replacement of slow unpack

[cyl]

this block of code took about 25 seconds in my computer

open A,"a_500mb_file";
binmode A;
while(sysread(A,$x,256)){
#do nothing
}
close A;

and this took 215 seconds

open A,"a_500mb_file";
binmode A;
while(sysread(A,$x,256)){
my @c=unpack('C*',$x);
}
close A;

So why is unpack so slow here? I thought it very fast before. Do I use
it in a wrong way or is there any replacement? Thanks.

 

[John W. Krahn]

this block of code took about 25 seconds in my computer

open A,"a_500mb_file";
binmode A;
while(sysread(A,$x,256)){
#do nothing
}
close A;

and this took 215 seconds

open A,"a_500mb_file";
binmode A;
while(sysread(A,$x,256)){
my @c=unpack('C*',$x);
}
close A;

So why is unpack so slow here? I thought it very fast before. Do I use
it in a wrong way or is there any replacement? Thanks.

Yes it is slow, but it is faster then the alternatives. The lesson here is
try to use the contents of $x without dividing it up into individual characters.



John

 

[cyl]

John W. Krahn 寫é“：

Yes it is slow, but it is faster then the alternatives. The lesson here is
try to use the contents of $x without dividing it up into individual characters.

In my case, the contents are network packet. If I don't unpack it, is
there other way to analyze it (like parsing IP header) with the binary
string?

 

[xhoster]

John W. Krahn =E5=AF=AB=E9=81=93=EF=BC=9A


In my case, the contents are network packet. If I don't unpack it, is
there other way to analyze it (like parsing IP header) with the binary
string?

Don't unpack the whole thing, just the header.

Xho

 

[John W. Krahn]

John W. Krahn 寫é“：


In my case, the contents are network packet. If I don't unpack it, is
there other way to analyze it (like parsing IP header) with the binary
string?

It depends on how you want to analyse it.


John

 

[cyl]

At last, I wrote the parsing code in C and use swig to wrap it so I can
still use perl to write non-critical parts. The improvement is amazing.
The time is reduced to 1/5.