need help with a bash script

ghostdog74 · 10-08-2009, 04:24 AM

Quote:

Originally Posted by vonbiber

Code:

rm -f output.txt
cat input.txt | while read line
do
    x=$(echo $line | awk '{print $1}'
    y=${x%%.*}.000
    #new line is eg: '00:00:01.000'
    #followed by '   Trusted       <integer> = 0x0000000000000001 (1)'
    newline=${y}"$(echo $line | sed 's?^[^ \t]*??')"
    echo "$newline" >> output.txt
done

no need cat.

ghostdog74 · 10-08-2009, 04:26 AM

Quote:

Originally Posted by lutusp

The title of the article is "Bash Shell Programming in Linux". Nowhere does it say or require that all the discussed commands are built-ins. That isn't the point. The point is for newbies to learn how to get results using a Bash shell session, not to limit themselves to Bash built-ins.

you might want to change the title to "Shell programming in Linux" instead.

ghostdog74 · 10-08-2009, 04:29 AM

Quote:

Originally Posted by smeezekitty

not quite right
here is some example text from the file

Code:

00:00:01.879   Trusted       <integer> = 0x0000000000000001 (1)
00:00:01.879   PCIDeviceNo   <integer> = 0x0000000000000001 (1)
00:00:01.879   PCIFunctionNo <integer> = 0x0000000000000001 (1)
00:00:01.879 
00:00:01.879 [/Devices/piix3ide/0/Config/] (level 4)
00:00:01.880   Type <string>  = "PIIX4" (cch=6)
         ^^^
         \|/
          V
          |_this needs to be changed to zero on every line

if you are sure you want to change only the last 3 digits of the first column, then use awk

Code:

# awk '{gsub(/\.???$/,".000",$1)}1' file
00:00:01.879.000 Trusted <integer> = 0x0000000000000001 (1)
00:00:01.879.000 PCIDeviceNo <integer> = 0x0000000000000001 (1)
00:00:01.879.000 PCIFunctionNo <integer> = 0x0000000000000001 (1)
00:00:01.879.000
00:00:01.879.000 [/Devices/piix3ide/0/Config/] (level 4)
00:00:01.880.000 Type <string> = "PIIX4" (cch=6)

smeezekitty · 10-08-2009, 12:38 PM

Quote:

Originally Posted by ghostdog74

if you are sure you want to change only the last 3 digits of the first column, then use awk

Code:

# awk '{gsub(/\.???$/,".000",$1)}1' file
00:00:01.879.000 Trusted <integer> = 0x0000000000000001 (1)
00:00:01.879.000 PCIDeviceNo <integer> = 0x0000000000000001 (1)
00:00:01.879.000 PCIFunctionNo <integer> = 0x0000000000000001 (1)
00:00:01.879.000
00:00:01.879.000 [/Devices/piix3ide/0/Config/] (level 4)
00:00:01.880.000 Type <string> = "PIIX4" (cch=6)

bash-2.04$ awk
bash: awk: command not found

lutusp · 10-08-2009, 03:54 PM

Quote:

Originally Posted by ghostdog74

you might want to change the title to "Shell programming in Linux" instead.

I assure you, if I do that, someone will surely complain that the article is far too reliant on Bash and not enough on csh, ksh, zsh, Perl, Ruby, Python etc., etc. and that I am maliciously misleading my readers.

Believe me, no matter what you write or how you write it, someone will complain about it, usually someone who doesn't write articles himself.

ghostdog74 · 10-08-2009, 07:25 PM

Quote:

Originally Posted by smeezekitty

bash-2.04$ awk
bash: awk: command not found

let me guess. you are using bash 2.04 on windows...right?

smeezekitty · 10-08-2009, 08:10 PM

why you think?

gnashley · 10-09-2009, 02:45 AM

lutusp, please accept my apology for being so critical. I was having a fit of orneriness...

My guess is that you are doing this on DSL or on 'My own OS' whihc don't have awk.

Here, this works using sh:

Code:

#!/bin/sh

while read LINE ; do
	if [ "${LINE#* }" != "" ] ; then
		echo ${LINE%.*}.000 ${LINE#* }
	else
		echo ${LINE%.*}.000
	fi
done < data.txt

smeezekitty · 10-09-2009, 02:59 AM

Quote:

Originally Posted by gnashley

lutusp, please accept my apology for being so critical. I was having a fit of orneriness...

My guess is that you are doing this on DSL or on 'My own OS' whihc don't have awk.

Here, this works using sh:

Code:

#!/bin/sh

while read LINE ; do
	if [ "${LINE#* }" != "" ] ; then
		echo ${LINE%.*}.000 ${LINE#* }
	else
		echo ${LINE%.*}.000
	fi
done < data.txt

stark verzerrt

lutusp · 10-09-2009, 04:21 AM

Quote:

Originally Posted by gnashley

lutusp, please accept my apology for being so critical. I was having a fit of orneriness...

My guess is that you are doing this on DSL or on 'My own OS' whihc don't have awk.

Here, this works using sh:

Code:

#!/bin/sh

while read LINE ; do
	if [ "${LINE#* }" != "" ] ; then
		echo ${LINE%.*}.000 ${LINE#* }
	else
		echo ${LINE%.*}.000
	fi
done < data.txt

Umm, I already solved this in a more efficient way earlier in this thread:

Code:

while read line
do
    echo ${line/.[0-9][0-9][0-9]/.000}
done

Granted, this thread is getting long, but still.

ghostdog74 · 10-09-2009, 10:09 AM

Quote:

Originally Posted by gnashley

Here, this works using sh:

Code:

#!/bin/sh

while read LINE ; do
	if [ "${LINE#* }" != "" ] ; then
		echo ${LINE%.*}.000 ${LINE#* }
	else
		echo ${LINE%.*}.000
	fi
done < data.txt

your sh is linked to bash, correct? because those aren't sh (bourne) syntax.

konsolebox · 10-09-2009, 10:49 AM

Quote:

Originally Posted by lutusp

Umm, I already solved this in a more efficient way earlier in this thread:

Code:

while read line
do
    echo ${line/.[0-9][0-9][0-9]/.000}
done

Sorry, i didn't read this code properly. The code's not only simpler to the referred code but mine as well.