String Manipulation Alternatives to RegEx

[Tom]

I have string that is 2.5 million bytes long.

I tried using Regular Expressions to look for patterns and replace the
pattern found with a pre-defined text. This works great on some computers
but on some others (win 2K server 1gig of ram) it times out!

So I am looking for an alternative method as I've been told Regular
Expressions are expensive.

Here is my pattern;
white spaces followed by a new line (\n) followed by more spaces plus
a predefined string.

For each occurance, I want to replace this pattern with a predefined string,
thereby removing all the blank lines, cr lf.
Does anyone know of an alternative method to try?
Can you provide an example?
Thanks

Tom

[Nicholas Paldino [.NET/C# MVP]]

Tom,

The regular expression doesn't time out, rather, the request is timing
out. You have to change the executionTimeout value in the httpRuntime
element in the web.config file to extend the timeout of the request.

Also, I would seriously consider using something else if you have 2.5
million bytes of text to search through, as using a regular expression is
not the most efficient manner (you have to load the text to be searched on
as a string, and allocating THAT much continuous memory is going to kill
you).

I would look for third party components that can work on a file, or work
on a stream.

Hope this helps.
--
- Nicholas Paldino [.NET/C# MVP]
- mv*@spam.guard.caspershouse.com

"Tom" <To*@discussions.microsoft.com> wrote in message
news:A1**********************************@microsof t.com...

I have string that is 2.5 million bytes long.

I tried using Regular Expressions to look for patterns and replace the
pattern found with a pre-defined text. This works great on some computers
but on some others (win 2K server 1gig of ram) it times out!

So I am looking for an alternative method as I've been told Regular
Expressions are expensive.

Here is my pattern;
white spaces followed by a new line (\n) followed by more spaces plus
a predefined string.

For each occurance, I want to replace this pattern with a predefined
string,
thereby removing all the blank lines, cr lf.
Does anyone know of an alternative method to try?
Can you provide an example?
Thanks

Tom

[Tom]

Can you recommend any third party tools?

Also, what's weired is I have several RegEx on this string. All of the
other RegEx's work very fast! So I don't know if the file size is the issue.
Maybe, its my RegEx?

Thanks

Tom
"Nicholas Paldino [.NET/C# MVP]" wrote:

Tom,

The regular expression doesn't time out, rather, the request is timing
out. You have to change the executionTimeout value in the httpRuntime
element in the web.config file to extend the timeout of the request.

Also, I would seriously consider using something else if you have 2.5
million bytes of text to search through, as using a regular expression is
not the most efficient manner (you have to load the text to be searched on
as a string, and allocating THAT much continuous memory is going to kill
you).

I would look for third party components that can work on a file, or work
on a stream.

Hope this helps.
--
- Nicholas Paldino [.NET/C# MVP]
- mv*@spam.guard.caspershouse.com

"Tom" <To*@discussions.microsoft.com> wrote in message
news:A1**********************************@microsof t.com...

I have string that is 2.5 million bytes long.

I tried using Regular Expressions to look for patterns and replace the
pattern found with a pre-defined text. This works great on some computers
but on some others (win 2K server 1gig of ram) it times out!

So I am looking for an alternative method as I've been told Regular
Expressions are expensive.

Here is my pattern;
white spaces followed by a new line (\n) followed by more spaces plus
a predefined string.

For each occurance, I want to replace this pattern with a predefined
string,
thereby removing all the blank lines, cr lf.
Does anyone know of an alternative method to try?
Can you provide an example?
Thanks

Tom

[Niki Estner]

"Tom" <To*@discussions.microsoft.com> wrote in
news:A1**********************************@microsof t.com...

I have string that is 2.5 million bytes long.

I tried using Regular Expressions to look for patterns and replace the
pattern found with a pre-defined text. This works great on some computers
but on some others (win 2K server 1gig of ram) it times out!
Regular expressions don't have timeouts.
So I am looking for an alternative method as I've been told Regular
Expressions are expensive.

Here is my pattern;
white spaces followed by a new line (\n) followed by more spaces plus
a predefined string.
Maybe you should post that pattern.
It somehow sounds as if you had a pattern like "\s*\n\s*PredefinedString".
Is that the case? Also post your Regex settings.
For each occurance, I want to replace this pattern with a predefined
string,
thereby removing all the blank lines, cr lf.
Does anyone know of an alternative method to try?

- "for" loop
- "while" loop
- "foreach" loop

Niki

[Jay B. Harlow [MVP - Outlook]]

Tom,

Does anyone know of an alternative method to try?
Rather then read the entire string into memory, have you considered
processing one or two lines at a time using a StreamReader & a state
machine?

Basically read the first line, if its blank, remember it, and read the next
line, if its the pattern, drop the first line, change the next line, write
the next line, read a new first line.

If the first line is not blank, write it

If the next line is not the pattern, write the first line, making the next
line the new first line.

If ???

I would probably work out what the state machine for the above before
attempting to code...

Hope this helps
Jay


"Tom" <To*@discussions.microsoft.com> wrote in message
news:A1**********************************@microsof t.com...I have string that is 2.5 million bytes long.

I tried using Regular Expressions to look for patterns and replace the
pattern found with a pre-defined text. This works great on some computers
but on some others (win 2K server 1gig of ram) it times out!

So I am looking for an alternative method as I've been told Regular
Expressions are expensive.

Here is my pattern;
white spaces followed by a new line (\n) followed by more spaces plus
a predefined string.

For each occurance, I want to replace this pattern with a predefined
string,
thereby removing all the blank lines, cr lf.
Does anyone know of an alternative method to try?
Can you provide an example?
Thanks

Tom