Efficiency of IsDigit, IsSymbol, IsLetter, etc

[flynn]

I am writing a string parser and wondering how efficient it is to call the Is* and .Char() methods?

 

Performance is an issue in this case since I need to make multiple passes through the string to 1) remove HTML tags , 2) remove CR-NL's and 3) parse the string for multiple combinations of characters. The strings can be up to 100k in size.

 

In C, I can compare characters using:

if str[x] = '\n' ...

and it is very fast since there is no function call overhead.

 

At the moment, I am using this type of code. Relatively speaking, how fast is the .Chars() method? I'm assuming I can't view the assembly source like I could with VB6.

 

if (s.chars(x) = "5") then ...

 

Should I be worried about the performance of the Is* methods?

[Afraits]

If you are performing string parsing - perhaps regular expressions may be useful for you, although more overhead they are quite flexible and powerful and so may work out better in the long run, check the regular expressions for examples and as a place to post questions.

 

For your actual question however I have no idea :)

[IngisKahn]

Both ways are just as slow. Stick with Regular Expressions. Used with the Compiled option they are extreamly fast.

[Talyrond]

If performance is an issue I would use a profiler, you will be amassed where bottle necks occur

 

Talyrond