I have this file (numbers.txt):

number_0
number_1
banana
number_2
orange
banana
number_3
number_4
number_1
apple
number_2
number_5
number_6

In the following sed command

$ sed -e '/number_1/,/number_2/ !d ' numbers.txt > out.txt
$ cat out.txt
number_1
banana
number_2
number_1
apple
number_2

I had twice answears.
However, I want only the last one.
How can I get this?
I want only the last pattern.

You could invert your logic:
By the way, you don't need "-e" to precede a single sed command. Also, you can separate commands with a semicolon instead. At least with gnu sed.

Thanks for your help.

If I add another number_1 (there is only one number_2 in the text), I will have a new problem.


number --> add
number_1 --> add
number_0
number_1
banana
number_2
orange
banana
number_3
number_4
number_1
apple
number_2
number_5
number_6
number_1 --> add
number_100 --> add

$ sed -n '/number_1/,/number_2/{
> p;
> /number_2/q }' numbers.txt
number_1
number_0
number_1
banana
number_2

I solved this with:

$ sed : numbers.txt | sed -n '/number_1/,/number_2/{p;/number_2/q }' | sed '1! G ; h ; $! d' | \
> sed -n '/number_2/,/number_1/{p; /number_1/q}' | sed '1! G ; h ; $! d'
number_1
banana
number_2

Is another way to solve this?

don't quite really understand what you are trying to get. where's the
number_1
apple
number_2
?
maybe you want to describe clearly what you are trying to get.?

Oh! I thought that I had got to solve it.

The text below is a book from the Bible. In portuguese is "Eclesiastes".

If you noted, the chapter 1 begin with " and finish with (Ecl 1), the chapter
2 with " and finish with (Ecl 2) and so on.

If I have a file than has only one chapter, I can read the verse that I choose to read.
However, when I put two chapters inside it I do not get to choose the chapter and the verse
that I want to read.
For example, read chapter 2 and verse 4.

I started to learn sed trying to do this!

"1. Palavras do Eclesiastes, filho de Davi, rei de Jerusalém.
2. Vaidade das vaidades, diz o Eclesiastes, vaidade das vaidades! Tudo é vaidade.
3. Que proveito tira o homem de todo o trabalho com que se afadiga debaixo do sol?
4. Uma geração passa, outra vem; mas a terra sempre subsiste.
5. O sol se levanta, o sol se põe; apressa-se a voltar a seu lugar; em seguida, se levanta de novo.
.
.
.
18. Porque no acúmulo de sabedoria, acumula-se tristeza, e que aumenta a ciência, aumenta a dor".
(Ecl 1)

"1. Eu disse comigo mesmo: Vamos, tentemos a alegria e gozemos o prazer. Mas isso é também vaidade.
2. Do riso eu disse: Loucura! e da alegria: Para que serve?
3. Resolvi entregar minha carne ao vinho, enquanto meu espírito se . . .
4. Empreendi grandes trabalhos, construí para mim casas e plantei vinhas;
5. fiz jardins e pomares, onde plantei árvores frutíferas de toda espécie;
.
.
.
25. pois, quem come e bebe, senão graças a ele? Àquele que lhe é agradável Deus dá sabedoria, ... passa".
(Ecl 2)

There is a pattern that is very useful.
sed -n '/<pattern>/,/^$/'

If you organize your data so that sections are separated by blank lines, then this will select the entire section. ( Such as a chapter ).

So if you had:
(Ecl 1)
1. ...
2. ...
...

(Ecl 2)
1. ...
2. ...
...

Then you could use
To print out a particular verse like: showverse 'Ecl 1' 2
You could test for the number of arguments, and if there is just one argument run:
instead to print out the entire chapter.

When I run the codes there is the following message:

sed: -e expressão #1, caractere 5: unexpected `,'

expressão = expression
caractere character

I could not find a solution.
I "googled", but I did not the answer.
However, I am learning very much about sed!!!