Shortest match regex - any help

Horlock

Hi hope this a suitable place to put this.

I'm trying to extract an expression in visual basic .net (but any regex help greatly appriciated).

my starting string is something like:

hello <personname> john <or/> fred <or/> terry <or/> bill <or/> jane </personname>

I want grab the person names!

As a start I would be happy with how to grab john.

My guess was

RegularExpressions.Regex.Replace(input, ".*<personname>(.*)<or/>.*", "$1").ToString

however this grabs

john <or/> fred <or/> terry <or/> bill <or/> jane

Unfortunately I can't just do

RegularExpressions.Regex.Replace(input, ".*>(^<*)<.*", "$1").ToString

because there is other markup in the string which is needed (ie the original string is actually slightly more like

hello <personname> john <or/> <dontlike> fred </dontlike> <or/> terry <or/> bill <or/> jane </personname>

Any help greatfully recieved (assuming it's possible - and I've posted in the correct place).

Ta

irnbru_2

Horlock wrote:

hello <personname> john <or/> fred <or/> terry <or/> bill <or/> jane </personname>

Is that actually what you're given? It's not part of some properly tag-balanced document?

Horlock

irnbru wrote:

Is that actually what you're given? It's not part of some properly tag-balanced document?

It is part of a tag-balance document - if that helps

Chippy_Minton

Horlock wrote:

It is part of a tag-balance document - if that helps

In that case it might be easier to parse it as an XML document - see http://www.codeproject.com/vb/net/parsefilecode.asp.

Horlock

Chippy_Minton wrote:

In that case it might be easier to parse it as an XML document - see http://www.codeproject.com/vb/net/parsefilecode.asp.

Wow that would probably be marvelous if I hadn't gone so far in the code already - still it might be worth my while to backtrack and use it - but there are quite a few classes which might need to be scrapped - though it may end more elegantly - I'll have to think about it.

Thanks loads though - afraid I can only click thanks once!

Horlock

Just found an easy work around - I think - but I'm not going to implement it tonight.

If I start from the other end.

Ie first I find Jane
then I find bill etc etc

I'm fairly confident that that will work