diff options
| author | Father Chrysostomos <sprout@cpan.org> | 2012-09-22 17:54:12 -0700 |
|---|---|---|
| committer | Father Chrysostomos <sprout@cpan.org> | 2012-09-22 18:50:50 -0700 |
| commit | 5255171e6cd0accee6f76ea2980e32b3b5b8e171 (patch) | |
| tree | f198188a949f401983d4493ed514365d2c09213c /pp.c | |
| parent | d8e299374017857bcc055c31b6d4a808fb862100 (diff) | |
| download | perl-5255171e6cd0accee6f76ea2980e32b3b5b8e171.tar.gz | |
[perl #94490] const fold should not trigger special split " "
The easiest way to fix this was to move the special handling out of
the regexp engine. Now a flag is set on the split op itself for
this case. A real regexp is still created, as that is the most
convenient way to propagate locale settings, and it prevents the
need to rework pp_split to handle a null regexp.
This also means that custom regexp plugins no longer need to handle
split specially (which they all do currently).
Diffstat (limited to 'pp.c')
| -rw-r--r-- | pp.c | 7 |
1 files changed, 4 insertions, 3 deletions
@@ -5299,6 +5299,7 @@ PP(pp_split) STRLEN len; const char *s = SvPV_const(sv, len); const bool do_utf8 = DO_UTF8(sv); + const bool skipwhite = PL_op->op_flags & OPf_SPECIAL; const char *strend = s + len; PMOP *pm; REGEXP *rx; @@ -5329,7 +5330,7 @@ PP(pp_split) rx = PM_GETRE(pm); TAINT_IF(get_regex_charset(RX_EXTFLAGS(rx)) == REGEX_LOCALE_CHARSET && - (RX_EXTFLAGS(rx) & (RXf_WHITE | RXf_SKIPWHITE))); + (RX_EXTFLAGS(rx) & RXf_WHITE || skipwhite)); RX_MATCH_UTF8_set(rx, do_utf8); @@ -5369,7 +5370,7 @@ PP(pp_split) } base = SP - PL_stack_base; orig = s; - if (RX_EXTFLAGS(rx) & RXf_SKIPWHITE) { + if (skipwhite) { if (do_utf8) { while (*s == ' ' || is_utf8_space((U8*)s)) s += UTF8SKIP(s); @@ -5391,7 +5392,7 @@ PP(pp_split) if (!limit) limit = maxiters + 2; - if (RX_EXTFLAGS(rx) & RXf_WHITE) { + if (RX_EXTFLAGS(rx) & RXf_WHITE || skipwhite) { while (--limit) { m = s; /* this one uses 'm' and is a negative test */ |