From e1eb54b99d592c460fea4b0f897f5e8e61c96c07 Mon Sep 17 00:00:00 2001 From: nobu Date: Wed, 17 Apr 2019 05:34:46 +0000 Subject: string.c: improve splitting into chars * string.c (rb_str_split_m): improve splitting into chars by an empty string, without a regexp. Comparison: to_chars-1 built-ruby: 1273527.6 i/s compare-ruby: 189423.3 i/s - 6.72x slower to_chars-10 built-ruby: 120993.5 i/s compare-ruby: 37075.8 i/s - 3.26x slower to_chars-100 built-ruby: 15646.4 i/s compare-ruby: 4012.1 i/s - 3.90x slower to_chars-1000 built-ruby: 1295.1 i/s compare-ruby: 408.5 i/s - 3.17x slower git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67582 b2dd03c8-39d4-4d8f-98ff-823fe69b080e --- benchmark/string_split.yml | 7 +++++++ string.c | 30 ++++++++++++++++++++---------- 2 files changed, 27 insertions(+), 10 deletions(-) create mode 100644 benchmark/string_split.yml diff --git a/benchmark/string_split.yml b/benchmark/string_split.yml new file mode 100644 index 0000000000..84ffe8f6a7 --- /dev/null +++ b/benchmark/string_split.yml @@ -0,0 +1,7 @@ +prelude: | + str0 = [*0..9].join("") +benchmark: + to_chars-1: str0.split('') + to_chars-10: (str0 * 10).split('') + to_chars-100: (str0 * 100).split('') + to_chars-1000: (str0 * 1000).split('') diff --git a/string.c b/string.c index 3bf6f42e2a..153fcc59b7 100644 --- a/string.c +++ b/string.c @@ -7759,7 +7759,7 @@ rb_str_split_m(int argc, VALUE *argv, VALUE str) rb_encoding *enc; VALUE spat; VALUE limit; - enum {awk, string, regexp} split_type; + enum {awk, string, regexp, chars} split_type; long beg, end, i = 0, empty_count = -1; int lim = 0; VALUE result, tmp; @@ -7801,8 +7801,7 @@ rb_str_split_m(int argc, VALUE *argv, VALUE str) split_type = string; if (RSTRING_LEN(spat) == 0) { /* Special case - split into chars */ - spat = rb_reg_regcomp(spat); - split_type = regexp; + split_type = chars; } else if (rb_enc_asciicompat(enc2) == 1) { if (RSTRING_LEN(spat) == 1 && RSTRING_PTR(spat)[0] == ' ') { @@ -7823,9 +7822,9 @@ rb_str_split_m(int argc, VALUE *argv, VALUE str) if (result) result = rb_ary_new(); beg = 0; + char *ptr = RSTRING_PTR(str); + char *eptr = RSTRING_END(str); if (split_type == awk) { - char *ptr = RSTRING_PTR(str); - char *eptr = RSTRING_END(str); char *bptr = ptr; int skip = 1; unsigned int c; @@ -7884,10 +7883,8 @@ rb_str_split_m(int argc, VALUE *argv, VALUE str) } } else if (split_type == string) { - char *ptr = RSTRING_PTR(str); char *str_start = ptr; char *substr_start = ptr; - char *eptr = RSTRING_END(str); char *sptr = RSTRING_PTR(spat); long slen = RSTRING_LEN(spat); @@ -7908,8 +7905,21 @@ rb_str_split_m(int argc, VALUE *argv, VALUE str) } beg = ptr - str_start; } + else if (split_type == chars) { + char *str_start = ptr; + int n; + + mustnot_broken(str); + enc = rb_enc_get(str); + while (ptr < eptr && + (n = rb_enc_precise_mbclen(ptr, eptr, enc)) > 0) { + SPLIT_STR(ptr - str_start, n); + ptr += n; + if (!NIL_P(limit) && lim <= ++i) break; + } + beg = ptr - str_start; + } else { - char *ptr = RSTRING_PTR(str); long len = RSTRING_LEN(str); long start = beg; long idx; @@ -7924,14 +7934,14 @@ rb_str_split_m(int argc, VALUE *argv, VALUE str) break; } else if (last_null == 1) { - SPLIT_STR(beg, rb_enc_fast_mbclen(ptr+beg, ptr+len, enc)); + SPLIT_STR(beg, rb_enc_fast_mbclen(ptr+beg, eptr, enc)); beg = start; } else { if (start == len) start++; else - start += rb_enc_fast_mbclen(ptr+start,ptr+len,enc); + start += rb_enc_fast_mbclen(ptr+start,eptr,enc); last_null = 1; continue; } -- cgit v1.2.3