The following finding is a result of a 5 minute brainstorm we did with Alex Denisov: we were thinking about the fastest way to estimate Swift vs Objective-C code ratio in an iOS application. Transition to Swift can be very slow in a project which has a lot of Objective-C code so we want to have at least a rough understanding of how much of its code has to be migrated from Objective-C to Swift (in case if Apple discontinues support of Objective-C as of April 2017.).

We came up with a very simple heuristic: grep through the codebase and find @implementation and ^class declarations for Objective-C and Swift declarations respectively and divide Swift value by Objective-C value. Of course this heuristic is very rough however we can make enough sense out of it in the context of the app we are currently working on.

I have split original one-liner to be a more readable .sh script:


objc=`grep -r "^@implementation " $project  | wc -l | tr -d ' '`;
swift=`grep -r "^class " $project  | wc -l | tr -d ' '`;

ratio=$(bc <<< "scale=3; $swift / $objc");
ratio_perc=$(bc <<< "$ratio * 100");

echo "Swift classes: $swift";
echo "Objective-C classes: $objc";
printf "Ratio: %0.3g%%\n" $ratio_perc;
$ ./
Swift classes: 142
Objective-C classes: 749
Ratio: 18.9%

This script is very simple however I would love to have found it written by someone else before I started looking for it a few hours ago today. So it is me who is sharing this to the internet.