CODEWITHSUNDEEP
bashshellseduser-input
Preeyah
Preeyah

Reputation: 373

Bash Script - How do i remove line/word from text file?

i can't remove like one line from the text file after user inputs the book title and author(from the text file)

i tried this code

What i did:

    function remove_book
{
  echo "Title: "
  read title
  echo "Author: "
  read name

  echo $title $name < BookDB.txt
  echo "Book $title by $name successfully removed!"

  sed '/pattern to match/d' ./BookDB.txt  

  #echo "testing remove"
}

but then, it shows this

even it shows like this, they weren't actually removed from the file.. 

Title: 
The Hobbit
Author: 
J.R.R. Tolkien
The Hobbit J.R.R. Tolkien
Book The Hobbit by J.R.R. Tolkien successfully removed!
Harry Potter and the Deathly Hallows J.K. Rowling
The Maze Runner James Dashner
Life Without Limits Nick Vujicic
The Hobbit J.R.R. Tolkien

Desired Output:

1)  Add new book
2)  Remove existing book info
3)  Update book info and quantity
4)  Search for book by title/author
5)  Process a book sold
6)  Inventory summary report
7)  Quit
Please enter your option: 2
Title : Biography of Crocodile Dundee
Author : Crox Swamplund
    Error! Book does not exists!


1)  Add new book
2)  Remove existing book info
3)  Update book info and quantity
4)  Search for book by title/author
5)  Process a book sold
6)  Inventory summary report
7)  Quit
Please enter your option: 2
Title : C++ for dummies
Author : Edward Scissorhands
Book Title ‘C++ for dummies’ removed successfully!

(the book and author's names in the Desired Output are just examples)

FROM

Harry Potter and the Deathly Hallows J.K. Rowling
The Maze Runner James Dashner
Life Without Limits Nick Vujicic
The Hobbit J.R.R. Tolkien [remove one line, like this Hobbit book]

TO

Harry Potter and the Deathly Hallows J.K. Rowling
The Maze Runner James Dashner
Life Without Limits Nick Vujicic

how do i remove one line? after user inputs the title and author's name? help me please thanks! :)

Upvotes: 2

Views: 1834

Answers (2)

Wintermute
Wintermute

Reputation: 44023

Do not do this with sed. At some point, someone somewhere will write a book titled "Good times/bad times" or "Ca$h ca$h ca$h!!!1! Make $$$ in your spare time!" or "What you always wanted to know (and never dared to ask)" and sed will screw it up horribly because the special characters have meaning for its pattern matching engine.

You can do it with GNU awk like this:

awk -i inplace -v title="$title" -v author="$name" '$0 != title " " author' BookDB.txt

This will select all lines from the file that are not exactly the contents of $title followed by one space followed by the contents of $name. Since the shell variables are not substituted into the awk code but passed through awk's -v parameter, no interpretation of special characters takes place.

Also: Are you sure you want to do it inplace? I like to keep a backup in case an operation goes wrong. Like

cp BookDB.txt BookDB.txt~
awk -v title="$title" -v author="$name" '$0 != title " " author' BookDB.txt~ > BookDB.txt

Then if things go wrong or you remove the wrong book, rolling back is easy. Also, this will work with other awks than the one from GNU.

Alternatively, you could use grep like this:

cp BookDB.txt BookDB.txt~
grep -vxF "$title $name" BookDB.txt~ > BookDB.txt

Where -x tells grep to match only if the match is the whole line and -F tells it to take the pattern as a fixed string rather than a regex.

Upvotes: 3

Kent
Kent

Reputation: 195049

you may want to add (gnu?) sed's -i option

read man sed to know more info about -i

expand a bit...

If you want the changed done by sed to be save in your file, you can use -i option, edit files in place. An example:

kent$  cat f
1
2
3
4
5

kent$  sed -i '/2/d' f

kent$  cat f
1
3
4
5

Upvotes: 1

Related Questions