@lde/pipeline 0.6.27 → 0.6.28
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/README.md +79 -31
- package/dist/sparql/selector.d.ts +3 -4
- package/dist/sparql/selector.d.ts.map +1 -1
- package/dist/sparql/selector.js +4 -4
- package/dist/stage.d.ts +4 -5
- package/dist/stage.d.ts.map +1 -1
- package/dist/stage.js +4 -7
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -2,63 +2,111 @@
|
|
|
2
2
|
|
|
3
3
|
Framework for building RDF data processing pipelines with SPARQL.
|
|
4
4
|
|
|
5
|
-
## Features
|
|
6
|
-
|
|
7
|
-
- **Pipeline** — orchestrates steps that process DCAT datasets
|
|
8
|
-
- **PipelineBuilder** — fluent API for constructing pipelines from steps and selectors
|
|
9
|
-
- **PipelineConfig** — load pipeline configuration from YAML/JSON files
|
|
10
|
-
- **SparqlConstructExecutor** — streaming SPARQL CONSTRUCT with template substitution and variable bindings
|
|
11
|
-
- **Distribution analysis** — probe and analyze dataset distributions
|
|
12
|
-
|
|
13
5
|
## Components
|
|
14
6
|
|
|
15
7
|
A **Pipeline** consists of:
|
|
16
8
|
|
|
17
|
-
-
|
|
18
|
-
-
|
|
9
|
+
- a **Dataset Selector** that selects which datasets to process
|
|
10
|
+
- a **Distribution Resolver** that resolves each dataset to a usable SPARQL endpoint
|
|
19
11
|
- one or more **Stages**, each consisting of:
|
|
20
|
-
- an optional **Selector** that
|
|
21
|
-
- one or more **Executors** that generate triples
|
|
12
|
+
- an optional **Item Selector** that selects resources (as variable bindings) for fan-out
|
|
13
|
+
- one or more **Executors** that generate triples
|
|
22
14
|
|
|
23
15
|
### Dataset Selector
|
|
24
16
|
|
|
25
|
-
Selects datasets, either manually
|
|
17
|
+
Selects datasets, either manually or by querying a DCAT Dataset Registry:
|
|
26
18
|
|
|
27
|
-
|
|
19
|
+
```typescript
|
|
20
|
+
// From a registry
|
|
21
|
+
const selector = new RegistrySelector({
|
|
22
|
+
registry: new Client(new URL('https://example.com/sparql')),
|
|
23
|
+
});
|
|
28
24
|
|
|
29
|
-
|
|
25
|
+
// Manual
|
|
26
|
+
const selector = new ManualDatasetSelection([dataset]);
|
|
27
|
+
```
|
|
30
28
|
|
|
31
|
-
|
|
29
|
+
### Item Selector
|
|
32
30
|
|
|
33
|
-
|
|
34
|
-
If not, and a valid RDF datadump is available, that is imported to a local SPARQL server.
|
|
31
|
+
Selects resources from the distribution and fans out executor calls per batch of results. Implements the `ItemSelector` interface:
|
|
35
32
|
|
|
36
|
-
|
|
33
|
+
```typescript
|
|
34
|
+
interface ItemSelector {
|
|
35
|
+
select(distribution: Distribution): AsyncIterable<VariableBindings>;
|
|
36
|
+
}
|
|
37
|
+
```
|
|
37
38
|
|
|
38
|
-
|
|
39
|
+
The distribution is received at run time, so selectors don't need the endpoint URL at construction time. Use `SparqlItemSelector` for SPARQL-based selection with automatic pagination:
|
|
39
40
|
|
|
40
|
-
|
|
41
|
-
|
|
41
|
+
```typescript
|
|
42
|
+
new SparqlItemSelector({
|
|
43
|
+
query: 'SELECT DISTINCT ?class WHERE { ?s a ?class }',
|
|
44
|
+
});
|
|
45
|
+
```
|
|
46
|
+
|
|
47
|
+
For dynamic queries that depend on the distribution, implement `ItemSelector` directly:
|
|
48
|
+
|
|
49
|
+
```typescript
|
|
50
|
+
const itemSelector: ItemSelector = {
|
|
51
|
+
select: (distribution) => {
|
|
52
|
+
const query = buildQuery(distribution);
|
|
53
|
+
return new SparqlItemSelector({ query }).select(distribution);
|
|
54
|
+
},
|
|
55
|
+
};
|
|
56
|
+
```
|
|
42
57
|
|
|
43
58
|
### Executor
|
|
44
59
|
|
|
60
|
+
Generates RDF triples. `SparqlConstructExecutor` runs a SPARQL CONSTRUCT query with template substitution and variable bindings:
|
|
61
|
+
|
|
62
|
+
```typescript
|
|
63
|
+
const executor = new SparqlConstructExecutor({
|
|
64
|
+
query: 'CONSTRUCT { ?s ?p ?o } WHERE { ?s ?p ?o }',
|
|
65
|
+
});
|
|
66
|
+
```
|
|
67
|
+
|
|
68
|
+
### Writer
|
|
69
|
+
|
|
70
|
+
Writes generated quads to a destination:
|
|
71
|
+
|
|
72
|
+
- `SparqlUpdateWriter` — writes to a SPARQL endpoint via UPDATE queries
|
|
73
|
+
- `FileWriter` — writes to local files
|
|
74
|
+
|
|
45
75
|
## Usage
|
|
46
76
|
|
|
47
77
|
```typescript
|
|
48
78
|
import {
|
|
49
|
-
|
|
79
|
+
Pipeline,
|
|
80
|
+
Stage,
|
|
50
81
|
SparqlConstructExecutor,
|
|
51
|
-
|
|
82
|
+
SparqlItemSelector,
|
|
83
|
+
SparqlUpdateWriter,
|
|
84
|
+
ManualDatasetSelection,
|
|
85
|
+
SparqlDistributionResolver,
|
|
52
86
|
} from '@lde/pipeline';
|
|
53
87
|
|
|
54
|
-
|
|
55
|
-
|
|
56
|
-
|
|
57
|
-
|
|
58
|
-
|
|
59
|
-
|
|
88
|
+
const pipeline = new Pipeline({
|
|
89
|
+
name: 'example',
|
|
90
|
+
datasetSelector: new ManualDatasetSelection([dataset]),
|
|
91
|
+
distributionResolver: new SparqlDistributionResolver(),
|
|
92
|
+
stages: [
|
|
93
|
+
new Stage({
|
|
94
|
+
name: 'per-class',
|
|
95
|
+
itemSelector: new SparqlItemSelector({
|
|
96
|
+
query: 'SELECT DISTINCT ?class WHERE { ?s a ?class }',
|
|
97
|
+
}),
|
|
98
|
+
executors: new SparqlConstructExecutor({
|
|
99
|
+
query:
|
|
100
|
+
'CONSTRUCT { ?class a <http://example.org/Class> } WHERE { ?s a ?class }',
|
|
101
|
+
}),
|
|
102
|
+
}),
|
|
103
|
+
],
|
|
104
|
+
writer: new SparqlUpdateWriter({
|
|
105
|
+
endpoint: new URL('http://localhost:7200/repositories/lde/statements'),
|
|
106
|
+
}),
|
|
60
107
|
});
|
|
61
|
-
|
|
108
|
+
|
|
109
|
+
await pipeline.run();
|
|
62
110
|
```
|
|
63
111
|
|
|
64
112
|
## Validation
|
|
@@ -1,11 +1,10 @@
|
|
|
1
|
+
import type { Distribution } from '@lde/dataset';
|
|
1
2
|
import { SparqlEndpointFetcher } from 'fetch-sparql-endpoint';
|
|
2
3
|
import type { ItemSelector } from '../stage.js';
|
|
3
4
|
import type { VariableBindings } from './executor.js';
|
|
4
5
|
export interface SparqlItemSelectorOptions {
|
|
5
6
|
/** SELECT query projecting at least one named variable. A LIMIT in the query sets the default page size. */
|
|
6
7
|
query: string;
|
|
7
|
-
/** SPARQL endpoint URL. */
|
|
8
|
-
endpoint: URL;
|
|
9
8
|
/** Results per page. Overrides any LIMIT in the query. @default 10 */
|
|
10
9
|
pageSize?: number;
|
|
11
10
|
/** Custom fetcher instance. */
|
|
@@ -15,6 +14,7 @@ export interface SparqlItemSelectorOptions {
|
|
|
15
14
|
* {@link ItemSelector} that pages through SPARQL SELECT results,
|
|
16
15
|
* yielding all projected variable bindings (NamedNode values only) per row.
|
|
17
16
|
*
|
|
17
|
+
* The endpoint URL comes from the {@link Distribution} passed to {@link select}.
|
|
18
18
|
* Pagination is an internal detail — consumers iterate binding rows directly.
|
|
19
19
|
* If the query contains a LIMIT, it is used as the default page size
|
|
20
20
|
* (can be overridden by the `pageSize` option). Pagination continues
|
|
@@ -22,10 +22,9 @@ export interface SparqlItemSelectorOptions {
|
|
|
22
22
|
*/
|
|
23
23
|
export declare class SparqlItemSelector implements ItemSelector {
|
|
24
24
|
private readonly parsed;
|
|
25
|
-
private readonly endpoint;
|
|
26
25
|
private readonly pageSize;
|
|
27
26
|
private readonly fetcher;
|
|
28
27
|
constructor(options: SparqlItemSelectorOptions);
|
|
29
|
-
|
|
28
|
+
select(distribution: Distribution): AsyncIterableIterator<VariableBindings>;
|
|
30
29
|
}
|
|
31
30
|
//# sourceMappingURL=selector.d.ts.map
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"selector.d.ts","sourceRoot":"","sources":["../../src/sparql/selector.ts"],"names":[],"mappings":"
|
|
1
|
+
{"version":3,"file":"selector.d.ts","sourceRoot":"","sources":["../../src/sparql/selector.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,cAAc,CAAC;AAEjD,OAAO,EAAE,qBAAqB,EAAE,MAAM,uBAAuB,CAAC;AAQ9D,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AAChD,OAAO,KAAK,EAAE,gBAAgB,EAAE,MAAM,eAAe,CAAC;AAKtD,MAAM,WAAW,yBAAyB;IACxC,4GAA4G;IAC5G,KAAK,EAAE,MAAM,CAAC;IACd,sEAAsE;IACtE,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,+BAA+B;IAC/B,OAAO,CAAC,EAAE,qBAAqB,CAAC;CACjC;AAED;;;;;;;;;GASG;AACH,qBAAa,kBAAmB,YAAW,YAAY;IACrD,OAAO,CAAC,QAAQ,CAAC,MAAM,CAAc;IACrC,OAAO,CAAC,QAAQ,CAAC,QAAQ,CAAS;IAClC,OAAO,CAAC,QAAQ,CAAC,OAAO,CAAwB;gBAEpC,OAAO,EAAE,yBAAyB;IAkBvC,MAAM,CACX,YAAY,EAAE,YAAY,GACzB,qBAAqB,CAAC,gBAAgB,CAAC;CAmC3C"}
|
package/dist/sparql/selector.js
CHANGED
|
@@ -6,6 +6,7 @@ const generator = new Generator();
|
|
|
6
6
|
* {@link ItemSelector} that pages through SPARQL SELECT results,
|
|
7
7
|
* yielding all projected variable bindings (NamedNode values only) per row.
|
|
8
8
|
*
|
|
9
|
+
* The endpoint URL comes from the {@link Distribution} passed to {@link select}.
|
|
9
10
|
* Pagination is an internal detail — consumers iterate binding rows directly.
|
|
10
11
|
* If the query contains a LIMIT, it is used as the default page size
|
|
11
12
|
* (can be overridden by the `pageSize` option). Pagination continues
|
|
@@ -13,7 +14,6 @@ const generator = new Generator();
|
|
|
13
14
|
*/
|
|
14
15
|
export class SparqlItemSelector {
|
|
15
16
|
parsed;
|
|
16
|
-
endpoint;
|
|
17
17
|
pageSize;
|
|
18
18
|
fetcher;
|
|
19
19
|
constructor(options) {
|
|
@@ -26,17 +26,17 @@ export class SparqlItemSelector {
|
|
|
26
26
|
throw new Error('Query must project at least one named variable (SELECT * is not supported)');
|
|
27
27
|
}
|
|
28
28
|
this.parsed = parsed;
|
|
29
|
-
this.endpoint = options.endpoint;
|
|
30
29
|
this.pageSize = options.pageSize ?? parsed.limit ?? 10;
|
|
31
30
|
this.fetcher = options.fetcher ?? new SparqlEndpointFetcher();
|
|
32
31
|
}
|
|
33
|
-
async *
|
|
32
|
+
async *select(distribution) {
|
|
33
|
+
const endpoint = distribution.accessUrl;
|
|
34
34
|
let offset = 0;
|
|
35
35
|
while (true) {
|
|
36
36
|
this.parsed.limit = this.pageSize;
|
|
37
37
|
this.parsed.offset = offset;
|
|
38
38
|
const paginatedQuery = generator.stringify(this.parsed);
|
|
39
|
-
const stream = (await this.fetcher.fetchBindings(
|
|
39
|
+
const stream = (await this.fetcher.fetchBindings(endpoint.toString(), paginatedQuery));
|
|
40
40
|
let pageSize = 0;
|
|
41
41
|
for await (const record of stream) {
|
|
42
42
|
const row = Object.fromEntries(Object.entries(record).filter(([, term]) => term.termType === 'NamedNode'));
|
package/dist/stage.d.ts
CHANGED
|
@@ -2,12 +2,10 @@ import { Dataset, Distribution } from '@lde/dataset';
|
|
|
2
2
|
import type { Executor, VariableBindings } from './sparql/executor.js';
|
|
3
3
|
import { NotSupported } from './sparql/executor.js';
|
|
4
4
|
import type { Writer } from './writer/writer.js';
|
|
5
|
-
/** An item selector, or a factory that receives the runtime distribution. */
|
|
6
|
-
export type ItemSelectorInput = ItemSelector | ((distribution: Distribution) => ItemSelector);
|
|
7
5
|
export interface StageOptions {
|
|
8
6
|
name: string;
|
|
9
7
|
executors: Executor | Executor[];
|
|
10
|
-
itemSelector?:
|
|
8
|
+
itemSelector?: ItemSelector;
|
|
11
9
|
/** Maximum number of bindings per executor call. @default 10 */
|
|
12
10
|
batchSize?: number;
|
|
13
11
|
/** Maximum concurrent in-flight executor batches. @default 10 */
|
|
@@ -22,7 +20,7 @@ export declare class Stage {
|
|
|
22
20
|
readonly name: string;
|
|
23
21
|
readonly stages: readonly Stage[];
|
|
24
22
|
private readonly executors;
|
|
25
|
-
private readonly
|
|
23
|
+
private readonly itemSelector?;
|
|
26
24
|
private readonly batchSize;
|
|
27
25
|
private readonly maxConcurrency;
|
|
28
26
|
constructor(options: StageOptions);
|
|
@@ -31,6 +29,7 @@ export declare class Stage {
|
|
|
31
29
|
private executeAll;
|
|
32
30
|
}
|
|
33
31
|
/** Selects items (as variable bindings) for executors to process. Pagination is an implementation detail. */
|
|
34
|
-
export interface ItemSelector
|
|
32
|
+
export interface ItemSelector {
|
|
33
|
+
select(distribution: Distribution): AsyncIterable<VariableBindings>;
|
|
35
34
|
}
|
|
36
35
|
//# sourceMappingURL=stage.d.ts.map
|
package/dist/stage.d.ts.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"stage.d.ts","sourceRoot":"","sources":["../src/stage.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,OAAO,EAAE,YAAY,EAAE,MAAM,cAAc,CAAC;AAErD,OAAO,KAAK,EAAE,QAAQ,EAAE,gBAAgB,EAAE,MAAM,sBAAsB,CAAC;AACvE,OAAO,EAAE,YAAY,EAAE,MAAM,sBAAsB,CAAC;AAEpD,OAAO,KAAK,EAAE,MAAM,EAAE,MAAM,oBAAoB,CAAC;AAGjD,
|
|
1
|
+
{"version":3,"file":"stage.d.ts","sourceRoot":"","sources":["../src/stage.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,OAAO,EAAE,YAAY,EAAE,MAAM,cAAc,CAAC;AAErD,OAAO,KAAK,EAAE,QAAQ,EAAE,gBAAgB,EAAE,MAAM,sBAAsB,CAAC;AACvE,OAAO,EAAE,YAAY,EAAE,MAAM,sBAAsB,CAAC;AAEpD,OAAO,KAAK,EAAE,MAAM,EAAE,MAAM,oBAAoB,CAAC;AAGjD,MAAM,WAAW,YAAY;IAC3B,IAAI,EAAE,MAAM,CAAC;IACb,SAAS,EAAE,QAAQ,GAAG,QAAQ,EAAE,CAAC;IACjC,YAAY,CAAC,EAAE,YAAY,CAAC;IAC5B,gEAAgE;IAChE,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,iEAAiE;IACjE,cAAc,CAAC,EAAE,MAAM,CAAC;IACxB,uDAAuD;IACvD,MAAM,CAAC,EAAE,KAAK,EAAE,CAAC;CAClB;AAED,MAAM,WAAW,UAAU;IACzB,UAAU,CAAC,EAAE,CAAC,iBAAiB,EAAE,MAAM,EAAE,cAAc,EAAE,MAAM,KAAK,IAAI,CAAC;CAC1E;AAED,qBAAa,KAAK;IAChB,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC;IACtB,QAAQ,CAAC,MAAM,EAAE,SAAS,KAAK,EAAE,CAAC;IAClC,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAa;IACvC,OAAO,CAAC,QAAQ,CAAC,YAAY,CAAC,CAAe;IAC7C,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAS;IACnC,OAAO,CAAC,QAAQ,CAAC,cAAc,CAAS;gBAE5B,OAAO,EAAE,YAAY;IAW3B,GAAG,CACP,OAAO,EAAE,OAAO,EAChB,YAAY,EAAE,YAAY,EAC1B,MAAM,EAAE,MAAM,EACd,OAAO,CAAC,EAAE,UAAU,GACnB,OAAO,CAAC,YAAY,GAAG,IAAI,CAAC;YAmBjB,eAAe;YA+Gf,UAAU;CAqBzB;AAUD,6GAA6G;AAC7G,MAAM,WAAW,YAAY;IAC3B,MAAM,CAAC,YAAY,EAAE,YAAY,GAAG,aAAa,CAAC,gBAAgB,CAAC,CAAC;CACrE"}
|
package/dist/stage.js
CHANGED
|
@@ -5,7 +5,7 @@ export class Stage {
|
|
|
5
5
|
name;
|
|
6
6
|
stages;
|
|
7
7
|
executors;
|
|
8
|
-
|
|
8
|
+
itemSelector;
|
|
9
9
|
batchSize;
|
|
10
10
|
maxConcurrency;
|
|
11
11
|
constructor(options) {
|
|
@@ -14,16 +14,13 @@ export class Stage {
|
|
|
14
14
|
this.executors = Array.isArray(options.executors)
|
|
15
15
|
? options.executors
|
|
16
16
|
: [options.executors];
|
|
17
|
-
this.
|
|
17
|
+
this.itemSelector = options.itemSelector;
|
|
18
18
|
this.batchSize = options.batchSize ?? 10;
|
|
19
19
|
this.maxConcurrency = options.maxConcurrency ?? 10;
|
|
20
20
|
}
|
|
21
21
|
async run(dataset, distribution, writer, options) {
|
|
22
|
-
if (this.
|
|
23
|
-
|
|
24
|
-
? this.itemSelectorInput(distribution)
|
|
25
|
-
: this.itemSelectorInput;
|
|
26
|
-
return this.runWithSelector(selector, dataset, distribution, writer, options);
|
|
22
|
+
if (this.itemSelector) {
|
|
23
|
+
return this.runWithSelector(this.itemSelector.select(distribution), dataset, distribution, writer, options);
|
|
27
24
|
}
|
|
28
25
|
const streams = await this.executeAll(dataset, distribution);
|
|
29
26
|
if (streams instanceof NotSupported) {
|