Separating sublists with colon

I am new to linux and am trying to parse a bunch of files that looks as follows -

• Some text
  • start list some other text
    • start sublist1
    • continue sublist1
  • more elements
  • more elements2
    • a sublist2
      • a sub-sublist1

Where all the spaces before the list are tabs. I need a way to parse the text so that a colon is added for sublists... so that it looks like the following at the end:

• Some text:
  • start list some other text:
    • start sublist1
    • continue sublist1
  • more elements
  • more elements2:
    • a sublist2:
      • a sub-sublist1
  • another element

So colons 开发者_StackOverflow中文版are added only when there is a sublist available.

I tried looking into sed and awk commands but I am was unable to find anything that stored the status of the previous line to be able to add the colon at the end. It does not have to be done in sed or awk, I have been trying these though and no luck. Any suggestions at all would help.

Somthing like that sould solve your problem:

awk '
    function countTabs(line) {
        tabs=0;
        i=0;
        while( substr(line,i++,1) == "\t")
            tabs++;
        return tabs;
     }
{
    line1 = $0;
    while( getline line2) {
        if ( countTabs(line1) < countTabs(line2))
           printf("%s:\n" , line1);
        else
           printf("%s\n",line1);
        line1 = line2;
    }
    print line2;
}'

something to try

awk '
{
    A[d++]=$0
    match($0,"[^[:blank:]]")
    if ( RSTART > t ){    A[d-1]=A[d-1]":"  }
    else{  gsub(/:$/,"",A[d-2])  }
    t=RSTART
}
END{
    for(i=0;i<=d;i++){
        print A[i]
    }
} ' file

output

$ cat file
Some text
        start list some other text
                start sublist1
                continue sublist1
        more elements
        more elements2
                a sublist2
                        a sub-sublist1
                                a sub-sublist2
        another element

$ ./shell.sh
Some text:
        start list some other text:
                start sublist1
                continue sublist1
        more elements
        more elements2
                a sublist2:
                        a sub-sublist1:
                                a sub-sublist2
        another element

This modified version of ghostdog74's script should get the job done:

awk '
{
    A[NR]=$0
    match($0,"[^[:blank:]]")
    if ( RSTART > t ){ A[NR-1]=A[NR-1]":" }
    t=RSTART
}
END{
    for(i=1; i<=NR+1; i++){
        print A[i]
    }
} ' file