reluctant quantifier doesn't work

Thread view:

alex_us01 - 09 Apr 2008 21:01 GMT

Hi all,

I use an regex in Java 6 as this:
"<a .*?some string here</a>"

and expect that .*? will match the smallest char sequence.
However, it matches the longest.

I give this regex to a method I wrote:
    private Matcher getMatcherMLDA(String regex, String input) {
        Pattern pat = Pattern.compile(regex,Pattern.MULTILINE |
Pattern.DOTALL);
        Matcher matcher = pat.matcher(input);
        return matcher;
    }

It looks like the reluctant quantifier doesn't work.
thanks!

Reply to this Message

Roedy Green - 09 Apr 2008 22:37 GMT

On Wed, 9 Apr 2008 13:01:50 -0700 (PDT), alex_us01
<arisalex@gmail.com> wrote, quoted or indirectly quoted someone who
said :

>I use an regex in Java 6 as this:
>"<a .*?some string here</a>"

Try this

"<a( .*)?some string here</a>"

Signature

Roedy Green Canadian Mind Products
The Java Glossary
http://mindprod.com

Reply to this Message

Gordon Beaton - 10 Apr 2008 06:59 GMT

> Try this
>
> "<a( .*)?some string here</a>"

Now it isn't reluctant anymore.

/gordon

--

Reply to this Message

alex_us01 - 30 Apr 2008 23:09 GMT

> > Try this
>
[quoted text clipped - 5 lines]
>
> --

True!
Therefore, this is not a solution.
Thanks, Gordon.

Reply to this Message

Gordon Beaton - 10 Apr 2008 06:58 GMT

> I use an regex in Java 6 as this:
> "<a .*?some string here</a>"
>
> and expect that .*? will match the smallest char sequence.
> However, it matches the longest.

First, you failed to show us how you are observing that.

Second, you need to realize that unless you specified a region, the
regexp is matched against the *entire* input, and the string matched
by the reluctant quantifier will match as much as necessary (but not
more) for the entire input to match.

You likely need to add an additional quantifier to the end of your
regexp to match the remainder of the input you aren't interested in,
thus allowing the reluctant quantifier to match less:

"<a .*?some string here</a>.*"

/gordon

--

Reply to this Message

Hendrik Maryns - 10 Apr 2008 13:34 GMT

Gordon Beaton schreef:

>> I use an regex in Java 6 as this:
>> "<a .*?some string here</a>"
[quoted text clipped - 14 lines]
>
> "<a .*?some string here</a>.*"

Or use find() instead of match().
(but you’ll want to introduce some capturing groups)

H.

Signature

Hendrik Maryns
http://tcl.sfs.uni-tuebingen.de/~hendrik/
==================
http://aouw.org
Ask smart questions, get good answers:
http://www.catb.org/~esr/faqs/smart-questions.html

Reply to this Message

alex_us01 - 30 Apr 2008 23:00 GMT

I do not use find() or match() at all.
Once I get the matcher via getMatcherMLDA(),
I use replaceFirst() or replaceAll() on it.

For example:

Matcher matcher = getMatcherMLDA(X,Y);
matcher.replaceAll("");
OR
matcher.replaceFirst("");

Let me know if any more clarification is needed. Thanks.

Reply to this Message

alex_us01 - 30 Apr 2008 23:12 GMT

> > I use an regex in Java 6 as this:
> > "<a .*?some string here</a>"
[quoted text clipped - 3 lines]
>
> First, you failed to show us how you are observing that.

Not quite. I chose not to.
It is too complicated.
I avoid posting long code unless asked.

> Second, you need to realize that unless you specified a region, the
> regexp is matched against the *entire* input, and the string matched
> by the reluctant quantifier will match as much as necessary (but not
> more) for the entire input to match.

Wrong.
I tried in a small example with the method I provided above and it
worked.
It did NOT need to match the *entire* input string
and it worked for a small pattern and small input.
Perhaps, I shall find an example that it doesn't work and post again.
(However, in the meantime, if you get an insight, let me know.)

Alex

Reply to this Message

alex_us01 - 30 Apr 2008 23:15 GMT

Alex wrote:

> It did NOT need to match the *entire* input string
> and it worked for a small pattern and small input.

Here is an example, that this worked:

---
Matcher matcher = ctr.getMatcherMLDA("<.*?>", "<alex> <alex>");
String out = matcher.replaceFirst("REPLACED");
---

When I ran this code, the out variable contains:
"REPLACED <alex>"

Reply to this Message

reluctant quantifier doesn't work

Free Magazines