wikitongues-db 0.1.0
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/LICENSE +21 -0
- package/README.md +172 -0
- package/data/processed/wikitongues_normalized.json +49337 -0
- package/data/processed/wikitongues_normalized.jsonl +863 -0
- package/dist/dataset-jRTVzgZX.d.mts +104 -0
- package/dist/dataset-jRTVzgZX.d.ts +104 -0
- package/dist/dataset.d.mts +1 -0
- package/dist/dataset.d.ts +1 -0
- package/dist/dataset.js +49351 -0
- package/dist/dataset.js.map +1 -0
- package/dist/dataset.mjs +49346 -0
- package/dist/dataset.mjs.map +1 -0
- package/dist/index.d.mts +394 -0
- package/dist/index.d.ts +394 -0
- package/dist/index.js +51242 -0
- package/dist/index.js.map +1 -0
- package/dist/index.mjs +51226 -0
- package/dist/index.mjs.map +1 -0
- package/package.json +57 -0
|
@@ -0,0 +1,104 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Domain types and interfaces for Wikitongues Database.
|
|
3
|
+
*/
|
|
4
|
+
interface LanguageData {
|
|
5
|
+
iso639_3: string;
|
|
6
|
+
bcp47: string;
|
|
7
|
+
name: string;
|
|
8
|
+
glottocode?: string | null;
|
|
9
|
+
autonym?: string | null;
|
|
10
|
+
dialect?: string | null;
|
|
11
|
+
}
|
|
12
|
+
interface SpeakerData {
|
|
13
|
+
name: string;
|
|
14
|
+
role?: string;
|
|
15
|
+
origin?: string | null;
|
|
16
|
+
}
|
|
17
|
+
interface ProvenanceData {
|
|
18
|
+
country_code?: string | null;
|
|
19
|
+
country_name?: string | null;
|
|
20
|
+
region?: string | null;
|
|
21
|
+
city?: string | null;
|
|
22
|
+
recorded_by?: string | null;
|
|
23
|
+
recording_date?: string | null;
|
|
24
|
+
}
|
|
25
|
+
interface TranscriptionData {
|
|
26
|
+
has_subtitles?: boolean;
|
|
27
|
+
available_subtitles?: string[];
|
|
28
|
+
native_text?: string | null;
|
|
29
|
+
english_translation?: string | null;
|
|
30
|
+
}
|
|
31
|
+
interface RawMetadataData {
|
|
32
|
+
title?: string;
|
|
33
|
+
tags?: string[];
|
|
34
|
+
}
|
|
35
|
+
interface VideoData {
|
|
36
|
+
id: string;
|
|
37
|
+
url: string;
|
|
38
|
+
duration_seconds: number;
|
|
39
|
+
upload_date: string;
|
|
40
|
+
license: string;
|
|
41
|
+
content_type: string;
|
|
42
|
+
primary_language: LanguageData;
|
|
43
|
+
additional_languages?: LanguageData[];
|
|
44
|
+
speakers?: SpeakerData[];
|
|
45
|
+
provenance?: ProvenanceData | null;
|
|
46
|
+
transcription?: TranscriptionData | null;
|
|
47
|
+
raw_metadata?: RawMetadataData | null;
|
|
48
|
+
}
|
|
49
|
+
interface FilterOptions {
|
|
50
|
+
language?: string;
|
|
51
|
+
country?: string;
|
|
52
|
+
speaker?: string;
|
|
53
|
+
contentType?: string;
|
|
54
|
+
license?: string;
|
|
55
|
+
creativeCommons?: boolean;
|
|
56
|
+
subtitles?: boolean;
|
|
57
|
+
minDuration?: number;
|
|
58
|
+
maxDuration?: number;
|
|
59
|
+
limit?: number;
|
|
60
|
+
}
|
|
61
|
+
interface LanguageSummary {
|
|
62
|
+
iso639_3: string;
|
|
63
|
+
bcp47: string;
|
|
64
|
+
name: string;
|
|
65
|
+
glottocode: string | null;
|
|
66
|
+
autonyms: string[];
|
|
67
|
+
dialects: string[];
|
|
68
|
+
video_count: number;
|
|
69
|
+
total_duration_seconds: number;
|
|
70
|
+
countries: string[];
|
|
71
|
+
}
|
|
72
|
+
interface CountrySummary {
|
|
73
|
+
country_code: string | null;
|
|
74
|
+
country_name: string;
|
|
75
|
+
video_count: number;
|
|
76
|
+
language_count: number;
|
|
77
|
+
languages: string[];
|
|
78
|
+
}
|
|
79
|
+
interface SpeakerSummary {
|
|
80
|
+
name: string;
|
|
81
|
+
roles: string[];
|
|
82
|
+
origins: string[];
|
|
83
|
+
video_count: number;
|
|
84
|
+
languages: string[];
|
|
85
|
+
}
|
|
86
|
+
interface DatasetStats {
|
|
87
|
+
total_videos: number;
|
|
88
|
+
total_languages: number;
|
|
89
|
+
total_countries: number;
|
|
90
|
+
total_duration_seconds: number;
|
|
91
|
+
total_duration_hours: number;
|
|
92
|
+
with_subtitles_count: number;
|
|
93
|
+
with_subtitles_percentage: number;
|
|
94
|
+
licenses: Record<string, number>;
|
|
95
|
+
content_types: Record<string, number>;
|
|
96
|
+
}
|
|
97
|
+
|
|
98
|
+
/**
|
|
99
|
+
* Static normalized Wikitongues dataset.
|
|
100
|
+
*/
|
|
101
|
+
|
|
102
|
+
declare const dataset: VideoData[];
|
|
103
|
+
|
|
104
|
+
export { type CountrySummary as C, type DatasetStats as D, type FilterOptions as F, type LanguageData as L, type ProvenanceData as P, type RawMetadataData as R, type SpeakerData as S, type TranscriptionData as T, type VideoData as V, type LanguageSummary as a, type SpeakerSummary as b, dataset as d };
|
|
@@ -0,0 +1,104 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Domain types and interfaces for Wikitongues Database.
|
|
3
|
+
*/
|
|
4
|
+
interface LanguageData {
|
|
5
|
+
iso639_3: string;
|
|
6
|
+
bcp47: string;
|
|
7
|
+
name: string;
|
|
8
|
+
glottocode?: string | null;
|
|
9
|
+
autonym?: string | null;
|
|
10
|
+
dialect?: string | null;
|
|
11
|
+
}
|
|
12
|
+
interface SpeakerData {
|
|
13
|
+
name: string;
|
|
14
|
+
role?: string;
|
|
15
|
+
origin?: string | null;
|
|
16
|
+
}
|
|
17
|
+
interface ProvenanceData {
|
|
18
|
+
country_code?: string | null;
|
|
19
|
+
country_name?: string | null;
|
|
20
|
+
region?: string | null;
|
|
21
|
+
city?: string | null;
|
|
22
|
+
recorded_by?: string | null;
|
|
23
|
+
recording_date?: string | null;
|
|
24
|
+
}
|
|
25
|
+
interface TranscriptionData {
|
|
26
|
+
has_subtitles?: boolean;
|
|
27
|
+
available_subtitles?: string[];
|
|
28
|
+
native_text?: string | null;
|
|
29
|
+
english_translation?: string | null;
|
|
30
|
+
}
|
|
31
|
+
interface RawMetadataData {
|
|
32
|
+
title?: string;
|
|
33
|
+
tags?: string[];
|
|
34
|
+
}
|
|
35
|
+
interface VideoData {
|
|
36
|
+
id: string;
|
|
37
|
+
url: string;
|
|
38
|
+
duration_seconds: number;
|
|
39
|
+
upload_date: string;
|
|
40
|
+
license: string;
|
|
41
|
+
content_type: string;
|
|
42
|
+
primary_language: LanguageData;
|
|
43
|
+
additional_languages?: LanguageData[];
|
|
44
|
+
speakers?: SpeakerData[];
|
|
45
|
+
provenance?: ProvenanceData | null;
|
|
46
|
+
transcription?: TranscriptionData | null;
|
|
47
|
+
raw_metadata?: RawMetadataData | null;
|
|
48
|
+
}
|
|
49
|
+
interface FilterOptions {
|
|
50
|
+
language?: string;
|
|
51
|
+
country?: string;
|
|
52
|
+
speaker?: string;
|
|
53
|
+
contentType?: string;
|
|
54
|
+
license?: string;
|
|
55
|
+
creativeCommons?: boolean;
|
|
56
|
+
subtitles?: boolean;
|
|
57
|
+
minDuration?: number;
|
|
58
|
+
maxDuration?: number;
|
|
59
|
+
limit?: number;
|
|
60
|
+
}
|
|
61
|
+
interface LanguageSummary {
|
|
62
|
+
iso639_3: string;
|
|
63
|
+
bcp47: string;
|
|
64
|
+
name: string;
|
|
65
|
+
glottocode: string | null;
|
|
66
|
+
autonyms: string[];
|
|
67
|
+
dialects: string[];
|
|
68
|
+
video_count: number;
|
|
69
|
+
total_duration_seconds: number;
|
|
70
|
+
countries: string[];
|
|
71
|
+
}
|
|
72
|
+
interface CountrySummary {
|
|
73
|
+
country_code: string | null;
|
|
74
|
+
country_name: string;
|
|
75
|
+
video_count: number;
|
|
76
|
+
language_count: number;
|
|
77
|
+
languages: string[];
|
|
78
|
+
}
|
|
79
|
+
interface SpeakerSummary {
|
|
80
|
+
name: string;
|
|
81
|
+
roles: string[];
|
|
82
|
+
origins: string[];
|
|
83
|
+
video_count: number;
|
|
84
|
+
languages: string[];
|
|
85
|
+
}
|
|
86
|
+
interface DatasetStats {
|
|
87
|
+
total_videos: number;
|
|
88
|
+
total_languages: number;
|
|
89
|
+
total_countries: number;
|
|
90
|
+
total_duration_seconds: number;
|
|
91
|
+
total_duration_hours: number;
|
|
92
|
+
with_subtitles_count: number;
|
|
93
|
+
with_subtitles_percentage: number;
|
|
94
|
+
licenses: Record<string, number>;
|
|
95
|
+
content_types: Record<string, number>;
|
|
96
|
+
}
|
|
97
|
+
|
|
98
|
+
/**
|
|
99
|
+
* Static normalized Wikitongues dataset.
|
|
100
|
+
*/
|
|
101
|
+
|
|
102
|
+
declare const dataset: VideoData[];
|
|
103
|
+
|
|
104
|
+
export { type CountrySummary as C, type DatasetStats as D, type FilterOptions as F, type LanguageData as L, type ProvenanceData as P, type RawMetadataData as R, type SpeakerData as S, type TranscriptionData as T, type VideoData as V, type LanguageSummary as a, type SpeakerSummary as b, dataset as d };
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
export { d as dataset, d as default } from './dataset-jRTVzgZX.mjs';
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
export { d as dataset, d as default } from './dataset-jRTVzgZX.js';
|