deeplake 0.1.0 → 0.2.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/browser/browser/client.d.ts +33 -0
- package/dist/browser/browser/client.d.ts.map +1 -0
- package/dist/browser/browser/client.js +312 -0
- package/dist/browser/browser/client.js.map +1 -0
- package/dist/browser/browser/index.d.ts +15 -0
- package/dist/browser/browser/index.d.ts.map +1 -0
- package/dist/browser/browser/index.js +11 -0
- package/dist/browser/browser/index.js.map +1 -0
- package/dist/browser/browser/wasm.d.ts +8 -0
- package/dist/browser/browser/wasm.d.ts.map +1 -0
- package/dist/browser/browser/wasm.js +53 -0
- package/dist/browser/browser/wasm.js.map +1 -0
- package/dist/browser/shared/api.d.ts +9 -0
- package/dist/browser/shared/api.d.ts.map +1 -0
- package/dist/browser/shared/api.js +67 -0
- package/dist/browser/shared/api.js.map +1 -0
- package/dist/browser/shared/credentials.d.ts +4 -0
- package/dist/browser/shared/credentials.d.ts.map +1 -0
- package/dist/browser/shared/credentials.js +24 -0
- package/dist/browser/shared/credentials.js.map +1 -0
- package/dist/browser/shared/database.d.ts +26 -0
- package/dist/browser/shared/database.d.ts.map +1 -0
- package/dist/browser/shared/database.js +67 -0
- package/dist/browser/shared/database.js.map +1 -0
- package/dist/browser/shared/dlref.d.ts +22 -0
- package/dist/browser/shared/dlref.d.ts.map +1 -0
- package/dist/browser/shared/dlref.js +48 -0
- package/dist/browser/shared/dlref.js.map +1 -0
- package/dist/browser/shared/errors.d.ts +22 -0
- package/dist/browser/shared/errors.d.ts.map +1 -0
- package/dist/browser/shared/errors.js +43 -0
- package/dist/browser/shared/errors.js.map +1 -0
- package/dist/browser/shared/schema.d.ts +7 -0
- package/dist/browser/shared/schema.d.ts.map +1 -0
- package/dist/browser/shared/schema.js +151 -0
- package/dist/browser/shared/schema.js.map +1 -0
- package/dist/browser/shared/token.d.ts +4 -0
- package/dist/browser/shared/token.d.ts.map +1 -0
- package/dist/browser/shared/token.js +69 -0
- package/dist/browser/shared/token.js.map +1 -0
- package/dist/browser/shared/types.d.ts +33 -0
- package/dist/browser/shared/types.d.ts.map +1 -0
- package/dist/browser/shared/types.js +2 -0
- package/dist/browser/shared/types.js.map +1 -0
- package/dist/browser/shared/wasm-common.d.ts +39 -0
- package/dist/browser/shared/wasm-common.d.ts.map +1 -0
- package/dist/browser/shared/wasm-common.js +287 -0
- package/dist/browser/shared/wasm-common.js.map +1 -0
- package/dist/node/node/client.d.ts +49 -0
- package/dist/node/node/client.d.ts.map +1 -0
- package/dist/node/node/client.js +670 -0
- package/dist/node/node/client.js.map +1 -0
- package/dist/node/node/formats/coco.d.ts +24 -0
- package/dist/node/node/formats/coco.d.ts.map +1 -0
- package/dist/node/node/formats/coco.js +342 -0
- package/dist/node/node/formats/coco.js.map +1 -0
- package/dist/node/node/formats/coco_panoptic.d.ts +19 -0
- package/dist/node/node/formats/coco_panoptic.d.ts.map +1 -0
- package/dist/node/node/formats/coco_panoptic.js +125 -0
- package/dist/node/node/formats/coco_panoptic.js.map +1 -0
- package/dist/node/node/formats/index.d.ts +7 -0
- package/dist/node/node/formats/index.d.ts.map +1 -0
- package/dist/node/node/formats/index.js +12 -0
- package/dist/node/node/formats/index.js.map +1 -0
- package/dist/node/node/formats/lerobot.d.ts +42 -0
- package/dist/node/node/formats/lerobot.d.ts.map +1 -0
- package/dist/node/node/formats/lerobot.js +351 -0
- package/dist/node/node/formats/lerobot.js.map +1 -0
- package/dist/node/node/index.d.ts +19 -0
- package/dist/node/node/index.d.ts.map +1 -0
- package/dist/node/node/index.js +75 -0
- package/dist/node/node/index.js.map +1 -0
- package/dist/node/node/mime.d.ts +2 -0
- package/dist/node/node/mime.d.ts.map +1 -0
- package/dist/node/node/mime.js +76 -0
- package/dist/node/node/mime.js.map +1 -0
- package/dist/node/node/normalizers/binary.d.ts +3 -0
- package/dist/node/node/normalizers/binary.d.ts.map +1 -0
- package/dist/node/node/normalizers/binary.js +51 -0
- package/dist/node/node/normalizers/binary.js.map +1 -0
- package/dist/node/node/normalizers/dict.d.ts +3 -0
- package/dist/node/node/normalizers/dict.d.ts.map +1 -0
- package/dist/node/node/normalizers/dict.js +17 -0
- package/dist/node/node/normalizers/dict.js.map +1 -0
- package/dist/node/node/normalizers/files.d.ts +3 -0
- package/dist/node/node/normalizers/files.d.ts.map +1 -0
- package/dist/node/node/normalizers/files.js +109 -0
- package/dist/node/node/normalizers/files.js.map +1 -0
- package/dist/node/node/normalizers/huggingface.d.ts +3 -0
- package/dist/node/node/normalizers/huggingface.d.ts.map +1 -0
- package/dist/node/node/normalizers/huggingface.js +158 -0
- package/dist/node/node/normalizers/huggingface.js.map +1 -0
- package/dist/node/node/normalizers/image.d.ts +3 -0
- package/dist/node/node/normalizers/image.d.ts.map +1 -0
- package/dist/node/node/normalizers/image.js +52 -0
- package/dist/node/node/normalizers/image.js.map +1 -0
- package/dist/node/node/normalizers/index.d.ts +9 -0
- package/dist/node/node/normalizers/index.d.ts.map +1 -0
- package/dist/node/node/normalizers/index.js +21 -0
- package/dist/node/node/normalizers/index.js.map +1 -0
- package/dist/node/node/normalizers/pdf.d.ts +3 -0
- package/dist/node/node/normalizers/pdf.d.ts.map +1 -0
- package/dist/node/node/normalizers/pdf.js +75 -0
- package/dist/node/node/normalizers/pdf.js.map +1 -0
- package/dist/node/node/normalizers/text.d.ts +4 -0
- package/dist/node/node/normalizers/text.d.ts.map +1 -0
- package/dist/node/node/normalizers/text.js +83 -0
- package/dist/node/node/normalizers/text.js.map +1 -0
- package/dist/node/node/normalizers/video.d.ts +3 -0
- package/dist/node/node/normalizers/video.d.ts.map +1 -0
- package/dist/node/node/normalizers/video.js +119 -0
- package/dist/node/node/normalizers/video.js.map +1 -0
- package/dist/node/node/storage.d.ts +15 -0
- package/dist/node/node/storage.d.ts.map +1 -0
- package/dist/node/node/storage.js +477 -0
- package/dist/node/node/storage.js.map +1 -0
- package/dist/node/node/wasm.d.ts +3 -0
- package/dist/node/node/wasm.d.ts.map +1 -0
- package/dist/node/node/wasm.js +49 -0
- package/dist/node/node/wasm.js.map +1 -0
- package/dist/node/shared/api.d.ts +9 -0
- package/dist/node/shared/api.d.ts.map +1 -0
- package/dist/node/shared/api.js +70 -0
- package/dist/node/shared/api.js.map +1 -0
- package/dist/node/shared/credentials.d.ts +4 -0
- package/dist/node/shared/credentials.d.ts.map +1 -0
- package/dist/node/shared/credentials.js +28 -0
- package/dist/node/shared/credentials.js.map +1 -0
- package/dist/node/shared/database.d.ts +26 -0
- package/dist/node/shared/database.d.ts.map +1 -0
- package/dist/node/shared/database.js +71 -0
- package/dist/node/shared/database.js.map +1 -0
- package/dist/node/shared/dlref.d.ts +22 -0
- package/dist/node/shared/dlref.d.ts.map +1 -0
- package/dist/node/shared/dlref.js +52 -0
- package/dist/node/shared/dlref.js.map +1 -0
- package/dist/node/shared/errors.d.ts +22 -0
- package/dist/node/shared/errors.d.ts.map +1 -0
- package/dist/node/shared/errors.js +53 -0
- package/dist/node/shared/errors.js.map +1 -0
- package/dist/node/shared/schema.d.ts +7 -0
- package/dist/node/shared/schema.d.ts.map +1 -0
- package/dist/node/shared/schema.js +157 -0
- package/dist/node/shared/schema.js.map +1 -0
- package/dist/node/shared/token.d.ts +4 -0
- package/dist/node/shared/token.d.ts.map +1 -0
- package/dist/node/shared/token.js +74 -0
- package/dist/node/shared/token.js.map +1 -0
- package/dist/node/shared/types.d.ts +33 -0
- package/dist/node/shared/types.d.ts.map +1 -0
- package/dist/node/shared/types.js +3 -0
- package/dist/node/shared/types.js.map +1 -0
- package/dist/node/shared/wasm-common.d.ts +39 -0
- package/dist/node/shared/wasm-common.d.ts.map +1 -0
- package/dist/node/shared/wasm-common.js +310 -0
- package/dist/node/shared/wasm-common.js.map +1 -0
- package/package.json +43 -18
- package/README.md +0 -39
- package/index.d.ts +0 -2
- package/index.js +0 -26
|
@@ -0,0 +1,83 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
|
|
3
|
+
if (k2 === undefined) k2 = k;
|
|
4
|
+
var desc = Object.getOwnPropertyDescriptor(m, k);
|
|
5
|
+
if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
|
|
6
|
+
desc = { enumerable: true, get: function() { return m[k]; } };
|
|
7
|
+
}
|
|
8
|
+
Object.defineProperty(o, k2, desc);
|
|
9
|
+
}) : (function(o, m, k, k2) {
|
|
10
|
+
if (k2 === undefined) k2 = k;
|
|
11
|
+
o[k2] = m[k];
|
|
12
|
+
}));
|
|
13
|
+
var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
|
|
14
|
+
Object.defineProperty(o, "default", { enumerable: true, value: v });
|
|
15
|
+
}) : function(o, v) {
|
|
16
|
+
o["default"] = v;
|
|
17
|
+
});
|
|
18
|
+
var __importStar = (this && this.__importStar) || (function () {
|
|
19
|
+
var ownKeys = function(o) {
|
|
20
|
+
ownKeys = Object.getOwnPropertyNames || function (o) {
|
|
21
|
+
var ar = [];
|
|
22
|
+
for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
|
|
23
|
+
return ar;
|
|
24
|
+
};
|
|
25
|
+
return ownKeys(o);
|
|
26
|
+
};
|
|
27
|
+
return function (mod) {
|
|
28
|
+
if (mod && mod.__esModule) return mod;
|
|
29
|
+
var result = {};
|
|
30
|
+
if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
|
|
31
|
+
__setModuleDefault(result, mod);
|
|
32
|
+
return result;
|
|
33
|
+
};
|
|
34
|
+
})();
|
|
35
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
36
|
+
exports.chunkText = chunkText;
|
|
37
|
+
exports.normalizeText = normalizeText;
|
|
38
|
+
const fs = __importStar(require("fs"));
|
|
39
|
+
const crypto = __importStar(require("crypto"));
|
|
40
|
+
function chunkText(text, chunkSize, overlap) {
|
|
41
|
+
if (text.length <= chunkSize) {
|
|
42
|
+
return text.trim() ? [text] : [];
|
|
43
|
+
}
|
|
44
|
+
const chunks = [];
|
|
45
|
+
let start = 0;
|
|
46
|
+
while (start < text.length) {
|
|
47
|
+
let end = start + chunkSize;
|
|
48
|
+
if (end >= text.length) {
|
|
49
|
+
const tail = text.slice(start).trim();
|
|
50
|
+
if (tail)
|
|
51
|
+
chunks.push(tail);
|
|
52
|
+
break;
|
|
53
|
+
}
|
|
54
|
+
const searchStart = Math.max(start + chunkSize - 100, start);
|
|
55
|
+
let bestBreak = end;
|
|
56
|
+
for (const sep of ['. ', '? ', '! ', '\n\n', '\n']) {
|
|
57
|
+
const pos = text.lastIndexOf(sep, end + 50);
|
|
58
|
+
if (pos >= searchStart && pos > start) {
|
|
59
|
+
bestBreak = pos + sep.length;
|
|
60
|
+
break;
|
|
61
|
+
}
|
|
62
|
+
}
|
|
63
|
+
const chunk = text.slice(start, bestBreak).trim();
|
|
64
|
+
if (chunk)
|
|
65
|
+
chunks.push(chunk);
|
|
66
|
+
const newStart = bestBreak - overlap;
|
|
67
|
+
start = newStart <= start ? bestBreak : newStart;
|
|
68
|
+
}
|
|
69
|
+
return chunks;
|
|
70
|
+
}
|
|
71
|
+
function normalizeText(filePath, chunkSize, chunkOverlap) {
|
|
72
|
+
const fileId = crypto.randomUUID();
|
|
73
|
+
const content = fs.readFileSync(filePath, 'utf-8');
|
|
74
|
+
const chunks = chunkText(content, chunkSize, chunkOverlap);
|
|
75
|
+
const name = filePath.split('/').pop() ?? filePath;
|
|
76
|
+
return {
|
|
77
|
+
id: chunks.map((_, i) => `${fileId}_${i}`),
|
|
78
|
+
file_id: Array(chunks.length).fill(fileId),
|
|
79
|
+
chunk_index: chunks.map((_, i) => i),
|
|
80
|
+
text: chunks.length > 0 ? chunks : [`Text: ${name}`],
|
|
81
|
+
};
|
|
82
|
+
}
|
|
83
|
+
//# sourceMappingURL=text.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"text.js","sourceRoot":"","sources":["../../../../src/node/normalizers/text.ts"],"names":[],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAIA,8BAwCC;AAED,sCAgBC;AA9DD,uCAAyB;AACzB,+CAAiC;AAGjC,SAAgB,SAAS,CACvB,IAAY,EACZ,SAAiB,EACjB,OAAe;IAEf,IAAI,IAAI,CAAC,MAAM,IAAI,SAAS,EAAE,CAAC;QAC7B,OAAO,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;IACnC,CAAC;IAED,MAAM,MAAM,GAAa,EAAE,CAAC;IAC5B,IAAI,KAAK,GAAG,CAAC,CAAC;IAEd,OAAO,KAAK,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC;QAC3B,IAAI,GAAG,GAAG,KAAK,GAAG,SAAS,CAAC;QAE5B,IAAI,GAAG,IAAI,IAAI,CAAC,MAAM,EAAE,CAAC;YACvB,MAAM,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,IAAI,EAAE,CAAC;YACtC,IAAI,IAAI;gBAAE,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YAC5B,MAAM;QACR,CAAC;QAED,MAAM,WAAW,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,GAAG,SAAS,GAAG,GAAG,EAAE,KAAK,CAAC,CAAC;QAC7D,IAAI,SAAS,GAAG,GAAG,CAAC;QAEpB,KAAK,MAAM,GAAG,IAAI,CAAC,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,CAAC,EAAE,CAAC;YACnD,MAAM,GAAG,GAAG,IAAI,CAAC,WAAW,CAAC,GAAG,EAAE,GAAG,GAAG,EAAE,CAAC,CAAC;YAC5C,IAAI,GAAG,IAAI,WAAW,IAAI,GAAG,GAAG,KAAK,EAAE,CAAC;gBACtC,SAAS,GAAG,GAAG,GAAG,GAAG,CAAC,MAAM,CAAC;gBAC7B,MAAM;YACR,CAAC;QACH,CAAC;QAED,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,SAAS,CAAC,CAAC,IAAI,EAAE,CAAC;QAClD,IAAI,KAAK;YAAE,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAE9B,MAAM,QAAQ,GAAG,SAAS,GAAG,OAAO,CAAC;QACrC,KAAK,GAAG,QAAQ,IAAI,KAAK,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,QAAQ,CAAC;IACnD,CAAC;IAED,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,SAAgB,aAAa,CAC3B,QAAgB,EAChB,SAAiB,EACjB,YAAoB;IAEpB,MAAM,MAAM,GAAG,MAAM,CAAC,UAAU,EAAE,CAAC;IACnC,MAAM,OAAO,GAAG,EAAE,CAAC,YAAY,CAAC,QAAQ,EAAE,OAAO,CAAC,CAAC;IACnD,MAAM,MAAM,GAAG,SAAS,CAAC,OAAO,EAAE,SAAS,EAAE,YAAY,CAAC,CAAC;IAC3D,MAAM,IAAI,GAAG,QAAQ,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,IAAI,QAAQ,CAAC;IAEnD,OAAO;QACL,EAAE,EAAE,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,GAAG,MAAM,IAAI,CAAC,EAAE,CAAC;QAC1C,OAAO,EAAE,KAAK,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC;QAC1C,WAAW,EAAE,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC;QACpC,IAAI,EAAE,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,SAAS,IAAI,EAAE,CAAC;KACrD,CAAC;AACJ,CAAC"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"video.d.ts","sourceRoot":"","sources":["../../../../src/node/normalizers/video.ts"],"names":[],"mappings":"AAKA,OAAO,KAAK,EAAE,KAAK,EAAE,MAAM,oBAAoB,CAAC;AAmBhD,wBAAuB,cAAc,CACnC,QAAQ,EAAE,MAAM,EAChB,aAAa,GAAE,MAAW,GACzB,cAAc,CAAC,KAAK,CAAC,CAwEvB"}
|
|
@@ -0,0 +1,119 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
|
|
3
|
+
if (k2 === undefined) k2 = k;
|
|
4
|
+
var desc = Object.getOwnPropertyDescriptor(m, k);
|
|
5
|
+
if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
|
|
6
|
+
desc = { enumerable: true, get: function() { return m[k]; } };
|
|
7
|
+
}
|
|
8
|
+
Object.defineProperty(o, k2, desc);
|
|
9
|
+
}) : (function(o, m, k, k2) {
|
|
10
|
+
if (k2 === undefined) k2 = k;
|
|
11
|
+
o[k2] = m[k];
|
|
12
|
+
}));
|
|
13
|
+
var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
|
|
14
|
+
Object.defineProperty(o, "default", { enumerable: true, value: v });
|
|
15
|
+
}) : function(o, v) {
|
|
16
|
+
o["default"] = v;
|
|
17
|
+
});
|
|
18
|
+
var __importStar = (this && this.__importStar) || (function () {
|
|
19
|
+
var ownKeys = function(o) {
|
|
20
|
+
ownKeys = Object.getOwnPropertyNames || function (o) {
|
|
21
|
+
var ar = [];
|
|
22
|
+
for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
|
|
23
|
+
return ar;
|
|
24
|
+
};
|
|
25
|
+
return ownKeys(o);
|
|
26
|
+
};
|
|
27
|
+
return function (mod) {
|
|
28
|
+
if (mod && mod.__esModule) return mod;
|
|
29
|
+
var result = {};
|
|
30
|
+
if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
|
|
31
|
+
__setModuleDefault(result, mod);
|
|
32
|
+
return result;
|
|
33
|
+
};
|
|
34
|
+
})();
|
|
35
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
36
|
+
exports.normalizeVideo = normalizeVideo;
|
|
37
|
+
const fs = __importStar(require("fs"));
|
|
38
|
+
const path = __importStar(require("path"));
|
|
39
|
+
const crypto = __importStar(require("crypto"));
|
|
40
|
+
const child_process_1 = require("child_process");
|
|
41
|
+
const util_1 = require("util");
|
|
42
|
+
const execFileAsync = (0, util_1.promisify)(child_process_1.execFile);
|
|
43
|
+
async function probeDuration(filePath) {
|
|
44
|
+
try {
|
|
45
|
+
const { stdout } = await execFileAsync('ffprobe', [
|
|
46
|
+
'-v', 'quiet',
|
|
47
|
+
'-print_format', 'json',
|
|
48
|
+
'-show_format',
|
|
49
|
+
filePath,
|
|
50
|
+
]);
|
|
51
|
+
const data = JSON.parse(stdout);
|
|
52
|
+
return parseFloat(data?.format?.duration ?? '0');
|
|
53
|
+
}
|
|
54
|
+
catch {
|
|
55
|
+
return 0;
|
|
56
|
+
}
|
|
57
|
+
}
|
|
58
|
+
async function* normalizeVideo(filePath, chunkInterval = 10) {
|
|
59
|
+
const fileId = crypto.randomUUID();
|
|
60
|
+
const filename = path.basename(filePath);
|
|
61
|
+
const duration = await probeDuration(filePath);
|
|
62
|
+
if (duration <= 0) {
|
|
63
|
+
yield {
|
|
64
|
+
id: [`${fileId}_0`],
|
|
65
|
+
file_id: [fileId],
|
|
66
|
+
chunk_index: [0],
|
|
67
|
+
start_time: [0.0],
|
|
68
|
+
end_time: [0.0],
|
|
69
|
+
video_data: [fs.readFileSync(filePath)],
|
|
70
|
+
thumbnail: [null],
|
|
71
|
+
text: [`Video: ${filename}`],
|
|
72
|
+
};
|
|
73
|
+
return;
|
|
74
|
+
}
|
|
75
|
+
const numChunks = Math.ceil(duration / chunkInterval);
|
|
76
|
+
const os = await Promise.resolve().then(() => __importStar(require('os')));
|
|
77
|
+
const tmpBase = os.tmpdir();
|
|
78
|
+
for (let idx = 0; idx < numChunks; idx++) {
|
|
79
|
+
const start = idx * chunkInterval;
|
|
80
|
+
const segDuration = Math.min(chunkInterval, duration - start);
|
|
81
|
+
const end = start + segDuration;
|
|
82
|
+
const tmpDir = fs.mkdtempSync(path.join(tmpBase, 'dl-video-'));
|
|
83
|
+
try {
|
|
84
|
+
const segPath = path.join(tmpDir, `seg_${idx}.mp4`);
|
|
85
|
+
try {
|
|
86
|
+
await execFileAsync('ffmpeg', ['-y', '-ss', String(start), '-t', String(segDuration), '-i', filePath, '-c', 'copy', segPath], { timeout: 300_000 });
|
|
87
|
+
}
|
|
88
|
+
catch {
|
|
89
|
+
continue;
|
|
90
|
+
}
|
|
91
|
+
let thumbnail = null;
|
|
92
|
+
const thumbPath = path.join(tmpDir, `thumb_${idx}.jpg`);
|
|
93
|
+
try {
|
|
94
|
+
await execFileAsync('ffmpeg', ['-y', '-ss', String((start + end) / 2), '-i', filePath, '-vframes', '1', '-s', '320x180', thumbPath], { timeout: 30_000 });
|
|
95
|
+
if (fs.existsSync(thumbPath)) {
|
|
96
|
+
thumbnail = fs.readFileSync(thumbPath);
|
|
97
|
+
}
|
|
98
|
+
}
|
|
99
|
+
catch {
|
|
100
|
+
thumbnail = null;
|
|
101
|
+
}
|
|
102
|
+
const videoData = fs.existsSync(segPath) ? fs.readFileSync(segPath) : null;
|
|
103
|
+
yield {
|
|
104
|
+
id: [`${fileId}_${idx}`],
|
|
105
|
+
file_id: [fileId],
|
|
106
|
+
chunk_index: [idx],
|
|
107
|
+
start_time: [Math.round(start * 1000) / 1000],
|
|
108
|
+
end_time: [Math.round(end * 1000) / 1000],
|
|
109
|
+
video_data: [videoData],
|
|
110
|
+
thumbnail: [thumbnail],
|
|
111
|
+
text: [`Video segment ${idx}: ${start.toFixed(1)}s - ${end.toFixed(1)}s`],
|
|
112
|
+
};
|
|
113
|
+
}
|
|
114
|
+
finally {
|
|
115
|
+
fs.rmSync(tmpDir, { recursive: true, force: true });
|
|
116
|
+
}
|
|
117
|
+
}
|
|
118
|
+
}
|
|
119
|
+
//# sourceMappingURL=video.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"video.js","sourceRoot":"","sources":["../../../../src/node/normalizers/video.ts"],"names":[],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAwBA,wCA2EC;AAnGD,uCAAyB;AACzB,2CAA6B;AAC7B,+CAAiC;AACjC,iDAAyC;AACzC,+BAAiC;AAGjC,MAAM,aAAa,GAAG,IAAA,gBAAS,EAAC,wBAAQ,CAAC,CAAC;AAE1C,KAAK,UAAU,aAAa,CAAC,QAAgB;IAC3C,IAAI,CAAC;QACH,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,aAAa,CAAC,SAAS,EAAE;YAChD,IAAI,EAAE,OAAO;YACb,eAAe,EAAE,MAAM;YACvB,cAAc;YACd,QAAQ;SACT,CAAC,CAAC;QACH,MAAM,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC;QAChC,OAAO,UAAU,CAAC,IAAI,EAAE,MAAM,EAAE,QAAQ,IAAI,GAAG,CAAC,CAAC;IACnD,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,CAAC,CAAC;IACX,CAAC;AACH,CAAC;AAEM,KAAK,SAAS,CAAC,CAAC,cAAc,CACnC,QAAgB,EAChB,gBAAwB,EAAE;IAE1B,MAAM,MAAM,GAAG,MAAM,CAAC,UAAU,EAAE,CAAC;IACnC,MAAM,QAAQ,GAAG,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACzC,MAAM,QAAQ,GAAG,MAAM,aAAa,CAAC,QAAQ,CAAC,CAAC;IAE/C,IAAI,QAAQ,IAAI,CAAC,EAAE,CAAC;QAClB,MAAM;YACJ,EAAE,EAAE,CAAC,GAAG,MAAM,IAAI,CAAC;YACnB,OAAO,EAAE,CAAC,MAAM,CAAC;YACjB,WAAW,EAAE,CAAC,CAAC,CAAC;YAChB,UAAU,EAAE,CAAC,GAAG,CAAC;YACjB,QAAQ,EAAE,CAAC,GAAG,CAAC;YACf,UAAU,EAAE,CAAC,EAAE,CAAC,YAAY,CAAC,QAAQ,CAAC,CAAC;YACvC,SAAS,EAAE,CAAC,IAAI,CAAC;YACjB,IAAI,EAAE,CAAC,UAAU,QAAQ,EAAE,CAAC;SAC7B,CAAC;QACF,OAAO;IACT,CAAC;IAED,MAAM,SAAS,GAAG,IAAI,CAAC,IAAI,CAAC,QAAQ,GAAG,aAAa,CAAC,CAAC;IACtD,MAAM,EAAE,GAAG,wDAAa,IAAI,GAAC,CAAC;IAC9B,MAAM,OAAO,GAAG,EAAE,CAAC,MAAM,EAAE,CAAC;IAE5B,KAAK,IAAI,GAAG,GAAG,CAAC,EAAE,GAAG,GAAG,SAAS,EAAE,GAAG,EAAE,EAAE,CAAC;QACzC,MAAM,KAAK,GAAG,GAAG,GAAG,aAAa,CAAC;QAClC,MAAM,WAAW,GAAG,IAAI,CAAC,GAAG,CAAC,aAAa,EAAE,QAAQ,GAAG,KAAK,CAAC,CAAC;QAC9D,MAAM,GAAG,GAAG,KAAK,GAAG,WAAW,CAAC;QAEhC,MAAM,MAAM,GAAG,EAAE,CAAC,WAAW,CAAC,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,WAAW,CAAC,CAAC,CAAC;QAC/D,IAAI,CAAC;YACH,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,OAAO,GAAG,MAAM,CAAC,CAAC;YACpD,IAAI,CAAC;gBACH,MAAM,aAAa,CACjB,QAAQ,EACR,CAAC,IAAI,EAAE,KAAK,EAAE,MAAM,CAAC,KAAK,CAAC,EAAE,IAAI,EAAE,MAAM,CAAC,WAAW,CAAC,EAAE,IAAI,EAAE,QAAQ,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,CAAC,EAC9F,EAAE,OAAO,EAAE,OAAO,EAAE,CACrB,CAAC;YACJ,CAAC;YAAC,MAAM,CAAC;gBACP,SAAS;YACX,CAAC;YAED,IAAI,SAAS,GAAkB,IAAI,CAAC;YACpC,MAAM,SAAS,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,SAAS,GAAG,MAAM,CAAC,CAAC;YACxD,IAAI,CAAC;gBACH,MAAM,aAAa,CACjB,QAAQ,EACR,CAAC,IAAI,EAAE,KAAK,EAAE,MAAM,CAAC,CAAC,KAAK,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,EAAE,QAAQ,EAAE,UAAU,EAAE,GAAG,EAAE,IAAI,EAAE,SAAS,EAAE,SAAS,CAAC,EACrG,EAAE,OAAO,EAAE,MAAM,EAAE,CACpB,CAAC;gBACF,IAAI,EAAE,CAAC,UAAU,CAAC,SAAS,CAAC,EAAE,CAAC;oBAC7B,SAAS,GAAG,EAAE,CAAC,YAAY,CAAC,SAAS,CAAC,CAAC;gBACzC,CAAC;YACH,CAAC;YAAC,MAAM,CAAC;gBACP,SAAS,GAAG,IAAI,CAAC;YACnB,CAAC;YAED,MAAM,SAAS,GAAG,EAAE,CAAC,UAAU,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,YAAY,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;YAE3E,MAAM;gBACJ,EAAE,EAAE,CAAC,GAAG,MAAM,IAAI,GAAG,EAAE,CAAC;gBACxB,OAAO,EAAE,CAAC,MAAM,CAAC;gBACjB,WAAW,EAAE,CAAC,GAAG,CAAC;gBAClB,UAAU,EAAE,CAAC,IAAI,CAAC,KAAK,CAAC,KAAK,GAAG,IAAI,CAAC,GAAG,IAAI,CAAC;gBAC7C,QAAQ,EAAE,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,GAAG,IAAI,CAAC,GAAG,IAAI,CAAC;gBACzC,UAAU,EAAE,CAAC,SAAS,CAAC;gBACvB,SAAS,EAAE,CAAC,SAAS,CAAC;gBACtB,IAAI,EAAE,CAAC,iBAAiB,GAAG,KAAK,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,OAAO,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,CAAC;aAC1E,CAAC;QACJ,CAAC;gBAAS,CAAC;YACT,EAAE,CAAC,MAAM,CAAC,MAAM,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;QACtD,CAAC;IACH,CAAC;AACH,CAAC"}
|
|
@@ -0,0 +1,15 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Node.js S3 storage client factory for the WASM deeplake engine.
|
|
3
|
+
*
|
|
4
|
+
* The C++ WASM code calls js_storage::client_factory()(type, config, mode?)
|
|
5
|
+
* to create a JS client object with methods like download, exists, list, etc.
|
|
6
|
+
* This module provides the factory that creates those client objects using
|
|
7
|
+
* the AWS SDK v3.
|
|
8
|
+
*
|
|
9
|
+
* Credential rotation: STS credentials are refreshed automatically before
|
|
10
|
+
* they expire (5 minutes before expiry, matching the native C++ behavior).
|
|
11
|
+
* The C++ side passes refresh metadata (orgId, dsName, token) through the
|
|
12
|
+
* config object so the JS side can call the backend API for fresh creds.
|
|
13
|
+
*/
|
|
14
|
+
export declare function createStorageFactory(wasmModule: any): (type: string, config: any, mode?: string) => any;
|
|
15
|
+
//# sourceMappingURL=storage.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"storage.d.ts","sourceRoot":"","sources":["../../../src/node/storage.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;GAYG;AAsKH,wBAAgB,oBAAoB,CAAC,UAAU,EAAE,GAAG,IAEhD,MAAM,MAAM,EACZ,QAAQ,GAAG,EACX,OAAO,MAAM,KACZ,GAAG,CA8FP"}
|