regex - How do I split an integer using an "if" condition? -

i've been looking days trying use awk, sed, cut, , tr possible solution problem. have data set thats delimited "@" shown below...

1@2@11@11/8@11/8@11@11/2 2@4@31 1/2@31 1/2@31/2@21@21/2 3@10@116 1/4@98@911 3/4@410@38 1/2 4@1@21@21/8@21/8@33@49 1/4 5@11@74@75@67 1/2@511 1/2@511 1/2 6@9@106@108 1/4@89 1/4@613 1/2@616 7@7@96@118 1/4@1313 1/2@715@717 3/4 8@12@127 3/4@129 3/4@1212 1/2@816 1/2@817 3/4 9@6@63@ 64 1/2@79@916 1/2@918 10@13@139 3/4@1311 1/4@1112@1017@1019 3/4 11@3@42@42@43 1/2@1118 1/2@1126 1/4 12@5@84 1/2@87@1011 3/4@1219 1/2@1228 1/4 13@8@52 1/2@53 1/2@57@1324@1332 3/4 

what split row numbers in first column (the ranks) rest of integers in other columns starting @ column 3 , on. end results this...

1@2@1@1@1@1/8@1@1/8@1@1@1@1/2 2@4@3@1 1/2@3@1 1/2@3@1/2@2@1@2@1/2 3@10@11@6 1/4@9@8@9@11 3/4@4@10@3@8 1/2 4@1@2@1@2@1/8@2@1/8@3@3@4@9 1/4 5@11@7@4@7@5@6@7 1/2@5@11 1/2@5@11 1/2 6@9@10@6@10@8 1/4@8@9 1/4@6@13 1/2@6@16 7@7@9@6@11@8 1/4@13@13 1/2@7@15@7@17 3/4 8@12@12@7 3/4@12@9 3/4@12@12 1/2@8@16 1/2@8@17 3/4 9@6@6@3@6@4 1/2@7@9@9@16 1/2@9@18 10@13@13@9 3/4@13@11 1/4@11@12@10@17@10@19 3/4 11@3@4@2@4@2@4@3 1/2@11@18 1/2@11@26 1/4 12@5@8@4 1/2@8@7@10@11 3/4@12@19 1/2@12@28 1/4 13@8@5@2 1/2@5@3 1/2@5@7@13@24@13@32 3/4 

i thinking use "if statement". "if integers start [2-9] split after 1 character, elif starts [1] , length equal 3 or more (before space , fraction) split firsts 2 characters." have no idea how how go solving problem. have thousands of similar files , need change structure of them, solution have ran through loop.

here's fun one:

perl -f@ -lape '$_ = join "@", shift(@f), shift(@f), map {s/(1\d|\d)(\d+)/$1\@$2/g; $_} @f' file 

with little commentary

perl -f@ -lape '     $_ = join "@",                # join following things, using "@"                shift(@f),          #   first field               shift(@f),          #   second field               map {               #   then, transform rest expr                   s{              #     search for:                       (1\d | \d)  #       1 plus digit, or digit                       (\d+)       #       followed 1 or more digits                    }{$1\@$2}xg;   #     add "@" in between                   $_              #     , return new string               } @f ' file 

the options:

• -a , -f@ -- split each line array @f using @ character separator
• -l -- handle line endings automatically
• -p -- automatically print variable $_ after processing each line