Re: Negating a subpattern

Another approach would be to use: $sections = preg_split ('#(\[template|wordlet .+\])#ism', $your_buffer, -1, PREG_SPLIT_DELIM_CAPTURE); produces: $sections[0] = empty if any matches, $your_buffer otherwise (I think ...) $sections[1] = "[template a1]" $sections[2] = "This is a test info" (and whitespace before and after) $sections[3] = ''[template a2]" $sections[4] = "This is another info" (and whitespace), etc. Then you can go back and run elements1, 3, etc., through preg_match() to pull out your prior subpatterns. "Acid" <acid@freenet.am> wrote in message news:002701c317c4$c0f55bc0$6400a8c0@acidxp... > Hi > > I have the following regexp > > #\[(template|wordlet) ([^\]]+)\]([^\[]*)#ism > > It gives me sections of my configuration file. As I understand it's > working like find [template or wordlet, then space, then something but ] > one ore more times, then something but [ zero or more times. > > Here is example: > > [template a1] > This is a test info > > [template a2] > This is another info > > [wordlet b1] > Some other stuff goes here. > > It works fine as long as under the section I don't have [ or ]. Now I > want to modify it in order to support square brackets as information > like this. > > [template a1] > This is another info which contains square brackets like this test[1234] > > [template a2... > > I tried to search across PHP docs but I didn't found how to describe the > following expression: > find [template or wordlet, then space, then something but ] one ore more > times, then something but [template or [wordlet zero or more times. At > the same time I want to keep all subpattern boundaries I defined as they > are in first expression. It seems to me that I can't negate a > subpattern... > > Any help will be appreciated. > > > > > > > >