Regex first match

Al · Nov 30, 2006

Hi,

I have the following code :

string sData=@"{\*\bkmkstart adresse1} FORMTEXT }{\rtlch \af0 \ltrch
\f1\fs22\insrsid16733897\charrsid10567516 {\*\datafield
000000000000000008616472657373653100000000000000000000000000}{\*\formfield{\fftype0\fftypetxt0{\*\ffname
adresse1}}}}}{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516
\u8194\'20\u8194\'20\u8194\'20\u8194\'20\u8194\'20}}}{\rtlch \af0
\ltrch \f1\fs22\insrsid14636597 {\*\bkmkend adresse1}";

string sReg = @"\{\\fldrslt \{\\rtlch.*sid[0-9]+ ";

Regex reg = new Regex(sReg,
RegexOptions.IgnoreCase|RegexOptions.Singleline);

MatchCollection mc = reg.Matches(sData);

string sResult = mc[0].Value;

sResult displays the following :
{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516
\u8194\'20\u8194\'20\u8194\'20\u8194\'20\u8194\'20}}}{\rtlch \af0
\ltrch \f1\fs22\insrsid14636597

In other words, it returns the !!! last !!! 'sid' contained in a
'{\rtlch' itself contained in a '{\fldrslt', followed by a certain
number of numbers and followed by a space.

But what I want is the following result:
{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516

In other words, I want regex to stop at the first 'sid' contained in a
'{\rtlch' itself contained in a '{\fldrslt', followed by a certain
number of numbers and followed by a space.

Can anyone give me the solution ?
Thanks
Al

Kevin Spencer · Nov 30, 2006

If you call the Match method, it will return only the first Match.

--
HTH,

Kevin Spencer
Microsoft MVP
Logostician
http://unclechutney.blogspot.com

Parabola is a mate of plane.

Jon Shemitz · Nov 30, 2006

Al said:
string sData=@"{\*\bkmkstart adresse1} FORMTEXT }{\rtlch \af0 \ltrch
\f1\fs22\insrsid16733897\charrsid10567516 {\*\datafield
000000000000000008616472657373653100000000000000000000000000}{\*\formfield{\fftype0\fftypetxt0{\*\ffname
adresse1}}}}}{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516
\u8194\'20\u8194\'20\u8194\'20\u8194\'20\u8194\'20}}}{\rtlch \af0
\ltrch \f1\fs22\insrsid14636597 {\*\bkmkend adresse1}";

string sReg = @"\{\\fldrslt \{\\rtlch.*sid[0-9]+ ";

Regex reg = new Regex(sReg,
RegexOptions.IgnoreCase|RegexOptions.Singleline);

MatchCollection mc = reg.Matches(sData);

string sResult = mc[0].Value;

sResult displays the following :
{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516
\u8194\'20\u8194\'20\u8194\'20\u8194\'20\u8194\'20}}}{\rtlch \af0
\ltrch \f1\fs22\insrsid14636597

In other words, it returns the !!! last !!! 'sid' contained in a
'{\rtlch' itself contained in a '{\fldrslt', followed by a certain
number of numbers and followed by a space.

But what I want is the following result:
{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516

But the sData doesn't contain that string! Your regex is indeed
finding the ONLY \fldrslt in your sample data.

Jesse Houwing · Nov 30, 2006

* Al wrote, On 30-11-2006 16:33:

Hi,

I have the following code :

string sData=@"{\*\bkmkstart adresse1} FORMTEXT }{\rtlch \af0 \ltrch
\f1\fs22\insrsid16733897\charrsid10567516 {\*\datafield
000000000000000008616472657373653100000000000000000000000000}{\*\formfield{\fftype0\fftypetxt0{\*\ffname
adresse1}}}}}{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516
\u8194\'20\u8194\'20\u8194\'20\u8194\'20\u8194\'20}}}{\rtlch \af0
\ltrch \f1\fs22\insrsid14636597 {\*\bkmkend adresse1}";

string sReg = @"\{\\fldrslt \{\\rtlch.*sid[0-9]+ ";

Regex reg = new Regex(sReg,
RegexOptions.IgnoreCase|RegexOptions.Singleline);

MatchCollection mc = reg.Matches(sData);

string sResult = mc[0].Value;

sResult displays the following :
{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516
\u8194\'20\u8194\'20\u8194\'20\u8194\'20\u8194\'20}}}{\rtlch \af0
\ltrch \f1\fs22\insrsid14636597

In other words, it returns the !!! last !!! 'sid' contained in a
'{\rtlch' itself contained in a '{\fldrslt', followed by a certain
number of numbers and followed by a space.

But what I want is the following result:
{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516

In other words, I want regex to stop at the first 'sid' contained in a
'{\rtlch' itself contained in a '{\fldrslt', followed by a certain
number of numbers and followed by a space.

Can anyone give me the solution ?
Thanks
Al

You're bumping into the fact that regex is greedy by default. This is
due to performance reasons originally and due to the specs that say that
a regex parser should look for the first and longest available answer.
Though it might settle for a shorter one, this is never guaranteed.

To solve your issue you should make your expression reluctant instead of
greedy (.* becomes .*?):

@"\{\\fldrslt \{\\rtlch.*?sid[0-9]+ ";
^^^

which will cause it to match the smallest possible string instead.

Jesse Houwing

Al · Dec 1, 2006

Thanks Jesse,
this works perfectly.
Al
Jesse Houwing a écrit :

* Al wrote, On 30-11-2006 16:33:

Hi,

I have the following code :

string sData=@"{\*\bkmkstart adresse1} FORMTEXT }{\rtlch \af0 \ltrch
\f1\fs22\insrsid16733897\charrsid10567516 {\*\datafield
000000000000000008616472657373653100000000000000000000000000}{\*\formfield{\fftype0\fftypetxt0{\*\ffname
adresse1}}}}}{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516
\u8194\'20\u8194\'20\u8194\'20\u8194\'20\u8194\'20}}}{\rtlch \af0
\ltrch \f1\fs22\insrsid14636597 {\*\bkmkend adresse1}";

string sReg = @"\{\\fldrslt \{\\rtlch.*sid[0-9]+ ";

Regex reg = new Regex(sReg,
RegexOptions.IgnoreCase|RegexOptions.Singleline);

MatchCollection mc = reg.Matches(sData);

string sResult = mc[0].Value;

sResult displays the following :
{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516
\u8194\'20\u8194\'20\u8194\'20\u8194\'20\u8194\'20}}}{\rtlch \af0
\ltrch \f1\fs22\insrsid14636597

In other words, it returns the !!! last !!! 'sid' contained in a
'{\rtlch' itself contained in a '{\fldrslt', followed by a certain
number of numbers and followed by a space.

But what I want is the following result:
{\fldrslt {\rtlch \af0 \ltrch
\f1\fs22\lang1024\langfe1024\noproof\insrsid10567516

In other words, I want regex to stop at the first 'sid' contained in a
'{\rtlch' itself contained in a '{\fldrslt', followed by a certain
number of numbers and followed by a space.

Can anyone give me the solution ?
Thanks
Al

Click to expand...

You're bumping into the fact that regex is greedy by default. This is
due to performance reasons originally and due to the specs that say that
a regex parser should look for the first and longest available answer.
Though it might settle for a shorter one, this is never guaranteed.

To solve your issue you should make your expression reluctant instead of
greedy (.* becomes .*?):

@"\{\\fldrslt \{\\rtlch.*?sid[0-9]+ ";
^^^

which will cause it to match the smallest possible string instead.

Jesse Houwing

Jon Shemitz · Dec 1, 2006

Jon said:
But the sData doesn't contain that string! Your regex is indeed
finding the ONLY \fldrslt in your sample data.

Ooops - misread the question. (Hadn't had my coffee yet.)

Regex first match

Al

Kevin Spencer

Jon Shemitz

Jesse Houwing

Al

Jon Shemitz