matching only the first string within quotes with grep

MadeInGermany · 12-19-2016, 07:13 AM

Regarding sed
The /2 option for the s command does not work, because the rest of the line is absorbed by the first match, so the 2nd match does not apply.
However, the \(...\)\{n\} reference can do it!

Code:

sed 's/\([^"]*\("[^"]*"\)\)\{1\}.*/\2/' sampleLine
"Appendix"
sed 's/\([^"]*\("[^"]*"\)\)\{2\}.*/\2/' sampleLine
"Full Program Listings"

The \2 references the inner \( \)

vincix · 01-03-2017, 03:27 AM

(From the Sed and Awk book)

Quote:

Substitution
We have already demonstrated many uses of the substitute command. Let’s look carefully at its syntax:
[address]s/pattern/replacement/flags
where the flags that modify the substitution are:
n A number (1 to 512) indicating that a replacement should be made for only the n th occurrence of the pattern.

It's true that it wouldn't be the nth occurrence in the whole file, but it works for each line where there's a match.

grail · 01-03-2017, 07:00 AM

Quote:

Originally Posted by vincix

(From the Sed and Awk book)

It's true that it wouldn't be the nth occurrence in the whole file, but it works for each line where there's a match.

Not sure what value this post was with regard to the original question??

vincix · 01-03-2017, 07:08 AM

The discussion evolved. In the 5th post and later there was the question of how to match a certain occurrence, but it was much simpler than expected with sed. That's the value with regard to the whole discussion.