aboutsummaryrefslogtreecommitdiff
path: root/telegram/formatter/formatter.go
blob: d6541cba78676b9b3060a41f95a7adbfc2a52698 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
package formatter

import (
	"sort"
	"unicode"

	log "github.com/sirupsen/logrus"
	"github.com/zelenin/go-tdlib/client"
)

// Insertion is a piece of text in given position
type Insertion struct {
	Offset int32
	Runes  []rune
}

// InsertionStack contains the sequence of insertions
// from the start or from the end
type InsertionStack []*Insertion

var boldRunesMarkdown = []rune("**")
var boldRunesXEP0393 = []rune("*")
var italicRunes = []rune("_")
var strikeRunesMarkdown = []rune("~~")
var strikeRunesXEP0393 = []rune("~")
var codeRunes = []rune("`")
var preRuneStart = []rune("```\n")
var preRuneEnd = []rune("\n```")

// rebalance pumps all the values until the given offset to current stack (growing
// from start) from given stack (growing from end); should be called
// before any insertions to the current stack at the given offset
func (s InsertionStack) rebalance(s2 InsertionStack, offset int32) (InsertionStack, InsertionStack) {
	for len(s2) > 0 && s2[len(s2)-1].Offset <= offset {
		s = append(s, s2[len(s2)-1])
		s2 = s2[:len(s2)-1]
	}

	return s, s2
}

// NewIterator is a second order function that sequentially scans and returns
// stack elements; starts returning nil when elements are ended
func (s InsertionStack) NewIterator() func() *Insertion {
	i := -1

	return func() *Insertion {
		i++
		if i < len(s) {
			return s[i]
		}
		return nil
	}
}

// SortEntities arranges the entities in traversal-ready order
func SortEntities(entities []*client.TextEntity) []*client.TextEntity {
	sortedEntities := make([]*client.TextEntity, len(entities))
	copy(sortedEntities, entities)

	sort.Slice(sortedEntities, func(i int, j int) bool {
		entity1 := entities[i]
		entity2 := entities[j]
		if entity1.Offset < entity2.Offset {
			return true
		} else if entity1.Offset == entity2.Offset {
			return entity1.Length > entity2.Length
		}
		return false
	})
	return sortedEntities
}

// MergeAdjacentEntities merges entities of a same kind
func MergeAdjacentEntities(entities []*client.TextEntity) []*client.TextEntity {
	mergedEntities := make([]*client.TextEntity, 0, len(entities))
	excludedIndices := make(map[int]bool)

	for i, entity := range entities {
		if excludedIndices[i] {
			continue
		}

		typ := entity.Type.TextEntityTypeType()
		start := entity.Offset
		end := start + entity.Length
		ei := make(map[int]bool)

		// collect continuations
		for j, entity2 := range entities[i+1:] {
			if entity2.Type.TextEntityTypeType() == typ && entity2.Offset == end {
				end += entity2.Length
				ei[j+i+1] = true
			}
		}

		// check for intersections with other entities
		var isIntersecting bool
		if len(ei) > 0 {
			for _, entity2 := range entities {
				entity2End := entity2.Offset + entity2.Length
				if (entity2.Offset < start && entity2End > start && entity2End < end) ||
					(entity2.Offset > start && entity2.Offset < end && entity2End > end) {
					isIntersecting = true
					break
				}
			}
		}

		if !isIntersecting {
			entity.Length = end - start
			for j := range ei {
				excludedIndices[j] = true
			}
		}
		mergedEntities = append(mergedEntities, entity)
	}

	return mergedEntities
}

// ClaspDirectives to the following span as required by XEP-0393
func ClaspDirectives(text string, entities []*client.TextEntity) []*client.TextEntity {
	alignedEntities := make([]*client.TextEntity, len(entities))
	copy(alignedEntities, entities)

	// transform the source text into a form with uniform runes and code points,
	// by duplicating the Basic Multilingual Plane
	doubledRunes := make([]rune, 0, len(text)*2)

	for _, cp := range text {
		if cp > 0x0000ffff {
			doubledRunes = append(doubledRunes, cp, cp)
		} else {
			doubledRunes = append(doubledRunes, cp)
		}
	}
	for i, entity := range alignedEntities {
		var dirty bool
		endOffset := entity.Offset + entity.Length

		if unicode.IsSpace(doubledRunes[entity.Offset]) {
			for j, r := range doubledRunes[entity.Offset+1:endOffset] {
				if !unicode.IsSpace(r) {
					dirty = true
					entity.Offset += int32(j+1)
					entity.Length -= int32(j+1)
					break
				}
			}
		}
		if unicode.IsSpace(doubledRunes[endOffset-1]) {
			for j := endOffset-2; j >= entity.Offset; j-- {
				if !unicode.IsSpace(doubledRunes[j]) {
					dirty = true
					entity.Length = j+1-entity.Offset
					break
				}
			}
		}

		if dirty {
			alignedEntities[i] = entity
		}
	}

	return alignedEntities
}

func markupBraces(entity *client.TextEntity, lbrace, rbrace []rune) (*Insertion, *Insertion) {
	return &Insertion{
			Offset: entity.Offset,
			Runes:  lbrace,
		}, &Insertion{
			Offset: entity.Offset + entity.Length,
			Runes:  rbrace,
		}
}

// EntityToMarkdown generates the wrapping Markdown tags
func EntityToMarkdown(entity *client.TextEntity) (*Insertion, *Insertion) {
	switch entity.Type.TextEntityTypeType() {
	case client.TypeTextEntityTypeBold:
		return markupBraces(entity, boldRunesMarkdown, boldRunesMarkdown)
	case client.TypeTextEntityTypeItalic:
		return markupBraces(entity, italicRunes, italicRunes)
	case client.TypeTextEntityTypeStrikethrough:
		return markupBraces(entity, strikeRunesMarkdown, strikeRunesMarkdown)
	case client.TypeTextEntityTypeCode:
		return markupBraces(entity, codeRunes, codeRunes)
	case client.TypeTextEntityTypePre:
		return markupBraces(entity, preRuneStart, preRuneEnd)
	case client.TypeTextEntityTypePreCode:
		preCode, _ := entity.Type.(*client.TextEntityTypePreCode)
		return markupBraces(entity, []rune("\n```"+preCode.Language+"\n"), codeRunes)
	case client.TypeTextEntityTypeTextUrl:
		textURL, _ := entity.Type.(*client.TextEntityTypeTextUrl)
		return markupBraces(entity, []rune("["), []rune("]("+textURL.Url+")"))
	}

	return nil, nil
}

// EntityToXEP0393 generates the wrapping XEP-0393 tags
func EntityToXEP0393(entity *client.TextEntity) (*Insertion, *Insertion) {
	if entity == nil || entity.Type == nil {
		return nil, nil
	}

	switch entity.Type.TextEntityTypeType() {
	case client.TypeTextEntityTypeBold:
		return markupBraces(entity, boldRunesXEP0393, boldRunesXEP0393)
	case client.TypeTextEntityTypeItalic:
		return markupBraces(entity, italicRunes, italicRunes)
	case client.TypeTextEntityTypeStrikethrough:
		return markupBraces(entity, strikeRunesXEP0393, strikeRunesXEP0393)
	case client.TypeTextEntityTypeCode:
		return markupBraces(entity, codeRunes, codeRunes)
	case client.TypeTextEntityTypePre:
		return markupBraces(entity, preRuneStart, preRuneEnd)
	case client.TypeTextEntityTypePreCode:
		preCode, _ := entity.Type.(*client.TextEntityTypePreCode)
		return markupBraces(entity, []rune("\n```"+preCode.Language+"\n"), codeRunes)
	case client.TypeTextEntityTypeTextUrl:
		textURL, _ := entity.Type.(*client.TextEntityTypeTextUrl)
		// non-standard, Pidgin-specific
		return markupBraces(entity, []rune{}, []rune(" <"+textURL.Url+">"))
	}

	return nil, nil
}

// Format traverses an already sorted list of entities and wraps the text in a markup
func Format(
	sourceText string,
	entities []*client.TextEntity,
	entityToMarkup func(*client.TextEntity) (*Insertion, *Insertion),
) string {
	if len(entities) == 0 {
		return sourceText
	}

	mergedEntities := SortEntities(ClaspDirectives(sourceText, MergeAdjacentEntities(SortEntities(entities))))

	startStack := make(InsertionStack, 0, len(sourceText))
	endStack := make(InsertionStack, 0, len(sourceText))

	// convert entities to a stack of brackets
	var maxEndOffset int32
	for _, entity := range mergedEntities {
		log.Debugf("%#v", entity)
		if entity.Length <= 0 {
			continue
		}

		endOffset := entity.Offset + entity.Length
		if endOffset > maxEndOffset {
			maxEndOffset = endOffset
		}

		startStack, endStack = startStack.rebalance(endStack, entity.Offset)

		startInsertion, endInsertion := entityToMarkup(entity)
		if startInsertion != nil {
			startStack = append(startStack, startInsertion)
		}
		if endInsertion != nil {
			endStack = append(endStack, endInsertion)
		}
	}
	// flush the closing brackets that still remain in endStack
	startStack, endStack = startStack.rebalance(endStack, maxEndOffset)

	// merge brackets into text
	markupRunes := make([]rune, 0, len(sourceText))

	nextInsertion := startStack.NewIterator()
	insertion := nextInsertion()
	var runeI int32

	for _, cp := range sourceText {
		for insertion != nil && insertion.Offset <= runeI {
			markupRunes = append(markupRunes, insertion.Runes...)
			insertion = nextInsertion()
		}

		markupRunes = append(markupRunes, cp)
		// skip two UTF-16 code units (not points actually!) if needed
		if cp > 0x0000ffff {
			runeI += 2
		} else {
			runeI++
		}
	}
	for insertion != nil {
		markupRunes = append(markupRunes, insertion.Runes...)
		insertion = nextInsertion()
	}

	return string(markupRunes)
}