File size: 5,362 Bytes
1185ec1
 
 
637b219
e02a62b
1185ec1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e40bd21
1185ec1
 
 
 
 
 
 
 
e40bd21
 
1185ec1
e02a62b
1185ec1
 
e40bd21
 
 
 
 
 
 
 
 
 
 
 
 
 
 
637b219
e40bd21
 
 
 
 
 
 
 
637b219
 
e40bd21
 
 
637b219
 
 
 
e40bd21
 
 
 
 
 
 
 
 
e02a62b
 
 
 
 
 
 
 
 
e40bd21
 
 
 
 
 
 
 
 
e02a62b
 
 
 
 
 
 
 
 
e40bd21
 
637b219
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e40bd21
1185ec1
 
 
 
 
 
637b219
1185ec1
3adb71a
1185ec1
 
 
 
3adb71a
637b219
 
1185ec1
 
 
 
 
 
 
 
 
 
637b219
1185ec1
3adb71a
1185ec1
 
 
 
3adb71a
637b219
 
1185ec1
 
e40bd21
1185ec1
 
 
e40bd21
637b219
1185ec1
 
 
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
import YAML from "yaml"
import { v4 as uuidv4 } from "uuid"

import { ClapHeader, ClapMeta, ClapModel, ClapProject, ClapScene, ClapSegment } from "./types"
import { getValidNumber } from "@/lib/getValidNumber"

/**
 * import a Clap file (from a plain text string)
 * 
 * note: it is not really async, because for some reason YAML.parse is a blocking call like for JSON,
 * they is no async version although we are now in the 20s not 90s
 */
export async function parseClap(inputStringOrBlob: string | Blob): Promise<ClapProject> {

  // Decompress the input blob using gzip
  const decompressor = new DecompressionStream('gzip');

  const inputBlob =
    typeof inputStringOrBlob === "string"
    ? new Blob([inputStringOrBlob], { type: "application/x-yaml" })
    : inputStringOrBlob;

  const decompressedStream = inputBlob.stream().pipeThrough(decompressor);

  // Convert the stream to text using a Response object
  const text = await new Response(decompressedStream).text();
  
  // Parse YAML string to raw data
  const rawData = YAML.parse(text);

  if (!Array.isArray(rawData) || rawData.length < 2) {
    throw new Error("invalid clap file (need a clap format header block and project metadata block)")
  }

  const maybeClapHeader = rawData[0] as ClapHeader

  if (maybeClapHeader.format !== "clap-0") {
    throw new Error("invalid clap file (sorry, but you can't make up version numbers like that)")
  }


  const maybeClapMeta = rawData[1] as ClapMeta

  const clapMeta: ClapMeta = {
    id: typeof maybeClapMeta.title === "string" ? maybeClapMeta.id : uuidv4(),
    title: typeof maybeClapMeta.title === "string" ? maybeClapMeta.title : "",
    description: typeof maybeClapMeta.description === "string" ? maybeClapMeta.description : "",
    licence: typeof maybeClapMeta.licence === "string" ? maybeClapMeta.licence : "",
    orientation: maybeClapMeta.orientation === "portrait" ? "portrait" : maybeClapMeta.orientation === "square" ? "square" : "landscape",
    width: getValidNumber(maybeClapMeta.width, 256, 8192, 1024),
    height: getValidNumber(maybeClapMeta.height, 256, 8192, 576),
    defaultVideoModel: typeof maybeClapMeta.defaultVideoModel === "string" ? maybeClapMeta.defaultVideoModel : "SVD",
    extraPositivePrompt: Array.isArray(maybeClapMeta.extraPositivePrompt) ? maybeClapMeta.extraPositivePrompt : []
  }

  /*
  in case we want to support streaming (mix of models and segments etc), we could do it this way:

  const maybeModelsOrSegments = rawData.slice(2)
  maybeModelsOrSegments.forEach((unknownElement: any) => {
    if (isValidNumber(unknownElement?.track)) {
      maybeSegments.push(unknownElement as ClapSegment)
    } else {
      maybeModels.push(unknownElement as ClapModel)
    }
  })
  */


  const expectedNumberOfModels = maybeClapHeader.numberOfModels || 0
  const expectedNumberOfScenes = maybeClapHeader.numberOfScenes || 0
  const expectedNumberOfSegments = maybeClapHeader.numberOfSegments || 0

  // note: we assume the order is strictly enforced!
  // if you implement streaming (mix of models and segments) you will have to rewrite this!

  const afterTheHeaders = 2
  const afterTheModels = afterTheHeaders + expectedNumberOfModels

  const afterTheScenes = afterTheModels + expectedNumberOfScenes

  // note: if there are no expected models, maybeModels will be empty
  const maybeModels = rawData.slice(afterTheHeaders, afterTheModels) as ClapModel[]

  // note: if there are no expected scenes, maybeScenes will be empty
  const maybeScenes = rawData.slice(afterTheModels, afterTheScenes) as ClapScene[]

  const maybeSegments = rawData.slice(afterTheScenes) as ClapSegment[]

  const clapModels: ClapModel[] = maybeModels.map(({
    id,
    category,
    triggerName,
    label,
    description,
    author,
    thumbnailUrl,
    seed,
    assetSourceType,
    assetUrl,
    age,
    gender,
    region,
    appearance,
    voiceVendor,
    voiceId,
  }) => ({
    // TODO: we should verify each of those, probably
    id,
    category,
    triggerName,
    label,
    description,
    author,
    thumbnailUrl,
    seed,
    assetSourceType,
    assetUrl,
    age,
    gender,
    region,
    appearance,
    voiceVendor,
    voiceId,
  }))

  const clapScenes: ClapScene[] = maybeScenes.map(({
    id,
    scene,
    line,
    rawLine,
    sequenceFullText,
    sequenceStartAtLine,
    sequenceEndAtLine,
    startAtLine,
    endAtLine,
    events,
  }) => ({
    id,
    scene,
    line,
    rawLine,
    sequenceFullText,
    sequenceStartAtLine,
    sequenceEndAtLine,
    startAtLine,
    endAtLine,
    events: events.map(e => e)
  }))
  
  const clapSegments: ClapSegment[] = maybeSegments.map(({
    id,
    track,
    startTimeInMs,
    endTimeInMs,
    category,
    modelId,
    sceneId,
    prompt,
    label,
    outputType,
    renderId,
    status,
    assetUrl,
    assetDurationInMs,
    createdBy,
    editedBy,
    outputGain,
    seed,
  }) => ({
    // TODO: we should verify each of those, probably
    id,
    track,
    startTimeInMs,
    endTimeInMs,
    category,
    modelId,
    sceneId,
    prompt,
    label,
    outputType,
    renderId,
    status,
    assetUrl,
    assetDurationInMs,
    createdBy,
    editedBy,
    outputGain,
    seed,
  }))

  return {
    meta: clapMeta,
    models: clapModels,
    scenes: clapScenes,
    segments: clapSegments
  }
}