| 1 |
|
package prosediff |
| 2 |
|
|
| 3 |
|
// Myers O(ND) sequence diff, used at three levels: blocks, words and code |
| 4 |
|
// lines. Sequences are always interned to []int first, so equality is exact |
| 5 |
|
// integer equality and there is no hash-collision risk. |
| 6 |
|
|
| 7 |
|
// Op is the role of a run in an edit script. |
| 8 |
|
type Op uint8 |
| 9 |
|
|
| 10 |
|
const ( |
| 11 |
|
// OpEqual marks content present, unchanged, in both revisions. |
| 12 |
|
OpEqual Op = iota |
| 13 |
|
// OpDelete marks content present only in the old revision. |
| 14 |
|
OpDelete |
| 15 |
|
// OpInsert marks content present only in the new revision. |
| 16 |
|
OpInsert |
| 17 |
|
) |
| 18 |
|
|
| 19 |
0 |
func (o Op) String() string { |
| 20 |
0 |
switch o { |
| 21 |
0 |
case OpEqual: |
| 22 |
0 |
return "equal" |
| 23 |
0 |
case OpDelete: |
| 24 |
0 |
return "delete" |
| 25 |
0 |
case OpInsert: |
| 26 |
0 |
return "insert" |
| 27 |
|
} |
| 28 |
0 |
return "unknown" |
| 29 |
|
} |
| 30 |
|
|
| 31 |
|
// edit is one run of an edit script: n consecutive elements sharing an Op. |
| 32 |
|
type edit struct { |
| 33 |
|
op Op |
| 34 |
|
n int |
| 35 |
|
} |
| 36 |
|
|
| 37 |
|
// interner maps arbitrary strings to dense ints so sequence comparison is |
| 38 |
|
// integer equality. |
| 39 |
|
type interner struct { |
| 40 |
|
ids map[string]int |
| 41 |
|
} |
| 42 |
|
|
| 43 |
101 |
func newInterner() *interner { return &interner{ids: make(map[string]int)} } |
| 44 |
|
|
| 45 |
1118 |
func (in *interner) id(s string) int { |
| 46 |
1118 |
if v, ok := in.ids[s]; ok { |
| 47 |
484 |
return v |
| 48 |
484 |
} |
| 49 |
634 |
v := len(in.ids) + 1 |
| 50 |
634 |
in.ids[s] = v |
| 51 |
634 |
return v |
| 52 |
|
} |
| 53 |
|
|
| 54 |
138 |
func (in *interner) all(ss []string) []int { |
| 55 |
138 |
out := make([]int, len(ss)) |
| 56 |
954 |
for i, s := range ss { |
| 57 |
954 |
out[i] = in.id(s) |
| 58 |
954 |
} |
| 59 |
138 |
return out |
| 60 |
|
} |
| 61 |
|
|
| 62 |
|
// diffInts returns the edit script turning a into b, as a coalesced run list. |
| 63 |
|
// The script is a minimal edit script (Myers), with common prefix and suffix |
| 64 |
|
// trimmed first so the expensive part only runs over the changed middle. |
| 65 |
512 |
func diffInts(a, b []int) []edit { |
| 66 |
512 |
pre := 0 |
| 67 |
512 |
for pre < len(a) && pre < len(b) && a[pre] == b[pre] { |
| 68 |
395 |
pre++ |
| 69 |
395 |
} |
| 70 |
512 |
suf := 0 |
| 71 |
512 |
for suf < len(a)-pre && suf < len(b)-pre && a[len(a)-1-suf] == b[len(b)-1-suf] { |
| 72 |
210 |
suf++ |
| 73 |
210 |
} |
| 74 |
512 |
var out []edit |
| 75 |
512 |
if pre > 0 { |
| 76 |
147 |
out = append(out, edit{OpEqual, pre}) |
| 77 |
147 |
} |
| 78 |
512 |
out = append(out, myers(a[pre:len(a)-suf], b[pre:len(b)-suf])...) |
| 79 |
512 |
if suf > 0 { |
| 80 |
130 |
out = append(out, edit{OpEqual, suf}) |
| 81 |
130 |
} |
| 82 |
512 |
return coalesce(out) |
| 83 |
|
} |
| 84 |
|
|
| 85 |
|
// vsnap is the V array of one Myers iteration, stored only over the diagonals |
| 86 |
|
// [-d, d] that iteration can reach. |
| 87 |
|
type vsnap struct { |
| 88 |
|
d int |
| 89 |
|
vals []int32 |
| 90 |
|
} |
| 91 |
|
|
| 92 |
6925 |
func (s vsnap) get(k int) int { return int(s.vals[k+s.d]) } |
| 93 |
|
|
| 94 |
512 |
func myers(a, b []int) []edit { |
| 95 |
512 |
n, m := len(a), len(b) |
| 96 |
512 |
switch { |
| 97 |
14 |
case n == 0 && m == 0: |
| 98 |
14 |
return nil |
| 99 |
44 |
case n == 0: |
| 100 |
44 |
return []edit{{OpInsert, m}} |
| 101 |
35 |
case m == 0: |
| 102 |
35 |
return []edit{{OpDelete, n}} |
| 103 |
|
} |
| 104 |
|
|
| 105 |
419 |
maxD := n + m |
| 106 |
419 |
v := make([]int32, 2*maxD+1) |
| 107 |
419 |
off := maxD |
| 108 |
419 |
trace := make([]vsnap, 0, 16) |
| 109 |
419 |
|
| 110 |
3560 |
for d := 0; d <= maxD; d++ { |
| 111 |
3560 |
done := false |
| 112 |
18067 |
for k := -d; k <= d; k += 2 { |
| 113 |
18067 |
var x int |
| 114 |
18067 |
if k == -d || (k != d && v[off+k-1] < v[off+k+1]) { |
| 115 |
13289 |
x = int(v[off+k+1]) // move down: consume one element of b |
| 116 |
13289 |
} else { |
| 117 |
4778 |
x = int(v[off+k-1]) + 1 // move right: consume one element of a |
| 118 |
4778 |
} |
| 119 |
18067 |
y := x - k |
| 120 |
18067 |
for x < n && y < m && a[x] == b[y] { |
| 121 |
1604 |
x++ |
| 122 |
1604 |
y++ |
| 123 |
1604 |
} |
| 124 |
18067 |
v[off+k] = int32(x) |
| 125 |
18067 |
if x >= n && y >= m { |
| 126 |
419 |
done = true |
| 127 |
419 |
break |
| 128 |
|
} |
| 129 |
|
} |
| 130 |
3560 |
snap := vsnap{d: d, vals: make([]int32, 2*d+1)} |
| 131 |
19668 |
for k := -d; k <= d; k += 2 { |
| 132 |
19668 |
snap.vals[k+d] = v[off+k] |
| 133 |
19668 |
} |
| 134 |
3560 |
trace = append(trace, snap) |
| 135 |
3560 |
if done { |
| 136 |
419 |
return backtrack(trace, n, m) |
| 137 |
419 |
} |
| 138 |
|
} |
| 139 |
0 |
panic("prosediff: myers did not converge") |
| 140 |
|
} |
| 141 |
|
|
| 142 |
|
// backtrack walks the saved V arrays from the end point back to the origin, |
| 143 |
|
// emitting the edit script in reverse and then flipping it. |
| 144 |
419 |
func backtrack(trace []vsnap, n, m int) []edit { |
| 145 |
419 |
var rev []edit |
| 146 |
4027 |
push := func(op Op) { |
| 147 |
4027 |
if len(rev) > 0 && rev[len(rev)-1].op == op { |
| 148 |
1770 |
rev[len(rev)-1].n++ |
| 149 |
1770 |
return |
| 150 |
1770 |
} |
| 151 |
2257 |
rev = append(rev, edit{op, 1}) |
| 152 |
|
} |
| 153 |
|
|
| 154 |
419 |
x, y := n, m |
| 155 |
3141 |
for d := len(trace) - 1; d > 0; d-- { |
| 156 |
3141 |
prev := trace[d-1] |
| 157 |
3141 |
k := x - y |
| 158 |
3141 |
var prevK int |
| 159 |
3141 |
if k == -d || (k != d && prev.get(k-1) < prev.get(k+1)) { |
| 160 |
1601 |
prevK = k + 1 |
| 161 |
1601 |
} else { |
| 162 |
1540 |
prevK = k - 1 |
| 163 |
1540 |
} |
| 164 |
3141 |
prevX := prev.get(prevK) |
| 165 |
3141 |
prevY := prevX - prevK |
| 166 |
3141 |
for x > prevX && y > prevY { |
| 167 |
886 |
push(OpEqual) |
| 168 |
886 |
x-- |
| 169 |
886 |
y-- |
| 170 |
886 |
} |
| 171 |
3141 |
if x == prevX { |
| 172 |
1601 |
push(OpInsert) |
| 173 |
1601 |
y-- |
| 174 |
1601 |
} else { |
| 175 |
1540 |
push(OpDelete) |
| 176 |
1540 |
x-- |
| 177 |
1540 |
} |
| 178 |
3141 |
x, y = prevX, prevY |
| 179 |
|
} |
| 180 |
419 |
for x > 0 && y > 0 { |
| 181 |
0 |
push(OpEqual) |
| 182 |
0 |
x-- |
| 183 |
0 |
y-- |
| 184 |
0 |
} |
| 185 |
|
// d == 0 leaves at most one of x, y non-zero only when the other |
| 186 |
|
// sequence was fully consumed on the diagonal, which the prefix trim |
| 187 |
|
// already handled; keep the guard rather than assume. |
| 188 |
419 |
for ; x > 0; x-- { |
| 189 |
0 |
push(OpDelete) |
| 190 |
0 |
} |
| 191 |
419 |
for ; y > 0; y-- { |
| 192 |
0 |
push(OpInsert) |
| 193 |
0 |
} |
| 194 |
|
|
| 195 |
419 |
out := make([]edit, 0, len(rev)) |
| 196 |
2257 |
for i := len(rev) - 1; i >= 0; i-- { |
| 197 |
2257 |
out = append(out, rev[i]) |
| 198 |
2257 |
} |
| 199 |
419 |
return coalesce(out) |
| 200 |
|
} |
| 201 |
|
|
| 202 |
948 |
func coalesce(in []edit) []edit { |
| 203 |
948 |
out := in[:0:0] |
| 204 |
4940 |
for _, e := range in { |
| 205 |
4940 |
if e.n == 0 { |
| 206 |
0 |
continue |
| 207 |
|
} |
| 208 |
4940 |
if len(out) > 0 && out[len(out)-1].op == e.op { |
| 209 |
0 |
out[len(out)-1].n += e.n |
| 210 |
0 |
continue |
| 211 |
|
} |
| 212 |
4940 |
out = append(out, e) |
| 213 |
|
} |
| 214 |
948 |
return out |
| 215 |
|
} |
| 216 |
|
|
| 217 |
|
// commonCount reports how many elements the edit script keeps equal. |
| 218 |
443 |
func commonCount(script []edit) int { |
| 219 |
443 |
n := 0 |
| 220 |
2420 |
for _, e := range script { |
| 221 |
2420 |
if e.op == OpEqual { |
| 222 |
916 |
n += e.n |
| 223 |
916 |
} |
| 224 |
|
} |
| 225 |
443 |
return n |
| 226 |
|
} |
| 227 |
|
|
| 228 |
|
// ratio is the classic 2*common/(len(a)+len(b)) similarity in [0,1]. |
| 229 |
44 |
func ratio(a, b []int) float64 { |
| 230 |
44 |
if len(a) == 0 && len(b) == 0 { |
| 231 |
1 |
return 1 |
| 232 |
1 |
} |
| 233 |
43 |
return 2 * float64(commonCount(diffInts(a, b))) / float64(len(a)+len(b)) |
| 234 |
|
} |