From f9126265bc03faa150614fa80fcd98e7056fb385 Mon Sep 17 00:00:00 2001 From: Karl Williamson Date: Fri, 13 May 2011 08:35:23 -0600 Subject: [PATCH] Assertion fails in multi-char regex match In '"s\N{U+DF}" =~ /\x{00DF}/i, the LHS folds to 'sss', the RHS to 'ss'. The bug occurs when the RHS tries to match the first two es's, but that splits the LHS \xDF character, which Perl doesn't know how to handle, and the assertion got triggered. (This is similar to [perl #72998].) The solution adopted here is to disallow a partial character match, as #72998 did as well. --- regexec.c | 10 ++++++---- t/re/pat_advanced.t | 4 ++++ 2 files changed, 10 insertions(+), 4 deletions(-) diff --git a/regexec.c b/regexec.c index 391fc16..fd90ad7 100644 --- a/regexec.c +++ b/regexec.c @@ -6726,9 +6726,12 @@ S_reginclass(pTHX_ const regexp * const prog, register const regnode * const n, STRLEN len; const char * const s = SvPV_const(sv, len); - if (len <= total_foldlen && memEQ(s, - (char*)folded, - len)) + if (len <= total_foldlen + && memEQ(s, (char*)folded, len) + + /* If 0, means matched a partial char. See + * [perl #90536] */ + && map_fold_len_back[len]) { /* Advance the target string ptr to account for @@ -6737,7 +6740,6 @@ S_reginclass(pTHX_ const regexp * const prog, register const regnode * const n, * length. */ if (lenp) { *lenp = map_fold_len_back[len]; - assert(*lenp != 0); /* Otherwise will loop */ } match = TRUE; break; diff --git a/t/re/pat_advanced.t b/t/re/pat_advanced.t index 6d7624d..2a510d1 100644 --- a/t/re/pat_advanced.t +++ b/t/re/pat_advanced.t @@ -2108,6 +2108,10 @@ EOP like("\x{00DF}", qr/[\x{1E9E}_]*/i, "\"\\x{00DF}\" =~ /[\\x{1E9E}_]*/i was looping"); } + { # Bug #90536, caused failed assertion + unlike("s\N{U+DF}", qr/^\x{00DF}/i, "\"s\\N{U+DF}\", qr/^\\x{00DF}/i"); + } + # !!! NOTE that tests that aren't at all likely to crash perl should go # a ways above, above these last ones. -- 2.7.4