awk onliner script to search files under a dir for 2 strings

I need to search for files which has string开发者_高级运维 str1 appearing before string str2. both strings are in separate lines.

for example, file1 looks like:

abc
def
str1
ghi
str2

file2 looks like:

abc
str2
def
ghi
str1
pqe

My search should return file1.

It should be a one liner script I can run on command line on unix.

Thanks in advance.

So, here is your awk one liner

awk -vRS="\0777" '/str1.*str2/{print FILENAME}' file*

Sed 1 liner for above:

F="file1" && test ! -z $(sed -n '/str1/,/str2/{/^str2$/p;}' "$F") && echo "$F"

F="file2" && test ! -z $(sed -n '/str1/,/str2/{/^str2$/p;}' "$F") && echo "$F"

OUTPUT

file1

And here is awk one liner

F="file1" && awk '{if ($0 == "str1") {a=NR} else if ($0 == "str2" && a> 0 && a<NR) {print FILENAME} }' $F

F="file2" && awk '{if ($0 == "str1") {a=NR} else if ($0 == "str2" && a> 0 && a<NR) {print FILENAME} }' $F

OUTPUT

file1

This is not exactly an one-liner but you can delete the newlines and the problem is solved :)

for file in $(ls) ; do
  awk "/str1/{found=1}/str2/{if(found) print \"$file\"}" $file
done

What does it do: for each file listed by ls, if str1 appears in it, the script marks it in a variable found:

/str1/{found=1}

then, when str2 appears in a line, it verifies if found is set. If so, prints the file name:

/str2/{
    if (found) 
        print "$file"
}

EDIT: there is still a more concise way to solve your problem, using find and xargs:

find . -print0 -maxdepth 1 | \
    xargs -0 -I{} awk '/str1/{found=1}/str2/{if(found) print "{}"}' "{}"

It is safer, too, because it handles files with spaces in their names. Also, you can extend it to search in subdirectories just removing the -maxdepth 1 option. Note that the awk script was not changed.

(There always is a good solution using find and xargs but this solution is always a bit hard to find :D )

HTH!