Filter cloc result for a certain folder and get statistic information

cloc_folder.bash

#!/bin/bash
for folder in `ls -l | grep "^d" | awk '{print $NF}'`;do 
    echo "counting lines for folder $folder" >> cloc.txt; 
    cloc.pl $folder >> cloc.txt;
done

filterCloc.awk cloc.txt > cloc_summary.txt

filterCloc.awk

#!/bin/awk -f
/counting lines for folder/{
    #print "current domain is", current_domain
    current_domain = $NF
}

/^C\+\+ /{
    code_stats[current_domain, "c++"] = $NF
}

/^C /{
    code_stats[current_domain, "c source"] = $NF
}

/^C\/C\+\+ Header /{
    code_stats[current_domain, "header"] = $NF
}

/^(make|CMake) /{
    code_stats[current_domain, "make"] += $NF
}

END{
    printf("%-40s %-10s %10s\n", "domain", "type", "loc");

    previous_domain = "";
    domain_total = 0;
    all_total = 0;

    n = asorti(code_stats, new_stat)
    for (i = 1; i <= n; i++){
        detail = new_stat[i]
        split(detail, indexes, SUBSEP);
        domain = indexes[1]
        type = indexes[2]
        loc = code_stats[detail]

        if (domain != previous_domain){
            printf("%-40s %-10s %10d\n", previous_domain, "total", domain_total);
            printf("------------------------------------------------------------------\n");
            all_total = all_total + domain_total
            domain_total = 0;
        }

        printf("%-40s %-10s %10d\n", domain, type, loc);
        domain_total = domain_total + loc
        previous_domain = domain
    }

    printf("%-40s %-10s %10d\n", previous_domain, "total", domain_total);
    printf("------------------------------------------------------------------\n");
    printf("%-40s %-10s %10d\n", "ALL DOMAINS", "total", all_total);
}

Key points:

• Array sort by index : asorti
• Multi-dimension array simulation
• cloc tool to count loc

cloc is a famous tool to count line of codes for SCM, its official site can be found here