Fewest Deletions for Matching Anagram

You are given two strings `source` and `target`. In one operation you may delete any single character from either string. Return the minimum total number of deletions required to make `source` and `target` **anagrams** of each other (i.e. they contain exactly the same multiset of characters).

Characters that remain in both strings after deletions must match in frequency.

**Example 1:**
```
Input: source = "kindly", target = "killing"
Output: 4
Explanation: Keep 'k','i','l' in each (3 characters each).
source deletions: 'n','d','y' (3 deletions)
target deletions: second 'l','i','n','g' — wait, target "killing" keeps k,i,l,l,i,n,g.
Actually keep common: k(1),i(1),l(1) → delete 'n','d','y' from source (3) and delete 'l','i','n','g' from target (4) → total 7?
Re-examine: freq(source)={k:1,i:1,n:1,d:1,l:1,y:1}, freq(target)={k:1,i:2,l:2,n:1,g:1}.
Overlap per char: k→1,i→1,l→1,n→1 = 4 kept. source has 6 chars, target has 7 chars. Deletions = (6-4)+(7-4) = 2+3 = 5.
Output: 5
```

**Example 2:**
```
Input: source = "broom", target = "motor"
Output: 4
Explanation: freq(source)={b:1,r:1,o:2,m:1}, freq(target)={m:1,o:1,t:1,r:1}.
Overlap: m→1,o→1,r→1 = 3 kept. Deletions = (5-3)+(5-3) = 4.
```