Mercurial > vim
changeset 5221:9982ec574beb v7.4a.036
updated for version 7.4a.036
Problem: "\p" in a regexp does not match double-width characters.
(Yukihiro Nakadaira)
Solution: Don't count display cells, use vim_isprintc().
| author | Bram Moolenaar <bram@vim.org> |
|---|---|
| date | Sun, 21 Jul 2013 17:06:00 +0200 |
| parents | 050893d44c33 |
| children | c8559a2d8e5f |
| files | src/regexp.c src/regexp_nfa.c src/testdir/test64.in src/testdir/test64.ok src/testdir/test95.in src/testdir/test95.ok src/version.c |
| diffstat | 7 files changed, 30 insertions(+), 5 deletions(-) [+] |
line wrap: on
line diff
--- a/src/regexp.c +++ b/src/regexp.c @@ -4563,14 +4563,14 @@ regmatch(scan) break; case PRINT: - if (ptr2cells(reginput) != 1) + if (!vim_isprintc(PTR2CHAR(reginput))) status = RA_NOMATCH; else ADVANCE_REGINPUT(); break; case SPRINT: - if (VIM_ISDIGIT(*reginput) || ptr2cells(reginput) != 1) + if (VIM_ISDIGIT(*reginput) || !vim_isprintc(PTR2CHAR(reginput))) status = RA_NOMATCH; else ADVANCE_REGINPUT(); @@ -5944,7 +5944,8 @@ regrepeat(p, maxcount) if (got_int) break; } - else if (ptr2cells(scan) == 1 && (testval || !VIM_ISDIGIT(*scan))) + else if (vim_isprintc(PTR2CHAR(scan)) == 1 + && (testval || !VIM_ISDIGIT(*scan))) { mb_ptr_adv(scan); }
--- a/src/regexp_nfa.c +++ b/src/regexp_nfa.c @@ -5749,12 +5749,12 @@ nfa_regmatch(prog, start, submatch, m) break; case NFA_PRINT: /* \p */ - result = ptr2cells(reginput) == 1; + result = vim_isprintc(PTR2CHAR(reginput)); ADD_STATE_IF_MATCH(t->state); break; case NFA_SPRINT: /* \P */ - result = !VIM_ISDIGIT(curc) && ptr2cells(reginput) == 1; + result = !VIM_ISDIGIT(curc) && vim_isprintc(PTR2CHAR(reginput)); ADD_STATE_IF_MATCH(t->state); break;
--- a/src/testdir/test64.in +++ b/src/testdir/test64.in @@ -228,6 +228,7 @@ STARTTEST :call add(tl, [2, '\v((ab)|c*)+', 'abcccaba', 'abcccab', '', 'ab']) :call add(tl, [2, '\v(a(c*)+b)+', 'acbababaaa', 'acbabab', 'ab', '']) :call add(tl, [2, '\v(a|b*)+', 'aaaa', 'aaaa', '']) +:call add(tl, [2, '\p*', 'aá ', 'aá ']) :" :" Test greedy-ness and lazy-ness :call add(tl, [2, 'a\{-2,7}','aaaaaaaaaaaaa', 'aa'])
--- a/src/testdir/test64.ok +++ b/src/testdir/test64.ok @@ -506,6 +506,9 @@ OK 2 - \v(a(c*)+b)+ OK 0 - \v(a|b*)+ OK 1 - \v(a|b*)+ OK 2 - \v(a|b*)+ +OK 0 - \p* +OK 1 - \p* +OK 2 - \p* OK 0 - a\{-2,7} OK 1 - a\{-2,7} OK 2 - a\{-2,7}
--- a/src/testdir/test95.in +++ b/src/testdir/test95.in @@ -29,6 +29,7 @@ STARTTEST :" this is not a normal "i" but 0xec :call add(tl, [2, '\p\+', 'ìa', 'ìa']) +:call add(tl, [2, '\p*', 'aあ', 'aあ']) :"""" Test recognition of some character classes :call add(tl, [2, '\i\+', '&*¨xx ', 'xx']) @@ -118,6 +119,16 @@ STARTTEST :endfor :unlet t tl e l +:" check that 'ambiwidth' does not change the meaning of \p +:set regexpengine=1 ambiwidth=single +:$put ='eng 1 ambi single: ' . match(\"\u00EC\", '\p') +:set regexpengine=1 ambiwidth=double +:$put ='eng 1 ambi double: ' . match(\"\u00EC\", '\p') +:set regexpengine=2 ambiwidth=single +:$put ='eng 2 ambi single: ' . match(\"\u00EC\", '\p') +:set regexpengine=2 ambiwidth=double +:$put ='eng 2 ambi double: ' . match(\"\u00EC\", '\p') + :/\%#=1^Results/,$wq! test.out ENDTEST
--- a/src/testdir/test95.ok +++ b/src/testdir/test95.ok @@ -17,6 +17,9 @@ OK 2 - [ม[:alpha:][=a=]]\+ OK 0 - \p\+ OK 1 - \p\+ OK 2 - \p\+ +OK 0 - \p* +OK 1 - \p* +OK 2 - \p* OK 0 - \i\+ OK 1 - \i\+ OK 2 - \i\+ @@ -113,3 +116,7 @@ OK 2 - \Zֹ\+ OK 0 - [^[=a=]]\+ OK 1 - [^[=a=]]\+ OK 2 - [^[=a=]]\+ +eng 1 ambi single: 0 +eng 1 ambi double: 0 +eng 2 ambi single: 0 +eng 2 ambi double: 0