[SOLVED] Text Processing

brainAcid · 05-19-2011, 02:22 PM

I cant find anything that accomplishes what I need. I want to find a 'pattern' and print every line after that until it reaches 'pattern2' until the end of file.

pattern=Trans
pattern2=Port

Quote:

Input File:

Trans

paragraph

Port

paragraph

Trans

paragraph

Port

paragraph

... (Repeats)

Quote:

Output:

Trans

paragraph

Trans

paragraph

Its a file with many occurrences and I only need to extract the paragraph after the string. Thanks!

Vrajgh · 05-19-2011, 03:30 PM

The following awk script reproduces the output you specify from your sample input.

Code:

#! /usr/bin/awk -f

/^Trans$/ { go=1}
/^Port$/  { go=0}

go==1 { print }

brainAcid · 05-19-2011, 03:34 PM

Sweet!

Thanks!

From IRC:

Quote:

sed -n '/TRANSLATION/,/PURPORT/p' out.txt

crts · 05-19-2011, 03:47 PM

Quote:

Originally Posted by brainAcid

Sweet!

Thanks!

From IRC:

Hi,

in your example you do not want to print the last line of the range, i.e. Port, so you might want to adjust the sed command slightly

Code:

sed -n '/Trans/,/Port/{/Port/!p}' file

Vrajgh · 05-20-2011, 04:06 AM

Excellent, I learnt something from this too - I have only ever used that sed notation to extract a single block of text and didn't realise that it extracts any number of blocks between matching lines. :-)

grail · 05-20-2011, 04:47 AM

Code:

awk '/Trans|Port/{g=!g}g' file

brainAcid · 05-20-2011, 06:10 AM

Thanks to everyone!! Very good community here!

I have learned a lot and have accomplished much.

Here is another one:

Quote:

INPUT:

TRANS
paragraph

TRANS
paragraph

...

Quote:

OUTPUT:

Bg1.1
paragraph

Bg1.2
paragraph

...

Now the 'Bg1.1' pattern is in a seperate sorted file.
How do I get each 'TRANS' replaced with 'Bg'?

Thanks!

brainAcid · 05-20-2011, 06:12 AM

http://www.linuxquestions.org/questi...-in-sed-62431/

Hmm...

crts · 05-20-2011, 06:43 AM

Quote:

Originally Posted by brainAcid

http://www.linuxquestions.org/questi...-in-sed-62431/

Hmm...

Not sure where you're going with this but you probably want something like this:

Code:

sed -n '/Trans/,/Port/{/Trans/ {R FILE
d};/Port/!p}' file

Notice, That it is a capital R in contrast to the linked example. It is also important to have line break after the filename FILE because everything that follows the R command will be interpreted as part of the filename.

brainAcid · 05-20-2011, 07:03 AM

Quote:

#!/bin/bash

patternFILE=v.head
workFILE=verses.txt

sed '/TRANSLATION/{R $patternFILE
d}' $workFILE > out.txt

I get output with the TRANS deleted but not replaced.
Using the -n option gives me a blank file.

Quote:

echo "Enter the file name"
read ips

index=0

while read line ; do
sed 's/TRANSLATION/'${line}'/g' verses.txt > out.txt
done < $ips

tried this but it only replaces TRANS with the last line in the file.

?

grail · 05-20-2011, 07:24 AM

Code:

awk '/Trans|Port/{g=!g;if(/Trans/)getline < "file2"}g' file

crts · 05-20-2011, 07:29 AM

Quote:

Originally Posted by brainAcid

I get output with the TRANS deleted but not replaced.

That is because you FILE is supposed to be the filename. Replace FILE with the actual name of your second file.
If you want to store the filename in a variable then you will have to use double-quotes instead of single-quotes to enable expansion. Like this

Code:

#!/bin/bash

patternFILE=v.head
workFILE=verses.txt
sed -n "/Trans/,/Port/{/Trans/ {R $patternFILE
d};/Port/! p}" $workFILE > out.txt

brainAcid · 05-20-2011, 07:31 AM

Say the same thing but replacing Trans. Lets forget about Port because I got that section parsed already. Now I need to replace each Trans with each line in file2. Thanks

grail · 05-20-2011, 07:33 AM

Was there something wrong with my solution?

crts · 05-20-2011, 07:39 AM

Quote:

Originally Posted by brainAcid

Say the same thing but replacing Trans.

Which post are you referring to? There was a typo in my previous post. Fixed that and it worked for me.
Also tried grail's solution and it also works fine.