how do i replace a text string in a file with a random string? (with sed etc)

steve51184 · 09-02-2010, 12:43 AM

hey all i have a line in a text file that has 40 random characters within a tag and i want to change the characters to a new set of 40 random characters (alphanumeric a-z 0-9 etc)

the line in the text file looks like this:

Quote:

<string>40randomcharacters</string>

how would i go about doing that?

also second question same as the above but how would i remove them instead of replacing them?

druuna · 09-02-2010, 01:13 AM

Hi,

I do hope your example is relevant, using tags might complicate things a bit:

Changing the 40 chars:
sed 's%<string>40randomcharacters</string>%<string>NEW40RANDOM</string>%' infile

Removing them:
sed 's%<string>40randomcharacters</string>%<string></string>%' infile

If you also want to remove the tags:
sed 's%<string>40randomcharacters</string>%%' infile

Hope this helps.

steve51184 · 09-02-2010, 01:23 AM

wow thank you for the fast reply and that is almost what I need

Basicly the 40randomcharacters is just that random and that command won't find that line :/

Also I'm wanting to replace it with a new random lot of 40 random characters..

In other words I need to replace everything within the <string> tag with 40 random characters

Sorry for the confusion and I hope that makes sense

druuna · 09-02-2010, 01:40 AM

Hi,

Totally my bad, it is still early where I live....

sed 's%<string>.*<string>%<string>NEW40RANDOM</string>%' infile

The above will replace anything within the string tags with NEW40RANDOM. You don't mention how these 40 new chars are generated, if they are in a variable then this should work (VARIABLE holds the new chars):

sed "s%<string>.*<string>%<string>$VARIABLE</string>%" infile

Mind the quoting, it switched from single to double.

Hope this is what you are looking for.

steve51184 · 09-02-2010, 01:44 AM

They will be genarated by this cod or part of it lol

Not sure how to do it but thats why I'm asking

ghostdog74 · 09-02-2010, 01:58 AM

Quote:

Originally Posted by steve51184

hey all i have a line in a text file that has 40 random characters within a tag and i want to change the characters to a new set of 40 random characters (alphanumeric a-z 0-9 etc)

If you have Python

Code:

import random
import string
alnum=string.ascii_letters+string.digits
s=""
for i in range(40):
    s=s+random.choice(alnum)
with open("file") as f:
    for line in f:
        if "</string>" in line:
            line=line.replace("<string>","<string>"+s)
        print line.rstrip()

druuna · 09-02-2010, 02:11 AM

Hi,

Took a little longer, but this works:

Code:

#!/bin/bash

nanosectime=`date '+%N'`
awk -v NST=$nanosectime '
BEGIN { 
split("ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789", choices, "")
srand(NST)
i = 1
while ( i <= 40 ) {
  printf("%s", choices[int(1 + rand() * 61)])
  i++
  }
print ""
}'

ghostdog74 · 09-02-2010, 02:24 AM

srand(systime()) might be sufficient..

druuna · 09-02-2010, 02:38 AM

Hi,

Quote:

Originally Posted by ghostdog74

srand(systime()) might be sufficient..

Doesn't systime() stop at seconds?

With the speed of modern computers you might end up feeding the same string a few times. The nanosectime=`date '+%N'` makes sure that the number is always different. Just a precaution on my side.

steve51184 · 09-02-2010, 09:56 AM

Quote:

Originally Posted by ghostdog74

If you have Python

Code:

import random
import string
alnum=string.ascii_letters+string.digits
s=""
for i in range(40):
    s=s+random.choice(alnum)
with open("file") as f:
    for line in f:
        if "</string>" in line:
            line=line.replace("<string>","<string>"+s)
        print line.rstrip()

not to sure about python.... how would i run that script and how do i set it to run on a specific file? save it as script.py and run it as './script.py file.txt' ?

Quote:

Originally Posted by druuna

Hi,

Took a little longer, but this works:

Code:

#!/bin/bash

nanosectime=`date '+%N'`
awk -v NST=$nanosectime '
BEGIN { 
split("ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789", choices, "")
srand(NST)
i = 1
while ( i <= 40 ) {
  printf("%s", choices[int(1 + rand() * 61)])
  i++
  }
print ""
}'

could that be integrated into the other sed script you posted before to make it 100% complete?

ghostdog74 · 09-02-2010, 09:58 AM

Save as a filename, run

Code:

$ python myscript.py

steve51184 · 09-02-2010, 10:03 AM

Quote:

Originally Posted by ghostdog74

Save as a filename, run

Code:

$ python myscript.py

so:

python myscript.py myfile.txt

?

druuna · 09-02-2010, 10:11 AM

Hi,

Code:

#!/bin/bash

nanosectime=`date '+%N'`
rndString=$(awk -v NST=$nanosectime '
BEGIN { 
split("ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789",
choices, "")
srand(NST)
i = 1
while ( i <= 40 ) {
  printf("%s", choices[int(1 + rand() * 61)])
  i++
  }
print ""
}')

sed -i.bak "s%<string>.*</string>%<string>$rndString</string>%" infile

This will change in place and creates a copy of the original with a .bak extension attached.

Hope this helps.

steve51184 · 09-02-2010, 10:11 AM

tried to run it as:

python myscript.py myfile.txt

and also changed 'file' t oteh location of myfile.txt but got this error:

Quote:

file.py:7: Warning: 'with' will become a reserved keyword in Python 2.6
File "file.py", line 9
as f:
^
SyntaxError: invalid syntax

steve51184 · 09-02-2010, 10:42 AM

perfect thank you all very much