bash Script to replace words

Question

Looking to run a script on the server to look at a path of a file and replace a word whereas it matches in the div.

So need to replace _toself to viewers where author equals a certain email [email protected]

URL=/var/www/sever/temp/fhyw1 FILE=user.txt

<div class='entry'>
  <div class='pageurl'>temp/fhyw1</div>
  <div class='context'>text</div>
  <div class='subject'>testing</div>
  <div class='notetext'></div>
  <div class='signed'>USER</div>
  <div class='author'>[email protected]</div>
  <div class='color'>0</div>
  <div class='visibility'>shared</div>
  <div class='to'>_toself</div>
  <div class='num'>4</div>
</div>
<div class='entry'>
  <div class='pageurl'>temp/fhyw1</div>
  <div class='context'>text</div>
  <div class='subject'>testing</div>
  <div class='notetext'></div>
  <div class='signed'>USER</div>
  <div class='author'>[email protected]</div>
  <div class='color'>0</div>
  <div class='visibility'>shared</div>
  <div class='to'>_viewers</div>
  <div class='num'>4</div>
</div>

potong · Accepted Answer · 2011-11-14 14:35:09Z

1

This sed solution might work for you:

 sed -e '/^<div class=.entry.>/,\_^</div>_{//!{H;d};\_^</div>_!{h;d};x;/author.>[email protected]/s/_toself/SUBSTITUTE TEXT/;p;x}' text_file

N.B. You will need to replace SUBSTITUE TEXT with the viewers,_viewers or whatever

The sed command allows all lines other than those between <div class=.entry.> and </dev>(. allows for single 'or double quotes ") to pass through unchanged. If the line begins with <div class=.entry.> it is copied to a register call the hold space (HS) and then the pattern space (PS) is deleted. All other lines are appended to the HS and then deleted accepting the line </div>. When this line appears the HS is swapped with the PS and if this multiline contains author.>[email protected] then SUBSTITUTE TEXT is substituted for _toself. The multiline is printed out regardless, then the PS replaces the HS and it in turn to printed out.

edited Nov 14, 2011 at 14:35

answered Nov 11, 2011 at 19:45

potong

59.3k6 gold badges55 silver badges92 bronze badges

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

2 Comments

Grimlockz Over a year ago

Thanks for the sed command seems to work ok but I'm getting blank lines inserted every div entry and /div ?

potong Over a year ago

I must have mucked up whilst copying and pasting. Solution corrected - a regex had gone astray \_^</div>_

ДМИТРИЙ МАЛИКОВ · Accepted Answer · 2011-11-11 17:22:06Z

We have some text

$> cat ./text 
<div class='entry'>
  <div class='pageurl'>temp/fhyw1</div>
  <div class='context'>text</div>
  <div class='subject'>testing</div>
  <div class='notetext'></div>
  <div class='signed'>USER</div>
  <div class='author'>[email protected]</div>
  <div class='color'>0</div>
  <div class='visibility'>shared</div>
  <div class='to'>_toself</div>
  <div class='num'>4</div>
</div>
<div class='entry'>
  <div class='pageurl'>temp/fhyw1</div>
  <div class='context'>text</div>
  <div class='subject'>testing</div>
  <div class='notetext'></div>
  <div class='signed'>USER</div>
  <div class='author'>[email protected]</div>
  <div class='color'>0</div>
  <div class='visibility'>shared</div>
  <div class='to'>_viewers</div>
  <div class='num'>4</div>
</div>

And we need to replace _toself 'to' value with viewers, but only in divs, where 'author' equals a [email protected]

I think sed can helps you, but you should have some experience with it to formulate all condition with sed syntax.

So, we can read file in while loop, cut it into a div-blocks and change one value by another only if blocks 'authors' value is equal some email.

#!/bin/bash

mail="[email protected]"
to_value_old=_toself
to_value_new=viewers

while IFS= read -r line; do
    if [[ -z "$( echo "$line" | grep -o -P "^<\/div>$" )" ]]; then
        entry_block="${entry_block}${line}\n"
    else
        entry_block="${entry_block}</div>\n"
        entry_block="$( echo -e "${entry_block}" )"
        if [[ -n "$( echo "${entry_block}" | grep -P "\<div class=\'author\'\>${mail}\<\/div\>" )" ]]; then
            entry_block="$( echo "${entry_block}" | sed -r -e "s/<div\ class='to'>${to_value_old}<\/div>/<div\ class='to'>${to_value_new}<\/div>/"  )"
        fi
        echo "${entry_block}"
        entry_block=""
    fi
done < ./text

And we get

$> ./div.sh 
<div class='entry'>
  <div class='pageurl'>temp/fhyw1</div>
  <div class='context'>text</div>
  <div class='subject'>testing</div>
  <div class='notetext'></div>
  <div class='signed'>USER</div>
  <div class='author'>[email protected]</div>
  <div class='color'>0</div>
  <div class='visibility'>shared</div>
  <div class='to'>viewers</div>
  <div class='num'>4</div>
</div>
<div class='entry'>
  <div class='pageurl'>temp/fhyw1</div>
  <div class='context'>text</div>
  <div class='subject'>testing</div>
  <div class='notetext'></div>
  <div class='signed'>USER</div>
  <div class='author'>[email protected]</div>
  <div class='color'>0</div>
  <div class='visibility'>shared</div>
  <div class='to'>_viewers</div>
  <div class='num'>4</div>
</div>

Done.

Thanks for this, i'm currently getting replace_anno.sh: 19: [[: not found replace_anno.sh: 19: [[: not found -e </div> when trying to run it with the text file

Didier Trosset · Accepted Answer · 2011-11-11 11:53:34Z

0

If you only want to replace all occurrences of _toself by something else, then sed will do the job perfectly.

sed 's/_toself/replacement_string/'

If you want to do this only within the div with a specified author then it's a bit more tricky.

answered Nov 11, 2011 at 11:53

Didier Trosset

37.7k14 gold badges89 silver badges126 bronze badges

2 Comments

Grimlockz Over a year ago

Yep going to need to do it for a specified author

Grimlockz Over a year ago

Not too much - Grepping and find hasn't helped too much. Looking at putting it through a while read

Collectives™ on Stack Overflow

bash Script to replace words

3 Answers 3

2 Comments

1 Comment

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

2 Comments

1 Comment

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related