[pcre-dev] [Bug 2325] Backreference erroneously seen as both…

Pàgina inicial
Delete this message
Autor: admin
Data:  
A: pcre-dev
Assumpte: [pcre-dev] [Bug 2325] Backreference erroneously seen as both set and unset in subroutine call
https://bugs.exim.org/show_bug.cgi?id=2325

--- Comment #3 from John Tattarakis <tattarakis@???> ---
Ah, good point, I should've connected the dots seeing as it starts happening
when jumping from 7 to 8!

Nevertheless, one could argue that there is still a discrepancy between what
PCRE sees as set with "(?(8)" and with "\8".

re: ^(a(?(8)\8))()()()()()()(b)(?1)$
str: "aba8"

Matches (in PCRE1), because presumably when pcrepattern.3's documentation says
"and there have not been that many capturing subpatterns", the meaning is
subtly different to "and that subpattern has been set".

>From your 2nd paragraph, the following is likely to exhibit the same result in

PCRE2:

re: ^(a(?(11)\11))()()()()()()()()()(b)(?1)$
str: "aba\t"

And this does in fact match (I just set up a means to test patterns in PCRE2
:D). Should this be accepted as intended behaviour?

--
You are receiving this mail because:
You are on the CC list for the bug.