| 1 | // Copyright 2022 The Go Authors. All rights reserved. |
|---|---|
| 2 | // Use of this source code is governed by a BSD-style |
| 3 | // license that can be found in the LICENSE file. |
| 4 | |
| 5 | package diff |
| 6 | |
| 7 | import ( |
| 8 | "bytes" |
| 9 | "unicode/utf8" |
| 10 | |
| 11 | "golang.org/x/tools/internal/diff/lcs" |
| 12 | ) |
| 13 | |
| 14 | // Strings computes the differences between two strings. |
| 15 | // The resulting edits respect rune boundaries. |
| 16 | func Strings(before, after string) []Edit { |
| 17 | if before == after { |
| 18 | return nil // common case |
| 19 | } |
| 20 | |
| 21 | if stringIsASCII(before) && stringIsASCII(after) { |
| 22 | // TODO(adonovan): opt: specialize diffASCII for strings. |
| 23 | return diffASCII([]byte(before), []byte(after)) |
| 24 | } |
| 25 | return diffRunes([]rune(before), []rune(after)) |
| 26 | } |
| 27 | |
| 28 | // Bytes computes the differences between two byte slices. |
| 29 | // The resulting edits respect rune boundaries. |
| 30 | func Bytes(before, after []byte) []Edit { |
| 31 | if bytes.Equal(before, after) { |
| 32 | return nil // common case |
| 33 | } |
| 34 | |
| 35 | if bytesIsASCII(before) && bytesIsASCII(after) { |
| 36 | return diffASCII(before, after) |
| 37 | } |
| 38 | return diffRunes(runes(before), runes(after)) |
| 39 | } |
| 40 | |
| 41 | func diffASCII(before, after []byte) []Edit { |
| 42 | diffs := lcs.DiffBytes(before, after) |
| 43 | |
| 44 | // Convert from LCS diffs. |
| 45 | res := make([]Edit, len(diffs)) |
| 46 | for i, d := range diffs { |
| 47 | res[i] = Edit{d.Start, d.End, string(after[d.ReplStart:d.ReplEnd])} |
| 48 | } |
| 49 | return res |
| 50 | } |
| 51 | |
| 52 | func diffRunes(before, after []rune) []Edit { |
| 53 | diffs := lcs.DiffRunes(before, after) |
| 54 | |
| 55 | // The diffs returned by the lcs package use indexes |
| 56 | // into whatever slice was passed in. |
| 57 | // Convert rune offsets to byte offsets. |
| 58 | res := make([]Edit, len(diffs)) |
| 59 | lastEnd := 0 |
| 60 | utf8Len := 0 |
| 61 | for i, d := range diffs { |
| 62 | utf8Len += runesLen(before[lastEnd:d.Start]) // text between edits |
| 63 | start := utf8Len |
| 64 | utf8Len += runesLen(before[d.Start:d.End]) // text deleted by this edit |
| 65 | res[i] = Edit{start, utf8Len, string(after[d.ReplStart:d.ReplEnd])} |
| 66 | lastEnd = d.End |
| 67 | } |
| 68 | return res |
| 69 | } |
| 70 | |
| 71 | // runes is like []rune(string(bytes)) without the duplicate allocation. |
| 72 | func runes(bytes []byte) []rune { |
| 73 | n := utf8.RuneCount(bytes) |
| 74 | runes := make([]rune, n) |
| 75 | for i := 0; i < n; i++ { |
| 76 | r, sz := utf8.DecodeRune(bytes) |
| 77 | bytes = bytes[sz:] |
| 78 | runes[i] = r |
| 79 | } |
| 80 | return runes |
| 81 | } |
| 82 | |
| 83 | // runesLen returns the length in bytes of the UTF-8 encoding of runes. |
| 84 | func runesLen(runes []rune) (len int) { |
| 85 | for _, r := range runes { |
| 86 | len += utf8.RuneLen(r) |
| 87 | } |
| 88 | return len |
| 89 | } |
| 90 | |
| 91 | // stringIsASCII reports whether s contains only ASCII. |
| 92 | // TODO(adonovan): combine when x/tools allows generics. |
| 93 | func stringIsASCII(s string) bool { |
| 94 | for i := 0; i < len(s); i++ { |
| 95 | if s[i] >= utf8.RuneSelf { |
| 96 | return false |
| 97 | } |
| 98 | } |
| 99 | return true |
| 100 | } |
| 101 | |
| 102 | func bytesIsASCII(s []byte) bool { |
| 103 | for i := 0; i < len(s); i++ { |
| 104 | if s[i] >= utf8.RuneSelf { |
| 105 | return false |
| 106 | } |
| 107 | } |
| 108 | return true |
| 109 | } |
| 110 |
Members