How to make Regex.Replace faster?

[Morgan Cheng]

In my case, I have to remove any line containing "0.000000" from input
string.
In below case, it takes about 100 ms for 2k size input string.
Regex.Replace(i nputString, ".*0\\.000000.* \n", "");
I want to optimize it, so i make a static member instance instead of
using static func of Regex;
static Regex filter= new Regex(".*0\\.00 0000.*\n",
RegexOptions.Co mpiled);
And use the static instance to replace the string.
filter.Replace( inputString, "");
Now, it takes about 200 ms. The platform is .net 1.1.
I don't understand why the non-static member function is slower thant
static one. The static function has to compile the regular expression
each time, right?

Is there any better solution to remove substring of specific pattern?
Perhaps I have to write my code to do it.

Thanks,
-Morgan

[Cor Ligthert [MVP]]

Morgan,

To make regex faster try to avoid it by using a replace or an other method.

Regex is approx. 100 times slower in simple replacements than by instance
"replace".
(This will be gained again in complex situation where you would have to do a
lot of replacements of course)

I hope this helps,

Cor
"Morgan Cheng" <mo************ @gmail.comschre ef in bericht
news:11******** ************@k7 0g2000cwa.googl egroups.com...

In my case, I have to remove any line containing "0.000000" from input
string.
In below case, it takes about 100 ms for 2k size input string.
Regex.Replace(i nputString, ".*0\\.000000.* \n", "");
I want to optimize it, so i make a static member instance instead of
using static func of Regex;
static Regex filter= new Regex(".*0\\.00 0000.*\n",
RegexOptions.Co mpiled);
And use the static instance to replace the string.
filter.Replace( inputString, "");
Now, it takes about 200 ms. The platform is .net 1.1.
I don't understand why the non-static member function is slower thant
static one. The static function has to compile the regular expression
each time, right?

Is there any better solution to remove substring of specific pattern?
Perhaps I have to write my code to do it.

Thanks,
-Morgan

[Jon Skeet [C# MVP]]

Morgan Cheng <mo************ @gmail.comwrote :

In my case, I have to remove any line containing "0.000000" from input
string.
In below case, it takes about 100 ms for 2k size input string.
Regex.Replace(i nputString, ".*0\\.000000.* \n", "");
I want to optimize it, so i make a static member instance instead of
using static func of Regex;
static Regex filter= new Regex(".*0\\.00 0000.*\n",
RegexOptions.Co mpiled);
And use the static instance to replace the string.
filter.Replace( inputString, "");
Now, it takes about 200 ms. The platform is .net 1.1.
I don't understand why the non-static member function is slower thant
static one. The static function has to compile the regular expression
each time, right?

Is there any better solution to remove substring of specific pattern?
Perhaps I have to write my code to do it.

I would try using a StringReader, read the data line by line, and
append any line which doesn't contain 0.000000 - simple string
operations. If you could provide a test program you've used to gather
your performance data, we could experiment a bit.

--
Jon Skeet - <sk***@pobox.co m>
http://www.pobox.com/~skeet Blog: http://www.msmvps.com/jon.skeet
If replying to the group, please do not mail me too

[Morgan Cheng]

Cor Ligthert [MVP] 写�：

Morgan,

To make regex faster try to avoid it by using a replace or an other method.

Do you mean it is not recommeded to trim string with Regex?
So, the best way for is to code manually handling the string?

Regex is approx. 100 times slower in simple replacements than by instance
"replace".
(This will be gained again in complex situation where you would have to do a
lot of replacements of course)

I hope this helps,

Cor
"Morgan Cheng" <mo************ @gmail.comschre ef in bericht
news:11******** ************@k7 0g2000cwa.googl egroups.com...

In my case, I have to remove any line containing "0.000000" from input
string.
In below case, it takes about 100 ms for 2k size input string.
Regex.Replace(i nputString, ".*0\\.000000.* \n", "");
I want to optimize it, so i make a static member instance instead of
using static func of Regex;
static Regex filter= new Regex(".*0\\.00 0000.*\n",
RegexOptions.Co mpiled);
And use the static instance to replace the string.
filter.Replace( inputString, "");
Now, it takes about 200 ms. The platform is .net 1.1.
I don't understand why the non-static member function is slower thant
static one. The static function has to compile the regular expression
each time, right?

Is there any better solution to remove substring of specific pattern?
Perhaps I have to write my code to do it.

Thanks,
-Morgan

[Cor Ligthert [MVP]]

Morgan,

>Do you mean it is not recommeded to trim string with Regex?
So, the best way for is to code manually handling the string?

If you can avoid Regex try that, but not in the case that you have to done a
lot of things for it, or make your code not readable for others anymore.
Than use Regex.

(And than you probably on the place that it performs in at least the same).

Just my opinion.

Cor
"Morgan Cheng" <mo************ @gmail.comschre ef in bericht
news:11******** **************@ b28g2000cwb.goo glegroups.com.. .

Cor Ligthert [MVP] ??:

Morgan,

To make regex faster try to avoid it by using a replace or an other
method.

Do you mean it is not recommeded to trim string with Regex?
So, the best way for is to code manually handling the string?

Regex is approx. 100 times slower in simple replacements than by instance
"replace".
(This will be gained again in complex situation where you would have to do
a
lot of replacements of course)

I hope this helps,

Cor
"Morgan Cheng" <mo************ @gmail.comschre ef in bericht
news:11******** ************@k7 0g2000cwa.googl egroups.com...

In my case, I have to remove any line containing "0.000000" from input
string.
In below case, it takes about 100 ms for 2k size input string.
Regex.Replace(i nputString, ".*0\\.000000.* \n", "");
I want to optimize it, so i make a static member instance instead of
using static func of Regex;
static Regex filter= new Regex(".*0\\.00 0000.*\n",
RegexOptions.Co mpiled);
And use the static instance to replace the string.
filter.Replace( inputString, "");
Now, it takes about 200 ms. The platform is .net 1.1.
I don't understand why the non-static member function is slower thant
static one. The static function has to compile the regular expression
each time, right?

Is there any better solution to remove substring of specific pattern?
Perhaps I have to write my code to do it.

Thanks,
-Morgan