package main

import (
	"bufio"
	"fmt"
	"os"
	"sort"
	"strconv"
)

var sc = bufio.NewScanner(os.Stdin)
var wr = bufio.NewWriter(os.Stdout)

func out(x ...interface{}) {
	fmt.Fprintln(wr, x...)
}

func getI() int {
	sc.Scan()
	i, e := strconv.Atoi(sc.Text())
	if e != nil {
		panic(e)
	}
	return i
}

func getF() float64 {
	sc.Scan()
	i, e := strconv.ParseFloat(sc.Text(), 64)
	if e != nil {
		panic(e)
	}
	return i
}

func getInts(N int) []int {
	ret := make([]int, N)
	for i := 0; i < N; i++ {
		ret[i] = getI()
	}
	return ret
}

func getS() string {
	sc.Scan()
	return sc.Text()
}

// min, max, asub, absなど基本関数
func max(a, b int) int {
	if a > b {
		return a
	}
	return b
}

func min(a, b int) int {
	if a < b {
		return a
	}
	return b
}

func asub(a, b int) int {
	if a > b {
		return a - b
	}
	return b - a
}

func abs(a int) int {
	if a >= 0 {
		return a
	}
	return -a
}

func lowerBound(a []int, x int) int {
	idx := sort.Search(len(a), func(i int) bool {
		return a[i] >= x
	})
	return idx
}

func upperBound(a []int, x int) int {
	idx := sort.Search(len(a), func(i int) bool {
		return a[i] > x
	})
	return idx
}

type pair struct {
	n, p int
}

var node [][]int
var N, M, Q int

func main() {
	defer wr.Flush()
	sc.Split(bufio.ScanWords)
	sc.Buffer([]byte{}, 1000000)
	// this template is new version.
	// use getI(), getS(), getInts(), getF()
	N, M, Q = getI(), getI(), getI()
	uf := Dsu(N + 1)
	node = make([][]int, N+1)

	// make joint
	for i := 0; i < M; i++ {
		u, v := getI()-1, getI()-1
		uf.Merge(u, v)
		node[u] = append(node[u], v)
		node[v] = append(node[v], u)
	}
	// count town
	n := make([]int, N)
	a := make([]int, Q)
	b := make([]int, Q)
	c := make([]int, Q)
	for i := 0; i < Q; i++ {
		a[i], b[i] = getI()-1, getI()-1
		if !uf.Same(a[i], b[i]) {
			n[a[i]]++
			n[b[i]]++
			c[i] = 2
		}
	}

	x := make([]pair, N)
	for i := 0; i < N; i++ {
		x[i] = pair{n[i], i}
	}
	sort.Slice(x, func(i, j int) bool {
		return x[i].n > x[j].n
	})
	// out(uf.Groups())
	for i := 0; i < N; i++ {
		u := x[i].p
		v := N
		if !uf.Same(u, v) {
			uf.Merge(u, v)
			node[u] = append(node[u], v)
			node[v] = append(node[v], u)
		}
	}
	// out(uf.Groups())
	// out(c)
	// out(x)
	l := newLCA(N, N+1, node)
	dist = make([]int, N+1)
	dfs2(N, -1, 0)
	// out(dist)
	ans := 0
	for i := 0; i < Q; i++ {
		r := l.lca(a[i], b[i])
		d := dist[a[i]] + dist[b[i]] - dist[r]*2 - c[i]
		ans += d
		// out("a,b", a[i], b[i], ":", r, d, c[i])
	}
	out(ans)
}

var dist []int

func dfs2(c, p, k int) {
	dist[c] = k
	for _, e := range node[c] {
		if e == p {
			continue
		}
		dfs2(e, c, k+1)
	}
}

/*
Lowest Common Ansestor
*/
type lca struct {
	G     [][]int
	vs    []int
	depth []int
	id    []int
	n     int
	k     int
	seg   *SegmentTree
}

const inf = int(1e16)

func newLCA(root, n int, node [][]int) *lca {
	var l lca
	l.n = n
	l.G = node
	l.vs = make([]int, n*2-1)
	l.depth = make([]int, n*2-1)
	l.id = make([]int, n)
	l.k = 0
	l.dfs(root, -1, 0)

	l.seg = SegtreeInit(n*2-1, Data{inf, 0})
	for i, e := range l.depth {
		l.seg.Set(i, Data{e, i})
	}
	l.seg.Update()
	return &l
}

func (l *lca) lca(u, v int) int {
	idx := l.seg.Query(min(l.id[u], l.id[v]), max(l.id[u], l.id[v]))
	return l.vs[idx.idx]
}

func (l *lca) dfs(v, p, d int) {
	l.id[v] = l.k
	l.vs[l.k] = v
	l.depth[l.k] = d
	l.k++
	// out(v, p, l.G[v], l.k)
	for i := 0; i < len(l.G[v]); i++ {
		if l.G[v][i] != p {
			l.dfs(l.G[v][i], v, d+1)
			l.vs[l.k] = v
			l.depth[l.k] = d
			l.k++
		}
	}
}

/*
  セグメント木(2020.05.24作成)
	SegtreeInitで初期化
	Setで値設定し、Updateで木作成
	Getで値を取得
	UpdateAtで個別のアイテムを更新
	Queryで区間の値を取得
	compareで比較方法変更可能
*/

// Data :
// Data型をstructすれば複数データが持てる
// compareも変更すること！
type Data struct {
	val, idx int
}

// SegmentTree :
type SegmentTree struct {
	inf    Data
	d      []Data
	offset int
}

// SegtreeInit :　nが要素数、valが初期値
func SegtreeInit(n int, val Data) *SegmentTree {
	var ret SegmentTree
	size := 1
	for size < n {
		size *= 2
	}
	ret.d = make([]Data, size*2)
	for i := 1; i < size*2; i++ {
		ret.d[i] = val
	}
	ret.offset = size
	ret.inf = val
	return &ret
}

// Set : 要素に値をセット（※木は更新されない）
func (s *SegmentTree) Set(idx int, val Data) {
	s.d[s.offset+idx] = val
}

// Get : 要素に値を取得
func (s *SegmentTree) Get(idx int) Data {
	return s.d[s.offset+idx]
}

// Update :
func (s *SegmentTree) Update() {
	N := s.offset
	off := s.offset

	for N > 1 {
		for i := off; i < off+N; i += 2 {
			p := i / 2
			l := i
			r := i + 1
			s.d[p] = s.compare(s.d[l], s.d[r])
		}
		off /= 2
		N /= 2
	}
}

// querySub :
// a, b ... 範囲
func (s *SegmentTree) querySub(a, b, k, l, r int) Data {
	if r <= a || b <= l {
		return s.inf
	}
	if a <= l && r <= b {
		return s.d[k]
	}
	return s.compare(
		s.querySub(a, b, k*2, l, (l+r)/2),
		s.querySub(a, b, k*2+1, (l+r)/2, r))
}

// Query :
// a, b ... 範囲 a <= x < bの範囲で検索
// [a, b)となっているのに注意
func (s *SegmentTree) Query(a, b int) Data {
	return s.querySub(a, b, 1, 0, s.offset)
}

// UpdateAt :
func (s *SegmentTree) UpdateAt(n int, val Data) {
	pos := s.offset + n
	s.d[pos] = val
	for pos > 1 {
		p := pos / 2
		l := p * 2
		r := p*2 + 1
		s.d[p] = s.compare(s.d[l], s.d[r])
		pos /= 2
	}
}

// compare :
// 比較関数（ここで比較方法を設定）
// ※min,maxを入れ替えるときなどは、Initの設定注意
func (s *SegmentTree) compare(l, r Data) Data {
	// 区間の合計の場合はinitを0にして下記
	// return l + r

	// 区間のminの場合はinfに最大値以上を設定して下記
	if l.val < r.val {
		return l
	}
	return r
}

//
// Disjoint Set Union: Union Find Tree
//

// DSU :
type DSU struct {
	parentOrSize []int
	n            int
}

// Dsu :
func Dsu(n int) *DSU {
	var d DSU
	d.n = n
	d.parentOrSize = make([]int, n)
	for i := 0; i < n; i++ {
		d.parentOrSize[i] = -1
	}
	return &d
}

// Merge :
func (d DSU) Merge(a, b int) int {
	x, y := d.Leader(a), d.Leader(b)
	if x == y {
		return x
	}
	if -d.parentOrSize[x] < -d.parentOrSize[y] {
		x, y = y, x
	}
	d.parentOrSize[x] += d.parentOrSize[y]
	d.parentOrSize[y] = x
	return x
}

// Same :
func (d DSU) Same(a, b int) bool {
	return d.Leader(a) == d.Leader(b)
}

// Leader :
func (d DSU) Leader(a int) int {
	if d.parentOrSize[a] < 0 {
		return a
	}
	d.parentOrSize[a] = d.Leader(d.parentOrSize[a])
	return d.parentOrSize[a]
}

// Size :
func (d DSU) Size(a int) int {
	return -d.parentOrSize[d.Leader(a)]
}

// Groups : original implement
func (d DSU) Groups() [][]int {
	m := make(map[int][]int)
	for i := 0; i < d.n; i++ {
		x := d.Leader(i)
		if x < 0 {
			m[i] = append(m[i], i)
		} else {
			m[x] = append(m[x], i)
		}
	}
	ret := make([][]int, len(m))
	idx := 0
	for _, e := range m {
		ret[idx] = make([]int, len(e))
		copy(ret[idx], e)
		idx++
	}
	return ret
}