Regex Help Required

[Matthias S.]

hello,

I've got the following regexpression:

"<span .*bbc_underline .*>(.*)</span>"

and the following input string:

"this <span class="bbc_unde rline">is underlined <span
class="bbc_stri kethrough">and striked through</span>text.</span>"

when I do a replace on the expression, I get the following result:

"this text."

I'm quite new to regular expressions. the goal in the expression above is to
yield the following result:

"this [u]is underlined [s]and striked through[/s]test.[u]"

but somehow, the first expression eats up more then it should. btw, I don't
know if or what is nested. I just have a couple of span's with given
classes, that will "turn into" square brackets. can somebody please help? A
ton of thanks in advance.

Matthias

[Ethan Strauss]

Regex is "greedy" by default. It will always find the largest chunk it can.
To make it "lazy" (find the smallest chunk it can), use "*?".
So, I think your Regex would be
"<span .*?bbc_underlin e.*?>(.*?)</span>"
Each of the ".*?" will find the minimum number of characters rather than the
maximum.
Give it a try.
Ethan
"Matthias S." <matthias&AtSig n&emvoid$dot$co mwrote in message
news:e2******** ******@TK2MSFTN GP05.phx.gbl...

hello,

I've got the following regexpression:

"<span .*bbc_underline .*>(.*)</span>"

and the following input string:

"this <span class="bbc_unde rline">is underlined <span
class="bbc_stri kethrough">and striked through</span>text.</span>"

when I do a replace on the expression, I get the following result:

"this text."

I'm quite new to regular expressions. the goal in the expression above is
to
yield the following result:

"this [u]is underlined [s]and striked through[/s]test.[u]"

but somehow, the first expression eats up more then it should. btw, I
don't
know if or what is nested. I just have a couple of span's with given
classes, that will "turn into" square brackets. can somebody please help?
A
ton of thanks in advance.

Matthias

[Matthias S.]

hey ethan,

thanks for your help, but it does not work correctly either, since it
matches the first occurence of </spaninstead of the last.

In the case I did not explain it correctly: What I want to achive is to
match the first occurence of <span SOMETHING "bbc_underline" with the last
possible occurence of </spanand replace everything within with $1.
I don't know how many nestings there are. The span classes I use come from a
predefined list (bbc_underline, bbc_italic, bbc_strikethrou gh and the like)

I would greatly appreceate further help.

Matthias

"Ethan Strauss" <ethan dot strauss at Promega dot comschrieb im
Newsbeitrag news:OD******** ******@TK2MSFTN GP03.phx.gbl...

Regex is "greedy" by default. It will always find the largest chunk it

can.

To make it "lazy" (find the smallest chunk it can), use "*?".
So, I think your Regex would be
"<span .*?bbc_underlin e.*?>(.*?)</span>"
Each of the ".*?" will find the minimum number of characters rather than

the

maximum.
Give it a try.
Ethan
"Matthias S." <matthias&AtSig n&emvoid$dot$co mwrote in message
news:e2******** ******@TK2MSFTN GP05.phx.gbl...

hello,

I've got the following regexpression:

"<span .*bbc_underline .*>(.*)</span>"

and the following input string:

"this <span class="bbc_unde rline">is underlined <span
class="bbc_stri kethrough">and striked through</span>text.</span>"

when I do a replace on the expression, I get the following result:

"this text."

I'm quite new to regular expressions. the goal in the expression above

is

to
yield the following result:

"this [u]is underlined [s]and striked through[/s]test.[u]"

but somehow, the first expression eats up more then it should. btw, I
don't
know if or what is nested. I just have a couple of span's with given
classes, that will "turn into" square brackets. can somebody please

help?

A
ton of thanks in advance.

Matthias

[=?ISO-8859-1?Q?G=F6ran_Andersson?=]

The only way of replacing nested tags is to start from the innermost tag.

Make a pattern that matches your tag, but only if there is not another
such tag inside it. Then you can successfully match the innermost tag
and replace it, then repeat the process until there are no more matches.

Matthias S. wrote:

hey ethan,

thanks for your help, but it does not work correctly either, since it
matches the first occurence of </spaninstead of the last.

In the case I did not explain it correctly: What I want to achive is to
match the first occurence of <span SOMETHING "bbc_underline" with the last
possible occurence of </spanand replace everything within with $1.
I don't know how many nestings there are. The span classes I use come from a
predefined list (bbc_underline, bbc_italic, bbc_strikethrou gh and the like)

I would greatly appreceate further help.

Matthias

"Ethan Strauss" <ethan dot strauss at Promega dot comschrieb im
Newsbeitrag news:OD******** ******@TK2MSFTN GP03.phx.gbl...

>Regex is "greedy" by default. It will always find the largest chunk it

can.

>To make it "lazy" (find the smallest chunk it can), use "*?".
So, I think your Regex would be
"<span .*?bbc_underlin e.*?>(.*?)</span>"
Each of the ".*?" will find the minimum number of characters rather than

the

>maximum.
Give it a try.
Ethan
"Matthias S." <matthias&AtSig n&emvoid$dot$co mwrote in message
news:e2******* *******@TK2MSFT NGP05.phx.gbl.. .

>>hello,

I've got the following regexpression:

"<span .*bbc_underline .*>(.*)</span>"

and the following input string:

"this <span class="bbc_unde rline">is underlined <span
class="bbc_st rikethrough">an d striked through</span>text.</span>"

when I do a replace on the expression, I get the following result:

"this text."

I'm quite new to regular expressions. the goal in the expression above

is

>>to
yield the following result:

"this [u]is underlined [s]and striked through[/s]test.[u]"

but somehow, the first expression eats up more then it should. btw, I
don't
know if or what is nested. I just have a couple of span's with given
classes, that will "turn into" square brackets. can somebody please

help?

>>A
ton of thanks in advance.

Matthias

--
Göran Andersson
_____
http://www.guffa.com

[Oliver Sturm]

Hello Matthias,

>thanks for your help, but it does not work correctly either, since it
matches the first occurence of </spaninstead of the last.

In the case I did not explain it correctly: What I want to achive is to
match the first occurence of <span SOMETHING "bbc_underline" with the last
possible occurence of </spanand replace everything within with $1.

If I understand you correctly, you want balanced matching. That is
actually possible in .NET - here's a post that explains how:
http://blogs.msdn.com/bclteam/archiv...15/396452.aspx

Let me know if that's what you need but you can't make it work - I don't
want to work on it unless it really helps you :-)
Oliver Sturm
--
http://www.sturmnet.org/blog