LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Programming (https://www.linuxquestions.org/questions/programming-9/)
-   -   Scramble characters in AWK (https://www.linuxquestions.org/questions/programming-9/scramble-characters-in-awk-807497/)

MTK358 05-12-2010 08:11 PM

Scramble characters in AWK
 
I see I'm finally posting an AWK question rather than an answer for a change :)

I wanted to make an AWK script that would scramble all the characters in each field, but leave the first and last characters where they were.

How would I do this?

Sergei Steshenko 05-12-2010 09:09 PM

Quote:

Originally Posted by MTK358 (Post 3966454)
...
How would I do this?

By reading the manual.

Sergei Steshenko 05-12-2010 09:36 PM

And, by the way, it conceptually is very similar to the ways it can be done in Perl - the same entities are involved. So I'm surprised you are asking it - you have already read Perl documentation on regular expressions, haven't you ?

grail 05-12-2010 10:15 PM

My only hint would be, think of options for what FS can be

Sergei Steshenko 05-12-2010 10:24 PM

Quote:

Originally Posted by grail (Post 3966534)
My only hint would be, think of options for what FS can be

It is definitely a possibility; even in 'awk' "there is more than one way to do it".

Sergei Steshenko 05-12-2010 10:25 PM

Oh, and by the way, starting from the way one would do it with his/her eyes and fingers still applies.

Sergei Steshenko 05-13-2010 01:22 AM

MTK358, if I remember correctly, the thing you want to do with 'awk' is very close to what you once wanted to do with Perl - I mean that OO "C" attempt of yours. I think in that case you meant to replace string middle leaving the beginning and the end intact. This time the beginning and the end are just single characters, but other than that it's the same task, isn't it ?

MTK358 05-13-2010 08:03 AM

Thanks for emotionally spoiling another project for me.

For some reason this makes me so emotional.

Is there a way to not have the "View Post" button on ignored posts?

And is there a way to have other members write as if no one replied?

The mods claim that it will make no difference here:

http://www.linuxquestions.org/questi...hreads-799110/

but they are WRONG WRONG WRONG!!!

:cry:

MTK358 05-13-2010 08:08 AM

Quote:

Originally Posted by Sergei Steshenko (Post 3966482)
By reading the manual.

I did read some if it, and I found a nice list of built-in functions.

Even so, it's not like the manual says how do do this.

Oh, I get it.

You just want me to look STUPID.

pixellany 05-13-2010 08:20 AM

MTK;
Please count to ten before continuing in this thread. What I see is a terse comment from Sergei, followed by several helpful comments. It would have been much better to respond to the latter.

Regardless, do not post anything else here except a discussion of the substance of your problem and the suggestions so far.

To discuss anything else, send me a PM or e-mail.

thank you

grail 05-13-2010 08:22 AM

Ok ... well he seems to be offline now, so do you have any questions I can help with?
Did you understand my hint?

grail 05-13-2010 08:23 AM

+1 pixellany to just continuing on the thread topic :)

MTK358 05-13-2010 08:51 AM

For example:

Quote:

Originally Posted by Sergei Steshenko (Post 3966489)
And, by the way, it conceptually is very similar to the ways it can be done in Perl - the same entities are involved. So I'm surprised you are asking it - you have already read Perl documentation on regular expressions, haven't you ?

I didn't even think that this has anything to do with regular expressions.

I just asked for an algorithm, or some hints on how to make one.

I noticed many times before that he thinks I don't read anything. There is a difference between not reading something and reading something that you have no clue about. It's as if he doesn't understand the concept of not understanding.

It also seems to me that he looks down upon me for being stupid for doing something before I fully understand it. But how else do you learn?

For example this little comment that really hurt me:

Quote:

Originally Posted by Sergei Steshenko (Post 3951033)
Just marvelous. You start a server and do not know where its log files are ? And how are you going to debug anything at all ?

Let me tell you something, Sergei! I was just doing that for fun.

I am not working in a big corporation setting up a huge corporate server without knowing where log files are.

I am sitting relaxed at home at my desktop, setting up a server just for the sake of learning how servers work, without any real intention of serving a public website, and to have a platform to learn PHP on.

Also, I find your terse language difficult to understand, and to a certain degree I find it condescending.

And one last thing: I just can't stand those cryptic puzzles and weird ways of doing things that so many members disagree on.

You are the polar opposite of LQ's fun, helpful atmosphere.

So if you want to help me, please understand these concepts first:
  • Not understanding
  • Doing stuff for fun
  • Not being so serious
  • Google searches that bring no relavent results
  • Not forcing me to do stuff "your way", which is different for everyone else's way.

If you disagree even slightly, I kindly ask you not to ever reply to my threads again. I would rather not get any replies at all than get a reply from you.

pixellany 05-13-2010 08:58 AM

MTK;
I asked you to keep this to the substance of your question. Because you ignored my request, you have been given a 1-day "time out".

Please contact me or any moderator privately to discuss further.

PTrenholme 05-13-2010 12:18 PM

Not a direct answer to your question, but an alternative:

While it's quite easy in gawk, I wanted to see if I could do it using bash. Here's how I did it:
Code:

#!/bin/bash
# Scramble all characters between the first an last character in each input token
#
scramble ()
{
  declare -a inside
  local first last i
  if [ ${#1} -le 3 ]
  then
#  If here, nothing to do
    scramble="${1}"
  else
#  We have something to do . . .
    first="${1:0:1}"
    last="${1:$((-1)):1}"
    i=1
    while [[ "${i}" -lt $((${#1}-1)) ]]
    do
      inside[${i}]=${1:$((i)):1}
      i=$((${i}+1))
    done
    scramble="${first}"
    for i in $(echo -n $(shuf -e "${inside[@]}"))
    do
      scramble="${scramble}"${i}
    done
    scramble="${scramble}""${last}"
  fi
}
for ok in "${@}"
do
  scramble "${ok}"
  echo -n "${scramble}"" "
done
echo

and here's how the output looks:
Code:

$ ./scramble this is a test
tihs is a test
$ ./scramble this is a test
this is a test
$ ./scramble this is a test
this is a test
$ ./scramble this is a test
tihs is a tset

Note the use of the shuf function to do the actual permutation. While a permutation function is fairly easy to write, I see no need to "reinvent the wheel."


All times are GMT -5. The time now is 06:39 PM.