The calculateDelta method in packages/cli/src/cli/utils/delta.ts uses a nested loop to detect renamed keys - for each
added key, it linearly scans ALL removed keys to find a matching content hash.

With large projects this scales quadratically:

• 100 renames → 10,000 comparisons
• 1,000 renames → 1,000,000 comparisons
• 5,000 renames → 25,000,000 comparisons

The loop at lines 62-70:

for (const addedKey of added) {
  const addedHash = md5(params.sourceData[addedKey]);
  for (const removedKey of removed) {
    if (params.checksums[removedKey] === addedHash) {
      renamed.push([removedKey, addedKey]);
      break;
    }
  }
}

This could be replaced with a Map<hash, removedKey> pre-built from the removed keys, turning each lookup from O(n) to
O(1). Overall complexity drops from O(n×m) to O(n+m).

This affects any project that renames a large number of translation keys at once (e.g. refactoring key naming
conventions).