use sed awk to replace variable defined in another file

Question

I have a users.txt file in the format:

User:CityID
Carl:0212
Syd:0312
Rick:9323

and a city.txt file in the format

Anaheim:0212
San Jose:0312

I need to replace every CityID in the users.txt with the city name from the city.txt file. How can I achieve this using sed and awk?

I can get the column of CityID's using awk using:

awk -F$'\:' '{print $2}'< users.txt

but how do it replace them with the corresponding city name?

Thank you.

When you redirect input from a file awk '...' < file, you lose the ability for awk to populate the often-useful FILENAME variable. Just let awk open the file itself awk '...' file. — Ed Morton
– Ed Morton, Commented Feb 23, 2014 at 20:19

jaypal singh · Accepted Answer · 2014-02-23 19:55:02Z

5

This would work:

awk 'BEGIN{FS=OFS=":"}NR==FNR{a[$2]=$1;next}{$2=a[$2]}1' city.txt user.txt

BEGIN{FS=OFS=":"} : We set the Input and Output Field Separator to :
NR==FNR{a[$2]=$1;next} : We read the city file and create an array to store City Name indexed at City ID
{$2=a[$2]} : We replace the second field in user.txt by referencing the array
1 : This is to print the line

answered Feb 23, 2014 at 19:55

jaypal singh

77.6k24 gold badges108 silver badges147 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

tripleee · Accepted Answer · 2014-02-23 19:57:55Z

2

You can use sed to transform city.txt into a sed script:

sed 's/\([^:]*\):\(.*\)/s%\2%\1%g/' city.txt

then feed that to a second sed instance to process users.txt:

sed 's/\([^:]*\):\(.*\)/s%\2%\1%g/' city.txt |
sed -f - users.txt

Not all sed variants will accept a script file on standard input; you'll have to resort to a temporary file in that case.

answered Feb 23, 2014 at 19:57

tripleee

192k37 gold badges318 silver badges367 bronze badges

2 Comments

BMW Over a year ago

+1 for this approach, but may has a bug, that can't identify some IDs, such as 0212 and 02120

tripleee Over a year ago

One could elaborate the generated script with word or field boundary markers, but given the sample data (fixed field length with leading zero padding) that didn't seem necessary. A quick and dirty fix would be s%:\2$%:\1%

Collectives™ on Stack Overflow

use sed awk to replace variable defined in another file

2 Answers 2

Comments

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related