On 09:40 Wed 05 Jun , Ingo Karkat wrote: > On 05-Jun-2013 09:10 +0200, Christian Brabandt wrote: > > > On Wed, June 5, 2013 08:56, Ingo Karkat wrote: > >> On 05-Jun-2013 08:03 +0200, Christian Brabandt wrote: > >> > >>> [...] the old engine has an long standing bug with \%V ;( > >>> > >>> e.g: > >>> > >>> vim -u NONE -N > >>> ifoobar<esc> > >>> 0ve<esc> > >>> /\%Vfoobar\%V > >>> > >>> I should have probably provided a fix long ago. > >> > >> Where is the bug, that there's no match?! That's intended, as the last > >> \%V matches zero-width _after_ the end of "foobar", but the visual > >> selection ends on the "r". You'd have to use /\%Vfooba\%Vr here. (The > >> example at :help \%V is wrong about this corner case, too.) > > > > I am not sure, it is intended. I certainly wouldn't expect this (and > > the doc is wrong in this regard, as you said). Since > > \%V is zero-width, I would expect the \%V to still match the end of the > > visual selection. > > Zero-width means: Matches at the current position, but does not consume > the character; i.e. the next match is made at the same position. > What is needed at the final atom is: Matches at the previous position > (that has already been consumed). I don't know how difficult that is to > implement in the regular expression engine(s), but I think such a new > \%<V atom would be helpful. > > In my opinion it is hard to change the existing \%V in a way to fit both > uses: It just matches anywhere inside a selection, and has no notion of > start or end of selection. Therefore, it's presumably difficult to > change its behavior at the end of the selection without introducing > off-by-one errors elsewhere. I'm already struggling to come up with a > good and precise documentation for the new behavior; the current one is > at least simple to grasp (when you understand zero-width matches). And > it would be a compatibility-breaking change. > > >> The general case is this ugly beast: /\%Vfoobar\%(\%V\|\%(\%V.\)\@<=\). > > > > Yes, very ugly and not easily understandable. > > > >> Since restricting the match to inside the entire selection is such a > >> common use case (see the related vis.vim plugin), I think it would be > >> worthwhile to have a special atom (e.g. \%<V) that matches if the > >> _preceding character_ is inside the selection, allowing this much nicer > >> pattern: /\%Vfoobar\%<V > > > > I can understand why this happens but nevertheless it is an annoying > > corner-case to consider when using the \%V atom, so I'd like to have > > this issue resolved. > > Yes, at least the help should warn about this corner case. I ran into it > (and came up with the complex correct regexp) in some of my plugins, but > I think \%V is one of the more obscure atoms few people use. Probably > more users rely on the vis.vim plugin, which solves the problem in a > completely different way. > > -- regards, ingo > > -- > -- > You received this message from the "vim_dev" maillist. > Do not top-post! Type your reply below the text you are replying to. > For more information, visit http://www.vim.org/maillist.php > > --- > You received this message because you are subscribed to the Google Groups > "vim_dev" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/groups/opt_out. >
And what about this pattern: /\%Vfooba\%Vr Thanks I did not know this pattern and it is very handy. I always felt that the search commands /, ? should have a range, so that one can limit the search to part of a text. What about having two commands :<range>/ and :<range>? for searches limited to the <range> of lines. Best regards, Marcin -- -- You received this message from the "vim_dev" maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit http://www.vim.org/maillist.php --- You received this message because you are subscribed to the Google Groups "vim_dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/groups/opt_out.
