How to join each double lines?

I have a text file,

a1
a2
b1
b2
c1
c2
...

I want to join by two lines so one can sort it:

a1:a2
b1:b2
c开发者_开发技巧1:c2
...

I'm using bash. the read function will eat the leading space, which is undesired. And I hate to write simple stupid C programs.

Then, I can use tr : "\n" to split the joined file back to two files.

paste -s -d ':\n' file should do it.

For example:

% cat f
a1
a2
b1
b2
% paste -s -d ':\n' f
a1:a2
b1:b2

sed 'N;s/\n/:/;' < srcfile > destfile

Looking over there I found an example that can be converted into:

sed '$!N;s/\n/:/' < file

INDEX=0
A=""
B=""

for i in `awk '{print $1}' input`
    do
    if [ $INDEX -eq 0 ]; then
        A=$i;
        let INDEX=1;
    fi

    if [ $INDEX -eq 1 ]; then
        B=$i;
        echo $A:$B
        let INDEX=0;
    fi
done

awk '{line=$0; printf line; if (getline) printf ":" $0; print ""}' inputfile

edit in-place with backup:

perl -i.bak -pe 's/\n\Z/:/ if $.%2' file

edit in-place no backup:

perl -i -pe 's/\n\Z/:/ if $.%2' file

Here is a solution in python:

#!/usr/bin/python3

def njoin(filename, outfn="", n=3, linesuffix=" "):
    if not outfn:
        outfn = filename + ".join"
    with open(filename) as infh, open(outfn, "w") as outfh:
        nline = 0
        for line in infh:
            if nline % n != n-1:
                line = line.rstrip() + linesuffix
            outfh.write(line)
            nline += 1

In you case you can use the function like this:

njoin("/path/to/file", n=2, linesuffixe=":")