Hi folks

I'm little confused with the regexp syntax.

To show the numbers from 3 to 7 I'm using:
[3-7]
How to show the numbers between 3 and 70 ?
kregexpeditor said, that [3-70] are the numbers 3 to 7 and 0 (or something like that)

I'm no expert but I'm pretty sure that it doesn't work that way. The [...] thing is used for identifying a single character that fits the criteria in the braces. "70" is not a single character, is it? No, it's not. What "[3-70]" is probably interpreted as is "match anything between 3 and 7 and also 0". I don't have a Linux box in front of me right now to test that out though.

Maybe if you explain exactly what you need to do i could help a little bit more.

-- the dudeMAN DavE

I think you're right

Well, I have one big wget log file and I want to see how many files are bigger then 5MB. The file size is like 3.3M , 5.7M , 11M, 11.3M, etc

I'm getting the size with:
cat logfile |grep "Length"|cut -f3 -d" "|grep M |wc -l

Any suggestions ?

I know that awk might work well for you here. Unfortunately I can't test it out right now (as I'm on a public access internet connection running Windows 2000), but it'd be something like this:
Now, I've never tried that (and can't at the moment) and I don't know if it'd work with the 'M' appended on the end. But I do know that with awk the first part is the condition part and the part in the curly-braces {} is the action part. It says "If the first column has a numeric value greater than 5.0 then print out the whole line". I'm assuming that column/field one is the column with the size value in it (the fields are separated by a space. You can change the field-separation character by specifying the -F switch for awk before you put the command in the single quotes).

Sorry I can't think of much beyond that.

-- the duDemAN daVE

You got my brain working on this! And it's too bad/soo sad that I don't have a Linux box in front of me! Arg!

Anyway, I found a page that explains how awk interprets comparisons (whether string or numeric): http://ou800doc.caldera.com/en/SDK_tools/P_NumStr.html

So, from what I read, I believe that if awk is comparing "6.7M" with "5.0" it'll compare it numerically, just like we want. Here's the quote:
If you've never used awk before, just know that $1 is the variable for the first field, $2 is for the second field, etc. And $0 is for the whole line. So, because I wouldn't know which field it is that the number is at, I couldn't write the exact command for you. But you can figure that out.

Let me know what you figure out.

-- the dudemaN DAVEEe

awk sounds good. I've made a little test.

$cat testfile
245M
67M
112M
999M
4M
7M
5M

$cat testfile |awk '$1 > 5 {print $0}'
67M
999M
7M
5M

Unfortunately it has some problems with numbers, begining with 1-4. 245M is bigger than 5M. The "45M" part is probably like a string to him

And also my size is in format (14.6M). With the brackets awk recognize it as a string (i think). I can cut -c 2,3,4,5,6 and the result will be 14.6M, but some numbers are in (10M) format, so they will stay 10M)

Stuck again

change delimiter, since i dont know which logfile you talking about, i can only give you the idea:
awk -FM '$1 > 5'

print is default action, so it can be skipped.