ // vfprintf mutates the string without updating coderange`
	`1169`	`+ ENC_CODERANGE_CLEAR(result);`
`1166`	`1170`	`rb_str_resize(result, (char *)f._p - RSTRING_PTR(result));`
`1167`	`1171`	`#undef f`
`1168`	`1172`	`}`
`@@ -1183,7 +1187,6 @@ rb_enc_vsprintf(rb_encoding enc, const char fmt, va_list ap)`
`1183`	`1187`	`rb_enc_associate(result, enc);`
`1184`	`1188`	`}`
`1185`	`1189`	`ruby_vsprintf0(result, RSTRING_PTR(result), fmt, ap);`
`1186`		`-`
`1187`	`1190`	`return result;`
`1188`	`1191`	`}`
`1189`	`1192`
Original file line number	Diff line number	Diff line change
`@@ -2498,7 +2498,6 @@ rb_str_modify_expand(VALUE str, long expand)`
`2498`	`2498`	`else if (expand > 0) {`
`2499`	`2499`	`RESIZE_CAPA_TERM(str, len + expand, termlen);`
`2500`	`2500`	`}`
`2501`		`- ENC_CODERANGE_CLEAR(str);`
`2502`	`2501`	`}`
`2503`	`2502`
`2504`	`2503`	`/* As rb_str_modify(), but don't clear coderange */`
`@@ -3073,16 +3072,16 @@ rb_str_set_len(VALUE str, long len)`
`3073`	`3072`	`VALUE`
`3074`	`3073`	`rb_str_resize(VALUE str, long len)`
`3075`	`3074`	`{`
`3076`		`- long slen;`
`3077`		`- int independent;`
`3078`		`-`
`3079`	`3075`	`if (len < 0) {`
`3080`	`3076`	`rb_raise(rb_eArgError, "negative string size (or size too big)");`
`3081`	`3077`	`}`
`3082`	`3078`
`3083`		`- independent = str_independent(str);`
`3084`		`- ENC_CODERANGE_CLEAR(str);`
`3085`		`- slen = RSTRING_LEN(str);`
	`3079`	`+ int independent = str_independent(str);`
	`3080`	`+ long slen = RSTRING_LEN(str);`
	`3081`	`+`
	`3082`	`+ if (slen > len && ENC_CODERANGE(str) != ENC_CODERANGE_7BIT) {`
	`3083`	`+ ENC_CODERANGE_CLEAR(str);`
	`3084`	`+ }`
`3086`	`3085`
`3087`	`3086`	`{`
`3088`	`3087`	`long capa;`