Home / ubuntu / Questions / 1408154
Accepted
genderbee
Asked: 2022-05-13 04:37:32 +0800 CST

insert variables from 2 files based on 1 row

  • 772

I have 3 source files.

1.txt - template.

<field name="COL1" label="COL2" data-source="COL1" classes="attribute" category="Attribute"/>

2.txt - variables for COL1.

IDNName
BusinessUnit
WWB
IncentiveID

3.txt - variables for COL2.

IDN Name
Business Unit
WWB
Incentive ID

I need output like this (take 1st line from 2.txt and 3.txt and insert to template, then take 2nd line from 2.txt and 3.txt and insert to template, etc...)

<field name="IDNName" label="IDN Name" data-source="IDNName" classes="attribute" category="Attribute"/>
<field name="BusinessUnit" label="Business Unit" data-source="BusinessUnit" classes="attribute" category="Attribute"/>
<field name="WWB" label="COL2" data-source="WWB" classes="attribute" category="Attribute"/>
<field name="IncentiveID" label="Incentive ID" data-source="IncentiveID" classes="attribute" category="Attribute"/>

Is possible to do it by one for condition with sed? I tried, but I don't know how to combine 2 files in one row.

for i in $(cat 2.txt); do cat 1.txt | sed 's/COL1/'$i'/g'; echo; done

Thanks.

bash

1 Answers

  1. Best Answer

    If you know that there's a "safe" delimiter such as TAB that may not occur in your 2.txt or 3.txt files, then you could do

    $ paste 2.txt 3.txt | while IFS=$'\t' read -r src lbl; do 
    sed -e 's/"COL1"/"'"$src"'"/g' -e 's/"COL2"/"'"$lbl"'"/' 1.txt
  done
<field name="IDNName" label="IDN Name" data-source="IDNName" classes="attribute" category="Attribute"/>
<field name="BusinessUnit" label="Business Unit" data-source="BusinessUnit" classes="attribute" category="Attribute"/>
<field name="WWB" label="WWB" data-source="WWB" classes="attribute" category="Attribute"/>
<field name="IncentiveID" label="Incentive ID" data-source="IncentiveID" classes="attribute" category="Attribute"/>

    Alternatively with GNU parallel:

    $ parallel --link sed -e 's/"COL1"/"{1}"/g' -e 's/"COL2"/"{2}"/' 1.txt :::: 2.txt 3.txt
<field name="IDNName" label="IDN Name" data-source="IDNName" classes="attribute" category="Attribute"/>
<field name="BusinessUnit" label="Business Unit" data-source="BusinessUnit" classes="attribute" category="Attribute"/>
<field name="WWB" label="WWB" data-source="WWB" classes="attribute" category="Attribute"/>
<field name="IncentiveID" label="Incentive ID" data-source="IncentiveID" classes="attribute" category="Attribute"/>

    However if you're actually working with XML data, then you might want to use a proper XML tool - xmlstarlet is one possibility, but since I'm more familiar with jq here's an option using kislyuk's xq from yq: Command-line YAML/XML/TOML processor - jq wrapper for YAML, XML, TOML documents

    $ xq -x --rawfile src 2.txt --rawfile lbl 3.txt '
    foreach ([($src | split("\n")[0:-1]), ($lbl | split("\n")[0:-1])] | transpose[]) as $a
      (.; .field |= (."@name" |= $a[0] | ."@label" |= $a[1] | ."@data-source" |= $a[0]))
' 1.txt
<field name="IDNName" label="IDN Name" data-source="IDNName" classes="attribute" category="Attribute"></field>
<field name="BusinessUnit" label="Business Unit" data-source="BusinessUnit" classes="attribute" category="Attribute"></field>
<field name="WWB" label="WWB" data-source="WWB" classes="attribute" category="Attribute"></field>
<field name="IncentiveID" label="Incentive ID" data-source="IncentiveID" classes="attribute" category="Attribute"></field>
    • 3