acryl-datahub 0.15.0.2rc4__py3-none-any.whl → 0.15.0.2rc5__py3-none-any.whl
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.
Potentially problematic release.
This version of acryl-datahub might be problematic. Click here for more details.
- {acryl_datahub-0.15.0.2rc4.dist-info → acryl_datahub-0.15.0.2rc5.dist-info}/METADATA +2578 -2578
- {acryl_datahub-0.15.0.2rc4.dist-info → acryl_datahub-0.15.0.2rc5.dist-info}/RECORD +49 -45
- datahub/__init__.py +1 -1
- datahub/cli/delete_cli.py +3 -3
- datahub/cli/migrate.py +2 -2
- datahub/emitter/mcp_builder.py +27 -0
- datahub/emitter/rest_emitter.py +1 -1
- datahub/ingestion/api/source.py +2 -2
- datahub/ingestion/source/delta_lake/source.py +0 -5
- datahub/ingestion/source/demo_data.py +1 -1
- datahub/ingestion/source/fivetran/fivetran.py +1 -6
- datahub/ingestion/source/iceberg/iceberg.py +10 -3
- datahub/ingestion/source/iceberg/iceberg_common.py +49 -9
- datahub/ingestion/source/iceberg/iceberg_profiler.py +3 -1
- datahub/ingestion/source/kafka_connect/kafka_connect.py +1 -6
- datahub/ingestion/source/metabase.py +1 -6
- datahub/ingestion/source/mlflow.py +0 -5
- datahub/ingestion/source/nifi.py +0 -5
- datahub/ingestion/source/redash.py +0 -5
- datahub/ingestion/source/redshift/redshift.py +1 -0
- datahub/ingestion/source/snowflake/snowflake_config.py +13 -0
- datahub/ingestion/source/snowflake/snowflake_schema.py +5 -2
- datahub/ingestion/source/snowflake/snowflake_schema_gen.py +112 -20
- datahub/ingestion/source/snowflake/snowflake_tag.py +14 -4
- datahub/ingestion/source/snowflake/snowflake_v2.py +0 -6
- datahub/ingestion/source/sql/sql_types.py +1 -1
- datahub/ingestion/source/sql/sql_utils.py +5 -0
- datahub/ingestion/source/superset.py +1 -6
- datahub/ingestion/source/tableau/tableau.py +0 -6
- datahub/metadata/_schema_classes.py +314 -41
- datahub/metadata/_urns/urn_defs.py +54 -0
- datahub/metadata/com/linkedin/pegasus2avro/common/__init__.py +2 -0
- datahub/metadata/com/linkedin/pegasus2avro/metadata/key/__init__.py +2 -0
- datahub/metadata/com/linkedin/pegasus2avro/versionset/__init__.py +17 -0
- datahub/metadata/schema.avsc +296 -87
- datahub/metadata/schemas/DatasetKey.avsc +2 -1
- datahub/metadata/schemas/MLFeatureProperties.avsc +51 -0
- datahub/metadata/schemas/MLModelDeploymentProperties.avsc +51 -0
- datahub/metadata/schemas/MLModelGroupProperties.avsc +96 -23
- datahub/metadata/schemas/MLModelKey.avsc +2 -1
- datahub/metadata/schemas/MLModelProperties.avsc +96 -48
- datahub/metadata/schemas/MLPrimaryKeyProperties.avsc +51 -0
- datahub/metadata/schemas/MetadataChangeEvent.avsc +98 -71
- datahub/metadata/schemas/VersionProperties.avsc +216 -0
- datahub/metadata/schemas/VersionSetKey.avsc +26 -0
- datahub/metadata/schemas/VersionSetProperties.avsc +49 -0
- {acryl_datahub-0.15.0.2rc4.dist-info → acryl_datahub-0.15.0.2rc5.dist-info}/WHEEL +0 -0
- {acryl_datahub-0.15.0.2rc4.dist-info → acryl_datahub-0.15.0.2rc5.dist-info}/entry_points.txt +0 -0
- {acryl_datahub-0.15.0.2rc4.dist-info → acryl_datahub-0.15.0.2rc5.dist-info}/top_level.txt +0 -0
|
@@ -0,0 +1,216 @@
|
|
|
1
|
+
{
|
|
2
|
+
"type": "record",
|
|
3
|
+
"Aspect": {
|
|
4
|
+
"name": "versionProperties"
|
|
5
|
+
},
|
|
6
|
+
"name": "VersionProperties",
|
|
7
|
+
"namespace": "com.linkedin.pegasus2avro.common",
|
|
8
|
+
"fields": [
|
|
9
|
+
{
|
|
10
|
+
"Relationship": {
|
|
11
|
+
"entityTypes": [
|
|
12
|
+
"versionSet"
|
|
13
|
+
],
|
|
14
|
+
"name": "VersionOf"
|
|
15
|
+
},
|
|
16
|
+
"Searchable": {
|
|
17
|
+
"queryByDefault": false
|
|
18
|
+
},
|
|
19
|
+
"java": {
|
|
20
|
+
"class": "com.linkedin.pegasus2avro.common.urn.Urn"
|
|
21
|
+
},
|
|
22
|
+
"type": "string",
|
|
23
|
+
"name": "versionSet",
|
|
24
|
+
"doc": "The linked Version Set entity that ties multiple versioned assets together",
|
|
25
|
+
"Urn": "Urn",
|
|
26
|
+
"entityTypes": [
|
|
27
|
+
"versionSet"
|
|
28
|
+
]
|
|
29
|
+
},
|
|
30
|
+
{
|
|
31
|
+
"Searchable": {
|
|
32
|
+
"/versionTag": {
|
|
33
|
+
"fieldName": "version",
|
|
34
|
+
"queryByDefault": false
|
|
35
|
+
}
|
|
36
|
+
},
|
|
37
|
+
"type": {
|
|
38
|
+
"type": "record",
|
|
39
|
+
"name": "VersionTag",
|
|
40
|
+
"namespace": "com.linkedin.pegasus2avro.common",
|
|
41
|
+
"fields": [
|
|
42
|
+
{
|
|
43
|
+
"type": [
|
|
44
|
+
"null",
|
|
45
|
+
"string"
|
|
46
|
+
],
|
|
47
|
+
"name": "versionTag",
|
|
48
|
+
"default": null
|
|
49
|
+
},
|
|
50
|
+
{
|
|
51
|
+
"type": [
|
|
52
|
+
"null",
|
|
53
|
+
{
|
|
54
|
+
"type": "record",
|
|
55
|
+
"name": "MetadataAttribution",
|
|
56
|
+
"namespace": "com.linkedin.pegasus2avro.common",
|
|
57
|
+
"fields": [
|
|
58
|
+
{
|
|
59
|
+
"type": "long",
|
|
60
|
+
"name": "time",
|
|
61
|
+
"doc": "When this metadata was updated."
|
|
62
|
+
},
|
|
63
|
+
{
|
|
64
|
+
"java": {
|
|
65
|
+
"class": "com.linkedin.pegasus2avro.common.urn.Urn"
|
|
66
|
+
},
|
|
67
|
+
"type": "string",
|
|
68
|
+
"name": "actor",
|
|
69
|
+
"doc": "The entity (e.g. a member URN) responsible for applying the assocated metadata. This can\neither be a user (in case of UI edits) or the datahub system for automation.",
|
|
70
|
+
"Urn": "Urn"
|
|
71
|
+
},
|
|
72
|
+
{
|
|
73
|
+
"java": {
|
|
74
|
+
"class": "com.linkedin.pegasus2avro.common.urn.Urn"
|
|
75
|
+
},
|
|
76
|
+
"type": [
|
|
77
|
+
"null",
|
|
78
|
+
"string"
|
|
79
|
+
],
|
|
80
|
+
"name": "source",
|
|
81
|
+
"default": null,
|
|
82
|
+
"doc": "The DataHub source responsible for applying the associated metadata. This will only be filled out\nwhen a DataHub source is responsible. This includes the specific metadata test urn, the automation urn.",
|
|
83
|
+
"Urn": "Urn"
|
|
84
|
+
},
|
|
85
|
+
{
|
|
86
|
+
"type": {
|
|
87
|
+
"type": "map",
|
|
88
|
+
"values": "string"
|
|
89
|
+
},
|
|
90
|
+
"name": "sourceDetail",
|
|
91
|
+
"default": {},
|
|
92
|
+
"doc": "The details associated with why this metadata was applied. For example, this could include\nthe actual regex rule, sql statement, ingestion pipeline ID, etc."
|
|
93
|
+
}
|
|
94
|
+
],
|
|
95
|
+
"doc": "Information about who, why, and how this metadata was applied"
|
|
96
|
+
}
|
|
97
|
+
],
|
|
98
|
+
"name": "metadataAttribution",
|
|
99
|
+
"default": null
|
|
100
|
+
}
|
|
101
|
+
],
|
|
102
|
+
"doc": "A resource-defined string representing the resource state for the purpose of concurrency control"
|
|
103
|
+
},
|
|
104
|
+
"name": "version",
|
|
105
|
+
"doc": "Label for this versioned asset, is unique within a version set"
|
|
106
|
+
},
|
|
107
|
+
{
|
|
108
|
+
"Searchable": {
|
|
109
|
+
"/*/versionTag": {
|
|
110
|
+
"fieldName": "aliases",
|
|
111
|
+
"queryByDefault": false
|
|
112
|
+
}
|
|
113
|
+
},
|
|
114
|
+
"type": {
|
|
115
|
+
"type": "array",
|
|
116
|
+
"items": "com.linkedin.pegasus2avro.common.VersionTag"
|
|
117
|
+
},
|
|
118
|
+
"name": "aliases",
|
|
119
|
+
"default": [],
|
|
120
|
+
"doc": "Associated aliases for this versioned asset"
|
|
121
|
+
},
|
|
122
|
+
{
|
|
123
|
+
"type": [
|
|
124
|
+
"null",
|
|
125
|
+
"string"
|
|
126
|
+
],
|
|
127
|
+
"name": "comment",
|
|
128
|
+
"default": null,
|
|
129
|
+
"doc": "Comment documenting what this version was created for, changes, or represents"
|
|
130
|
+
},
|
|
131
|
+
{
|
|
132
|
+
"Searchable": {
|
|
133
|
+
"fieldName": "versionSortId",
|
|
134
|
+
"queryByDefault": false
|
|
135
|
+
},
|
|
136
|
+
"type": "string",
|
|
137
|
+
"name": "sortId",
|
|
138
|
+
"doc": "Sort identifier that determines where a version lives in the order of the Version Set.\nWhat this looks like depends on the Version Scheme. For sort ids generated by DataHub we use an 8 character string representation."
|
|
139
|
+
},
|
|
140
|
+
{
|
|
141
|
+
"type": [
|
|
142
|
+
"null",
|
|
143
|
+
{
|
|
144
|
+
"type": "record",
|
|
145
|
+
"name": "AuditStamp",
|
|
146
|
+
"namespace": "com.linkedin.pegasus2avro.common",
|
|
147
|
+
"fields": [
|
|
148
|
+
{
|
|
149
|
+
"type": "long",
|
|
150
|
+
"name": "time",
|
|
151
|
+
"doc": "When did the resource/association/sub-resource move into the specific lifecycle stage represented by this AuditEvent."
|
|
152
|
+
},
|
|
153
|
+
{
|
|
154
|
+
"java": {
|
|
155
|
+
"class": "com.linkedin.pegasus2avro.common.urn.Urn"
|
|
156
|
+
},
|
|
157
|
+
"type": "string",
|
|
158
|
+
"name": "actor",
|
|
159
|
+
"doc": "The entity (e.g. a member URN) which will be credited for moving the resource/association/sub-resource into the specific lifecycle stage. It is also the one used to authorize the change.",
|
|
160
|
+
"Urn": "Urn"
|
|
161
|
+
},
|
|
162
|
+
{
|
|
163
|
+
"java": {
|
|
164
|
+
"class": "com.linkedin.pegasus2avro.common.urn.Urn"
|
|
165
|
+
},
|
|
166
|
+
"type": [
|
|
167
|
+
"null",
|
|
168
|
+
"string"
|
|
169
|
+
],
|
|
170
|
+
"name": "impersonator",
|
|
171
|
+
"default": null,
|
|
172
|
+
"doc": "The entity (e.g. a service URN) which performs the change on behalf of the Actor and must be authorized to act as the Actor.",
|
|
173
|
+
"Urn": "Urn"
|
|
174
|
+
},
|
|
175
|
+
{
|
|
176
|
+
"type": [
|
|
177
|
+
"null",
|
|
178
|
+
"string"
|
|
179
|
+
],
|
|
180
|
+
"name": "message",
|
|
181
|
+
"default": null,
|
|
182
|
+
"doc": "Additional context around how DataHub was informed of the particular change. For example: was the change created by an automated process, or manually."
|
|
183
|
+
}
|
|
184
|
+
],
|
|
185
|
+
"doc": "Data captured on a resource/association/sub-resource level giving insight into when that resource/association/sub-resource moved into a particular lifecycle stage, and who acted to move it into that specific lifecycle stage."
|
|
186
|
+
}
|
|
187
|
+
],
|
|
188
|
+
"name": "sourceCreatedTimestamp",
|
|
189
|
+
"default": null,
|
|
190
|
+
"doc": "Timestamp reflecting when this asset version was created in the source system."
|
|
191
|
+
},
|
|
192
|
+
{
|
|
193
|
+
"type": [
|
|
194
|
+
"null",
|
|
195
|
+
"com.linkedin.pegasus2avro.common.AuditStamp"
|
|
196
|
+
],
|
|
197
|
+
"name": "metadataCreatedTimestamp",
|
|
198
|
+
"default": null,
|
|
199
|
+
"doc": "Timestamp reflecting when the metadata for this version was created in DataHub"
|
|
200
|
+
},
|
|
201
|
+
{
|
|
202
|
+
"Searchable": {
|
|
203
|
+
"fieldType": "BOOLEAN",
|
|
204
|
+
"queryByDefault": false
|
|
205
|
+
},
|
|
206
|
+
"type": [
|
|
207
|
+
"null",
|
|
208
|
+
"boolean"
|
|
209
|
+
],
|
|
210
|
+
"name": "isLatest",
|
|
211
|
+
"default": null,
|
|
212
|
+
"doc": "Marks whether this version is currently the latest. Set by a side effect and should not be modified by API."
|
|
213
|
+
}
|
|
214
|
+
],
|
|
215
|
+
"doc": "Properties about a versioned asset i.e. dataset, ML Model, etc."
|
|
216
|
+
}
|
|
@@ -0,0 +1,26 @@
|
|
|
1
|
+
{
|
|
2
|
+
"type": "record",
|
|
3
|
+
"Aspect": {
|
|
4
|
+
"name": "versionSetKey",
|
|
5
|
+
"keyForEntity": "versionSet",
|
|
6
|
+
"entityCategory": "core",
|
|
7
|
+
"entityAspects": [
|
|
8
|
+
"versionSetProperties"
|
|
9
|
+
]
|
|
10
|
+
},
|
|
11
|
+
"name": "VersionSetKey",
|
|
12
|
+
"namespace": "com.linkedin.pegasus2avro.metadata.key",
|
|
13
|
+
"fields": [
|
|
14
|
+
{
|
|
15
|
+
"type": "string",
|
|
16
|
+
"name": "id",
|
|
17
|
+
"doc": "ID of the Version Set, generated from platform + asset id / name"
|
|
18
|
+
},
|
|
19
|
+
{
|
|
20
|
+
"type": "string",
|
|
21
|
+
"name": "entityType",
|
|
22
|
+
"doc": "Type of entities included in version set, limits to a single entity type between linked versioned entities"
|
|
23
|
+
}
|
|
24
|
+
],
|
|
25
|
+
"doc": "Key for a Version Set entity"
|
|
26
|
+
}
|
|
@@ -0,0 +1,49 @@
|
|
|
1
|
+
{
|
|
2
|
+
"type": "record",
|
|
3
|
+
"Aspect": {
|
|
4
|
+
"name": "versionSetProperties"
|
|
5
|
+
},
|
|
6
|
+
"name": "VersionSetProperties",
|
|
7
|
+
"namespace": "com.linkedin.pegasus2avro.versionset",
|
|
8
|
+
"fields": [
|
|
9
|
+
{
|
|
10
|
+
"Searchable": {
|
|
11
|
+
"/*": {
|
|
12
|
+
"fieldType": "TEXT",
|
|
13
|
+
"queryByDefault": true
|
|
14
|
+
}
|
|
15
|
+
},
|
|
16
|
+
"type": {
|
|
17
|
+
"type": "map",
|
|
18
|
+
"values": "string"
|
|
19
|
+
},
|
|
20
|
+
"name": "customProperties",
|
|
21
|
+
"default": {},
|
|
22
|
+
"doc": "Custom property bag."
|
|
23
|
+
},
|
|
24
|
+
{
|
|
25
|
+
"Searchable": {
|
|
26
|
+
"queryByDefault": "false"
|
|
27
|
+
},
|
|
28
|
+
"java": {
|
|
29
|
+
"class": "com.linkedin.pegasus2avro.common.urn.Urn"
|
|
30
|
+
},
|
|
31
|
+
"type": "string",
|
|
32
|
+
"name": "latest",
|
|
33
|
+
"doc": "The latest versioned entity linked to in this version set",
|
|
34
|
+
"Urn": "Urn"
|
|
35
|
+
},
|
|
36
|
+
{
|
|
37
|
+
"type": {
|
|
38
|
+
"type": "enum",
|
|
39
|
+
"name": "VersioningScheme",
|
|
40
|
+
"namespace": "com.linkedin.pegasus2avro.versionset",
|
|
41
|
+
"symbols": [
|
|
42
|
+
"ALPHANUMERIC_GENERATED_BY_DATAHUB"
|
|
43
|
+
]
|
|
44
|
+
},
|
|
45
|
+
"name": "versioningScheme",
|
|
46
|
+
"doc": "What versioning scheme is being utilized for the versioned entities sort criterion. Static once set"
|
|
47
|
+
}
|
|
48
|
+
]
|
|
49
|
+
}
|
|
File without changes
|
{acryl_datahub-0.15.0.2rc4.dist-info → acryl_datahub-0.15.0.2rc5.dist-info}/entry_points.txt
RENAMED
|
File without changes
|
|
File without changes
|