问题
I was trying to use sed to do the following:
In file looks something like:
FirstLine SecondLineEEE AAAblablabla ForthLineEEE FifthLine LastLine
I want to remove EEE (but keep the rest of second line) and the whole line starting with AAA and keep other part of the file intact.
The expected result is (as seen, if the following line doesn't start with AAA, it will be kept, that's why I need to match multiple lines.)
FirstLine SecondLine ForthLineEEE FifthLine LastLine
How should I do it? Thanks in advance!
回答1:
This might work for you:
echo -e "FirstLine\nSecondLineEEE\nAAAblablabla\nLastLine" |
sed '/EEE$/{N;s/EEE\nAAA.*//}'
FirstLine
SecondLine
LastLine
回答2:
fge@erwin ~ $ sed '/EEE$/d; /^AAA/d' <<EOF
> FirstLine
> SecondLineEEE
> AAAblablabla
> LastLine
> EOF
FirstLine
LastLine
/re/d will remove all lines from the input matching regex re. Adapt the patterns to your needs.
回答3:
May be this could work -
sed '/EEE$/N;s/\(.*\)EEE\nAAA.*/\1/' filename
Test:
[jaypal:~/Temp] cat file
FirstLine
SecondLineEEE
AAAblablabla
ForthLineEEE
FifthLine
LastLine
[jaypal:~/Temp] sed '/EEE$/N;s/\(.*\)EEE\nAAA.*/\1/' file
FirstLine
SecondLine
ForthLineEEE
FifthLine
LastLine
来源：https://stackoverflow.com/questions/8635054/matching-pattern-span-multiple-line-and-remove-those-matching-lines