XML- Sed || Awk Bash script... Help!

Shell Programming and Scripting
1

Hi ! I'm working into my first bash script to make some xml modification and it's going to make me crazy lol .. so I decide to try into this forum to take some ideas from people that really know about this!

This is my situation I've and xml file with a lots of positional values with another tags too but this are the importants tags

<x>28.981743194648736</x>
<y>72.4853691731618</y>
<z>0.0</z>
<x>53.46605522341658</x>
<y>54.678596788603365</y>
<z>0.0</z>

and this continue with differents values in the same way
I need to change all the values of x tags by new ones, particulary with random numbers.. the problem that I've is that I have succes to change that values but..... ! , sed change me the value of all the <x> </x> tags with the same value and I need that all that values were differents.. here my code

#!/bin/bash
NUM=0

while [ $NUM -le 20 ]; do
NUM1=$(echo "scale=3;((${RANDOM})/1000)" |bc)
echo "$NUM1"
sed 's/<x>[-[:alnum:]./]\{1,\}<\/x>/<x>'$NUM1'<\/x>/' myfile.xml > newfile.xml

echo "\$NUM: $NUM"
let NUM=$NUM+1
done

THANKS FOR YOUR TIME ANY IDEA WOULD BE WELCOME !!!

2

line-basedtools like awk, sed, shells, aren't very good at handling xml because they're not recursive parsers. (Well, it parses THEIR language recursively, but doesn't parse text they read in recursively -- things like yacc do..) Usually you'd use a language like perl or python and load an xml module. Could you tell us what your actual goal is instead of the way you want to do it?

3

note that everytime your run this script, you get the same x value in same line. but in the file, every x-value should be unique. If you want to get different x-value in each run, change the srand seed please.

awk '/x/{srand(NR);v=rand()*50+1;$0="<x>"v"</x>"};1' yourFile

output:

<x>12.8894</x>
<y>72.4853691731618</y>
<z>0.0</z>
<x>37.5135</x>
<y>54.678596788603365</y>
<z>0.0</z>
1 Like
4

Thanks Corona668 and sk1418 for your time ! . .

sk1418 your code works like a charm! thanks for your help !, another question...... sorry if I bother you, I almost know nothing of awk , you can put the "output" in a new file like in sed ? and another thing ! you can define how many decimals you need ? like 3.444 or 3.44444444

sed 's/<x>[-[:alnum:]./]\{1,\}<\/x>/<x>'$NUM1'<\/x>/' cornodobeta2.xml > newfile.xml

I tried with > newfile.xml but no lucky

Thanks for your help !!

5

For your Q1: write the output to a new file is as easy as:

awk ... inputFile > yourNewFile

Q2: with precision control:

awk '/x/{srand(NR);v=rand()*50+1;printf  "<x>%.13f</x>\n",v};!/x/'

you can change the "%.13f" to your wish to control the precision. the output of above line with %.13f looks like:

<x>12.8893756120882</x>
<y>72.4853691731618</y>
<z>0.0</z>
<x>37.5135307593800</x>
<y>54.678596788603365</y>
<z>0.0</z>
1 Like
6

Man ... you're simply awesome!
Supose that in the future I need to change the <y> value too, with a value distinct to the x and also random the code will be this ?

awk '/x/{srand(NR);v=rand()*50+1;$0="<x>"v"</x>"};1' myfile.xml > filemod.xml
awk '/y/{srand(NR);v=rand()*50+1;$0="<y>"v"</y>"};1' myfile.xml > filemod.xml

because it doesn't give me the expected result .

Here I Post part of the exactly xml that I'm working .

<mote> 
      <interface_config> 
        se.sics.cooja.interfaces.Position 
        <x>53.39887462699737</x> 
        <y>66.79001043503861</y> 
        <z>0.0</z> 
      </interface_config> 
      <interface_config> 
        se.sics.cooja.contikimote.interfaces.ContikiMoteID 
        <id>4</id>

Thanks Again for your Time and your experience in this !

7

you can change the random seed to get different random number for your 'y'.
eg

awk '/y/{srand(NR*2);v=rand()*77+1;$0="<y>"v"</y>"};1' myfile.xml > filemod.xml

I think it would be ok. (I didn't test).

In fact the easiest way to get a new y is do some math calculation on generated x. e.g. sqr(x) or x + a magic number ...

1 Like
8

Hi sk1418 thanks for your help !, I'm having a problem with the script , when I run this into all the xml file , when it found some word with an x into the words for eg <success_ratio_tx>1.0</success_ratio_tx> it change the value too and the result is this <x>30.7032</x> ... It was my fault because I didn't tell you all my xml file, is there a way that awk just take the exactly <x>..</x> tags ?

Thanks Again for your time , and sorry to be so newbie in this.:wall:

9 
awk '/<x>/{srand(NR);v=rand()*50+1;printf  "<x>%.13f</x>\n",v};!/<x>/'
1 Like
10

Thanks man ! , Now it's working ok . I really appreciate your time helping me in this work !.