зеркало из https://github.com/github/ruby.git
* insns.def (toregexp): generate a regexp from strings instead of one
string. * re.c (rb_reg_new_ary): defined for toregexp. it concatenates strings after each string is preprocessed. * compile.c (compile_dstr_fragments): split from compile_dstr. (compile_dstr): call compile_dstr_fragments. (compile_dregx): defined for dynamic regexp. (iseq_compile_each): use compile_dregx for dynamic regexp. [ruby-dev:33400] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@15311 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This commit is contained in:
Родитель
65615986ec
Коммит
d5c8ad5359
15
ChangeLog
15
ChangeLog
|
@ -1,3 +1,18 @@
|
|||
Tue Jan 29 16:59:01 2008 Tanaka Akira <akr@fsij.org>
|
||||
|
||||
* insns.def (toregexp): generate a regexp from strings instead of one
|
||||
string.
|
||||
|
||||
* re.c (rb_reg_new_ary): defined for toregexp. it concatenates
|
||||
strings after each string is preprocessed.
|
||||
|
||||
* compile.c (compile_dstr_fragments): split from compile_dstr.
|
||||
(compile_dstr): call compile_dstr_fragments.
|
||||
(compile_dregx): defined for dynamic regexp.
|
||||
(iseq_compile_each): use compile_dregx for dynamic regexp.
|
||||
|
||||
[ruby-dev:33400]
|
||||
|
||||
Tue Jan 29 16:25:26 2008 NARUSE, Yui <naruse@ruby-lang.org>
|
||||
|
||||
* common.mk, ext/extmkf.rb: always make encdb.h.
|
||||
|
|
26
compile.c
26
compile.c
|
@ -1853,7 +1853,7 @@ iseq_set_sequence_stackcaching(rb_iseq_t *iseq, LINK_ANCHOR *anchor)
|
|||
|
||||
|
||||
static int
|
||||
compile_dstr(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE * node)
|
||||
compile_dstr_fragments(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE * node, int *cntp)
|
||||
{
|
||||
NODE *list = node->nd_next;
|
||||
VALUE lit = node->nd_lit;
|
||||
|
@ -1867,11 +1867,29 @@ compile_dstr(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE * node)
|
|||
cnt++;
|
||||
list = list->nd_next;
|
||||
}
|
||||
*cntp = cnt;
|
||||
|
||||
return COMPILE_OK;
|
||||
}
|
||||
|
||||
static int
|
||||
compile_dstr(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE * node)
|
||||
{
|
||||
int cnt;
|
||||
compile_dstr_fragments(iseq, ret, node, &cnt);
|
||||
ADD_INSN1(ret, nd_line(node), concatstrings, INT2FIX(cnt));
|
||||
return COMPILE_OK;
|
||||
}
|
||||
|
||||
static int
|
||||
compile_dregx(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE * node)
|
||||
{
|
||||
int cnt;
|
||||
compile_dstr_fragments(iseq, ret, node, &cnt);
|
||||
ADD_INSN2(ret, nd_line(node), toregexp, INT2FIX(node->nd_cflag), INT2FIX(cnt));
|
||||
return COMPILE_OK;
|
||||
}
|
||||
|
||||
static int
|
||||
compile_branch_condition(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE * cond,
|
||||
LABEL *then_label, LABEL *else_label)
|
||||
|
@ -4111,8 +4129,7 @@ iseq_compile_each(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE * node, int poped)
|
|||
break;
|
||||
}
|
||||
case NODE_DREGX:{
|
||||
compile_dstr(iseq, ret, node);
|
||||
ADD_INSN1(ret, nd_line(node), toregexp, INT2FIX(node->nd_cflag));
|
||||
compile_dregx(iseq, ret, node);
|
||||
|
||||
if (poped) {
|
||||
ADD_INSN(ret, nd_line(node), pop);
|
||||
|
@ -4128,8 +4145,7 @@ iseq_compile_each(rb_iseq_t *iseq, LINK_ANCHOR *ret, NODE * node, int poped)
|
|||
ADD_INSN2(ret, nd_line(node), onceinlinecache, 0, lend);
|
||||
ADD_INSN(ret, nd_line(node), pop);
|
||||
|
||||
compile_dstr(iseq, ret, node);
|
||||
ADD_INSN1(ret, nd_line(node), toregexp, INT2FIX(node->nd_cflag));
|
||||
compile_dregx(iseq, ret, node);
|
||||
|
||||
ADD_INSN1(ret, nd_line(node), setinlinecache, lstart);
|
||||
ADD_LABEL(ret, lend);
|
||||
|
|
17
insns.def
17
insns.def
|
@ -397,12 +397,19 @@ tostring
|
|||
*/
|
||||
DEFINE_INSN
|
||||
toregexp
|
||||
(rb_num_t opt)
|
||||
(VALUE str)
|
||||
(VALUE val)
|
||||
(rb_num_t opt, rb_num_t cnt)
|
||||
(...)
|
||||
(VALUE val) // inc += 1 - cnt;
|
||||
{
|
||||
volatile VALUE tmp = str; /* for GC */
|
||||
val = rb_reg_new_str(str, opt);
|
||||
VALUE rb_reg_new_ary(VALUE ary, int options);
|
||||
int i;
|
||||
VALUE ary = rb_ary_new2(cnt);
|
||||
RBASIC(ary)->klass = 0;
|
||||
for (i = 0; i < cnt; i++) {
|
||||
rb_ary_store(ary, cnt-i-1, TOPN(i));
|
||||
}
|
||||
POPN(cnt);
|
||||
val = rb_reg_new_ary(ary, opt);
|
||||
}
|
||||
|
||||
/**
|
||||
|
|
77
re.c
77
re.c
|
@ -1949,36 +1949,53 @@ rb_reg_check_preprocess(VALUE str)
|
|||
return Qnil;
|
||||
}
|
||||
|
||||
#if 0
|
||||
static VALUE
|
||||
rb_reg_preprocess_obj(VALUE str,
|
||||
rb_encoding **fixed_enc, onig_errmsg_buffer err)
|
||||
{
|
||||
VALUE buf;
|
||||
char *p, *end;
|
||||
rb_encoding *enc;
|
||||
|
||||
StringValue(str);
|
||||
p = RSTRING_PTR(str);
|
||||
end = p + RSTRING_LEN(str);
|
||||
enc = rb_enc_get(str);
|
||||
|
||||
buf = rb_reg_preprocess(p, end, enc, fixed_enc, err);
|
||||
RB_GC_GUARD(str);
|
||||
return buf;
|
||||
}
|
||||
|
||||
static VALUE
|
||||
rb_reg_preprocess_m(VALUE klass, VALUE obj)
|
||||
rb_reg_preprocess_dregexp(VALUE ary)
|
||||
{
|
||||
rb_encoding *fixed_enc = 0;
|
||||
onig_errmsg_buffer err = "";
|
||||
VALUE str = rb_reg_preprocess_obj(obj, &fixed_enc, err);
|
||||
if (str == Qnil)
|
||||
rb_raise(rb_eArgError, "%s", err);
|
||||
return rb_assoc_new(str, fixed_enc ? Qtrue : Qfalse);
|
||||
int i;
|
||||
VALUE result = 0;
|
||||
int argc = RARRAY_LEN(ary);
|
||||
VALUE *argv = RARRAY_PTR(ary);
|
||||
|
||||
if (argc == 0) {
|
||||
rb_raise(rb_eArgError, "no arguments given");
|
||||
}
|
||||
|
||||
for (i = 0; i < argc; i++) {
|
||||
VALUE str = argv[i];
|
||||
VALUE buf;
|
||||
char *p, *end;
|
||||
rb_encoding *enc;
|
||||
|
||||
StringValue(str);
|
||||
p = RSTRING_PTR(str);
|
||||
end = p + RSTRING_LEN(str);
|
||||
enc = rb_enc_get(str);
|
||||
|
||||
buf = rb_reg_preprocess(p, end, enc, &fixed_enc, err);
|
||||
RB_GC_GUARD(str);
|
||||
|
||||
if (buf == Qnil)
|
||||
rb_raise(rb_eArgError, "%s", err);
|
||||
|
||||
if (i == 0) {
|
||||
/* The encoding of the first fragment is the encoding
|
||||
* given by the regexp option or script encoding. */
|
||||
if (fixed_enc == 0) {
|
||||
rb_enc_copy(buf, str);
|
||||
}
|
||||
}
|
||||
|
||||
if (!result)
|
||||
result = buf;
|
||||
else
|
||||
rb_str_buf_append(result, buf);
|
||||
}
|
||||
|
||||
return result;
|
||||
}
|
||||
#endif
|
||||
|
||||
static int
|
||||
rb_reg_initialize(VALUE obj, const char *s, int len, rb_encoding *enc,
|
||||
|
@ -2084,6 +2101,12 @@ rb_reg_new_str(VALUE s, int options)
|
|||
return re;
|
||||
}
|
||||
|
||||
VALUE
|
||||
rb_reg_new_ary(VALUE ary, int opt)
|
||||
{
|
||||
return rb_reg_new_str(rb_reg_preprocess_dregexp(ary), opt);
|
||||
}
|
||||
|
||||
VALUE
|
||||
rb_enc_reg_new(const char *s, long len, rb_encoding *enc, int options)
|
||||
{
|
||||
|
@ -3042,10 +3065,6 @@ Init_Regexp(void)
|
|||
rb_define_singleton_method(rb_cRegexp, "last_match", rb_reg_s_last_match, -1);
|
||||
rb_define_singleton_method(rb_cRegexp, "try_convert", rb_reg_s_try_convert, 1);
|
||||
|
||||
#if 0
|
||||
rb_define_singleton_method(rb_cRegexp, "preprocess", rb_reg_preprocess_m, 1);
|
||||
#endif
|
||||
|
||||
rb_define_method(rb_cRegexp, "initialize", rb_reg_initialize_m, -1);
|
||||
rb_define_method(rb_cRegexp, "initialize_copy", rb_reg_init_copy, 1);
|
||||
rb_define_method(rb_cRegexp, "hash", rb_reg_hash, 0);
|
||||
|
|
|
@ -442,14 +442,14 @@ class TestM17N < Test::Unit::TestCase
|
|||
assert_raise(ArgumentError) { eval(s("/\#{r}\xc2\xa1/s")) }
|
||||
|
||||
r = /\xc2\xa1/e
|
||||
#assert_raise(ArgumentError) { eval(s("/\xc2\xa1\#{r}/s")) }
|
||||
#assert_raise(ArgumentError) { eval(s("/\#{r}\xc2\xa1/s")) }
|
||||
assert_raise(ArgumentError) { eval(s("/\xc2\xa1\#{r}/s")) }
|
||||
assert_raise(ArgumentError) { eval(s("/\#{r}\xc2\xa1/s")) }
|
||||
|
||||
r = eval(e("/\xc2\xa1/"))
|
||||
#assert_raise(ArgumentError) { /\xc2\xa1#{r}/s }
|
||||
assert_raise(ArgumentError) { /\xc2\xa1#{r}/s }
|
||||
|
||||
r = /\xc2\xa1/e
|
||||
#assert_raise(ArgumentError) { /\xc2\xa1#{r}/s }
|
||||
assert_raise(ArgumentError) { /\xc2\xa1#{r}/s }
|
||||
end
|
||||
|
||||
def test_begin_end_offset
|
||||
|
@ -574,7 +574,7 @@ class TestM17N < Test::Unit::TestCase
|
|||
}
|
||||
assert_regexp_fixed_ascii8bit(/#{}\xc2\xa1/n)
|
||||
assert_regexp_fixed_ascii8bit(/\xc2\xa1#{}/n)
|
||||
#assert_raise(SyntaxError) { s1, s2 = s('\xc2'), s('\xa1'); /#{s1}#{s2}/ }
|
||||
assert_nothing_raised { s1, s2 = a('\xc2'), a('\xa1'); /#{s1}#{s2}/ }
|
||||
end
|
||||
|
||||
def test_dynamic_eucjp_regexp
|
||||
|
@ -584,7 +584,7 @@ class TestM17N < Test::Unit::TestCase
|
|||
assert_raise(SyntaxError) { eval('/\xc2#{}/e') }
|
||||
assert_raise(SyntaxError) { eval('/#{}\xc2/e') }
|
||||
assert_raise(SyntaxError) { eval('/\xc2#{}\xa1/e') }
|
||||
#assert_raise(SyntaxError) { s1, s2 = e('\xc2'), e('\xa1'); /#{s1}#{s2}/ }
|
||||
assert_raise(ArgumentError) { s1, s2 = e('\xc2'), e('\xa1'); /#{s1}#{s2}/ }
|
||||
end
|
||||
|
||||
def test_dynamic_sjis_regexp
|
||||
|
@ -594,7 +594,7 @@ class TestM17N < Test::Unit::TestCase
|
|||
assert_raise(SyntaxError) { eval('/\x81#{}/s') }
|
||||
assert_raise(SyntaxError) { eval('/#{}\x81/s') }
|
||||
assert_raise(SyntaxError) { eval('/\x81#{}\xa1/s') }
|
||||
#assert_raise(SyntaxError) { s1, s2 = s('\x81'), s('\xa1'); /#{s1}#{s2}/ }
|
||||
assert_raise(ArgumentError) { s1, s2 = s('\x81'), s('\xa1'); /#{s1}#{s2}/ }
|
||||
end
|
||||
|
||||
def test_dynamic_utf8_regexp
|
||||
|
@ -604,7 +604,7 @@ class TestM17N < Test::Unit::TestCase
|
|||
assert_raise(SyntaxError) { eval('/\xc2#{}/u') }
|
||||
assert_raise(SyntaxError) { eval('/#{}\xc2/u') }
|
||||
assert_raise(SyntaxError) { eval('/\xc2#{}\xa1/u') }
|
||||
#assert_raise(SyntaxError) { s1, s2 = u('\xc2'), u('\xa1'); /#{s1}#{s2}/ }
|
||||
assert_raise(ArgumentError) { s1, s2 = u('\xc2'), u('\xa1'); /#{s1}#{s2}/ }
|
||||
end
|
||||
|
||||
def test_regexp_unicode
|
||||
|
@ -1080,6 +1080,6 @@ class TestM17N < Test::Unit::TestCase
|
|||
assert_regexp_usascii_literal('/\u1234#{%q"\x80"}/', nil, SyntaxError)
|
||||
assert_regexp_usascii_literal('/\u1234#{"\x80"}/', nil, SyntaxError)
|
||||
assert_regexp_usascii_literal('/\u1234\x80/', nil, SyntaxError)
|
||||
assert_regexp_usascii_literal('/\u1234#{}\x80/', nil, RegexpError)
|
||||
assert_regexp_usascii_literal('/\u1234#{}\x80/', nil, ArgumentError)
|
||||
end
|
||||
end
|
||||
|
|
Загрузка…
Ссылка в новой задаче