Feature #2571 » 0001-string.c-rb_str_hamming_distance-Init_String.patch
| ChangeLog | ||
|---|---|---|
|
Thu Jan 7 16:32:00 2010 Kenta Murata <mrkn@mrkn.jp>
|
||
|
* string.c (rb_str_hamming_distance, Init_String),
|
||
|
test/ruby/test_string.c (test_hamming_distance): Added support for
|
||
|
calculating hamming distance between two strings.
|
||
|
Thu Jan 7 07:56:09 2010 Ryan Davis <ryand-ruby@zenspider.com>
|
||
|
* lib/minitest/*.rb: Imported minitest 1.5.0 r5596.
|
||
| string.c | ||
|---|---|---|
|
*var = val;
|
||
|
}
|
||
|
/*
|
||
|
*/
|
||
|
VALUE
|
||
|
rb_str_hamming_distance(VALUE str1, VALUE str2)
|
||
|
{
|
||
|
long d = 0;
|
||
|
rb_encoding *enc;
|
||
|
char *p1, *p1end, *p2, *p2end;
|
||
|
if (TYPE(str2) != T_STRING) return Qnil;
|
||
|
enc = rb_enc_compatible(str1, str2);
|
||
|
if (!enc) return Qnil;
|
||
|
if (str_strlen(str1, enc) != str_strlen(str2, enc)) return Qnil;
|
||
|
p1 = RSTRING_PTR(str1); p1end = RSTRING_END(str1);
|
||
|
p2 = RSTRING_PTR(str2); p2end = RSTRING_END(str2);
|
||
|
if (single_byte_optimizable(str1) && single_byte_optimizable(str2)) {
|
||
|
while (p1 < p1end) {
|
||
|
if (*p1 != *p2) ++d;
|
||
|
++p1;
|
||
|
++p2;
|
||
|
}
|
||
|
}
|
||
|
else {
|
||
|
while (p1 < p1end && p2 < p2end) {
|
||
|
int l1, c1 = rb_enc_ascget(p1, p1end, &l1, enc);
|
||
|
int l2, c2 = rb_enc_ascget(p2, p2end, &l2, enc);
|
||
|
if (0 <= c1 && 0 <= c2) {
|
||
|
if (c1 != c2) ++d;
|
||
|
}
|
||
|
else {
|
||
|
int r, len;
|
||
|
l1 = rb_enc_mbclen(p1, p1end, enc);
|
||
|
l2 = rb_enc_mbclen(p2, p2end, enc);
|
||
|
len = l1 < l2 ? l1 : l2;
|
||
|
r = memcmp(p1, p2, len);
|
||
|
if (r != 0 || l1 != l2) ++d;
|
||
|
}
|
||
|
p1 += l1;
|
||
|
p2 += l2;
|
||
|
}
|
||
|
}
|
||
|
return LONG2NUM(d);
|
||
|
}
|
||
|
/*
|
||
|
* call-seq:
|
||
| ... | ... | |
|
rb_define_method(rb_cString, "partition", rb_str_partition, 1);
|
||
|
rb_define_method(rb_cString, "rpartition", rb_str_rpartition, 1);
|
||
|
rb_define_method(rb_cString, "hamming_distance", rb_str_hamming_distance, 1);
|
||
|
rb_define_method(rb_cString, "^", rb_str_hamming_distance, 1);
|
||
|
rb_define_method(rb_cString, "encoding", rb_obj_encoding, 0); /* in encoding.c */
|
||
|
rb_define_method(rb_cString, "force_encoding", rb_str_force_encoding, 1);
|
||
|
rb_define_method(rb_cString, "valid_encoding?", rb_str_valid_encoding_p, 0);
|
||
| test/ruby/test_string.rb | ||
|---|---|---|
|
assert_equal(s, k, '[ruby-dev:39068]')
|
||
|
assert_equal(Encoding::UTF_8, k.encoding, '[ruby-dev:39068]')
|
||
|
end
|
||
|
def test_hamming_distance
|
||
|
assert_nil("a" ^ "abc")
|
||
|
assert_equal(0, "" ^ "")
|
||
|
assert_equal(2, "abc" ^ "dbf")
|
||
|
assert_equal(1, "aXb" ^ "a\u{ff38}b")
|
||
|
assert_equal(1, "\u3042\u3044\u3046" ^ "\u3042\u3044\u3048")
|
||
|
assert_equal(3, "\u3042a\u3044\u3046" ^ "\u3042\u3044b\u3048")
|
||
|
assert_equal(1, "aXb".force_encoding("EUC-JP") ^ "aYb".force_encoding("Shift_JIS"))
|
||
|
assert_nil("\u3042\u3044\u3046".force_encoding("EUC-JP") ^ "\u3042\u3044\u3046".force_encoding("Shift_JIS"))
|
||
|
end
|
||
|
end
|
||