@lde/pipeline 0.6.11 → 0.6.12

package/dist/sparql/index.d.ts

@@ -1,3 +1,4 @@
 export { SparqlConstructExecutor, substituteQueryTemplates, NotSupported, readQueryFile, type ExecutableDataset, type Executor, type SparqlConstructExecuteOptions, type SparqlConstructExecutorOptions, type QuadStream, } from './executor.js';
 export { collect } from './collect.js';
+export { SparqlSelector, type SparqlSelectorOptions } from './selector.js';
 //# sourceMappingURL=index.d.ts.map

package/dist/sparql/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/sparql/index.ts"],"names":[],"mappings":"AAAA,OAAO,EACL,uBAAuB,EACvB,wBAAwB,EACxB,YAAY,EACZ,aAAa,EACb,KAAK,iBAAiB,EACtB,KAAK,QAAQ,EACb,KAAK,6BAA6B,EAClC,KAAK,8BAA8B,EACnC,KAAK,UAAU,GAChB,MAAM,eAAe,CAAC;AAEvB,OAAO,EAAE,OAAO,EAAE,MAAM,cAAc,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/sparql/index.ts"],"names":[],"mappings":"AAAA,OAAO,EACL,uBAAuB,EACvB,wBAAwB,EACxB,YAAY,EACZ,aAAa,EACb,KAAK,iBAAiB,EACtB,KAAK,QAAQ,EACb,KAAK,6BAA6B,EAClC,KAAK,8BAA8B,EACnC,KAAK,UAAU,GAChB,MAAM,eAAe,CAAC;AAEvB,OAAO,EAAE,OAAO,EAAE,MAAM,cAAc,CAAC;AAEvC,OAAO,EAAE,cAAc,EAAE,KAAK,qBAAqB,EAAE,MAAM,eAAe,CAAC"}

package/dist/sparql/index.js

@@ -1,2 +1,3 @@
 export { SparqlConstructExecutor, substituteQueryTemplates, NotSupported, readQueryFile, } from './executor.js';
 export { collect } from './collect.js';
+export { SparqlSelector } from './selector.js';

package/dist/sparql/selector.d.ts

@@ -0,0 +1,30 @@
+import { SparqlEndpointFetcher } from 'fetch-sparql-endpoint';
+import type { StageSelectorBindings, StageSelector } from '../stage.js';
+export interface SparqlSelectorOptions {
+    /** SELECT query projecting at least one named variable. A LIMIT in the query sets the default page size. */
+    query: string;
+    /** SPARQL endpoint URL. */
+    endpoint: URL;
+    /** Results per page. Overrides any LIMIT in the query. @default 10 */
+    pageSize?: number;
+    /** Custom fetcher instance. */
+    fetcher?: SparqlEndpointFetcher;
+}
+/**
+ * {@link StageSelector} that pages through SPARQL SELECT results,
+ * yielding all projected variable bindings (NamedNode values only) per row.
+ *
+ * Pagination is an internal detail — consumers iterate binding rows directly.
+ * If the query contains a LIMIT, it is used as the default page size
+ * (can be overridden by the `pageSize` option). Pagination continues
+ * until a page returns fewer results than the page size.
+ */
+export declare class SparqlSelector implements StageSelector {
+    private readonly parsed;
+    private readonly endpoint;
+    private readonly pageSize;
+    private readonly fetcher;
+    constructor(options: SparqlSelectorOptions);
+    [Symbol.asyncIterator](): AsyncIterableIterator<StageSelectorBindings>;
+}
+//# sourceMappingURL=selector.d.ts.map

package/dist/sparql/selector.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"selector.d.ts","sourceRoot":"","sources":["../../src/sparql/selector.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,qBAAqB,EAAE,MAAM,uBAAuB,CAAC;AAQ9D,OAAO,KAAK,EAAE,qBAAqB,EAAE,aAAa,EAAE,MAAM,aAAa,CAAC;AAKxE,MAAM,WAAW,qBAAqB;IACpC,4GAA4G;IAC5G,KAAK,EAAE,MAAM,CAAC;IACd,2BAA2B;IAC3B,QAAQ,EAAE,GAAG,CAAC;IACd,sEAAsE;IACtE,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,+BAA+B;IAC/B,OAAO,CAAC,EAAE,qBAAqB,CAAC;CACjC;AAED;;;;;;;;GAQG;AACH,qBAAa,cAAe,YAAW,aAAa;IAClD,OAAO,CAAC,QAAQ,CAAC,MAAM,CAAc;IACrC,OAAO,CAAC,QAAQ,CAAC,QAAQ,CAAM;IAC/B,OAAO,CAAC,QAAQ,CAAC,QAAQ,CAAS;IAClC,OAAO,CAAC,QAAQ,CAAC,OAAO,CAAwB;gBAEpC,OAAO,EAAE,qBAAqB;IAmBnC,CAAC,MAAM,CAAC,aAAa,CAAC,IAAI,qBAAqB,CAAC,qBAAqB,CAAC;CAkC9E"}

package/dist/sparql/selector.js

@@ -0,0 +1,57 @@
+import { SparqlEndpointFetcher } from 'fetch-sparql-endpoint';
+import { Generator, Parser, } from 'sparqljs';
+const parser = new Parser();
+const generator = new Generator();
+/**
+ * {@link StageSelector} that pages through SPARQL SELECT results,
+ * yielding all projected variable bindings (NamedNode values only) per row.
+ *
+ * Pagination is an internal detail — consumers iterate binding rows directly.
+ * If the query contains a LIMIT, it is used as the default page size
+ * (can be overridden by the `pageSize` option). Pagination continues
+ * until a page returns fewer results than the page size.
+ */
+export class SparqlSelector {
+    parsed;
+    endpoint;
+    pageSize;
+    fetcher;
+    constructor(options) {
+        const parsed = parser.parse(options.query);
+        if (parsed.type !== 'query' || parsed.queryType !== 'SELECT') {
+            throw new Error('Query must be a SELECT query');
+        }
+        const variables = parsed.variables.filter(isVariableTerm);
+        if (variables.length === 0) {
+            throw new Error('Query must project at least one named variable (SELECT * is not supported)');
+        }
+        this.parsed = parsed;
+        this.endpoint = options.endpoint;
+        this.pageSize = options.pageSize ?? parsed.limit ?? 10;
+        this.fetcher = options.fetcher ?? new SparqlEndpointFetcher();
+    }
+    async *[Symbol.asyncIterator]() {
+        let offset = 0;
+        while (true) {
+            this.parsed.limit = this.pageSize;
+            this.parsed.offset = offset;
+            const paginatedQuery = generator.stringify(this.parsed);
+            const stream = (await this.fetcher.fetchBindings(this.endpoint.toString(), paginatedQuery));
+            let pageSize = 0;
+            for await (const record of stream) {
+                const row = Object.fromEntries(Object.entries(record).filter(([, term]) => term.termType === 'NamedNode'));
+                if (Object.keys(row).length > 0) {
+                    yield row;
+                    pageSize++;
+                }
+            }
+            if (pageSize === 0 || pageSize < this.pageSize) {
+                return;
+            }
+            offset += pageSize;
+        }
+    }
+}
+function isVariableTerm(v) {
+    return 'termType' in v && v.termType === 'Variable';
+}

package/dist/stage.d.ts

@@ -1,4 +1,4 @@
-import type { Quad } from '@rdfjs/types';
+import type { NamedNode, Quad } from '@rdfjs/types';
 import type { ExecutableDataset, Executor } from './sparql/executor.js';
 import { NotSupported } from './sparql/executor.js';
 export interface StageOptions {
@@ -11,4 +11,9 @@ export declare class Stage {
     constructor(options: StageOptions);
     run(dataset: ExecutableDataset): Promise<AsyncIterable<Quad> | NotSupported>;
 }
+/** A single row of variable bindings from a selector query. */
+export type StageSelectorBindings = Record<string, NamedNode>;
+/** Stage-level selector that yields variable bindings for use in executor queries. Pagination is an implementation detail. */
+export interface StageSelector extends AsyncIterable<StageSelectorBindings> {
+}
 //# sourceMappingURL=stage.d.ts.map

package/dist/stage.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"stage.d.ts","sourceRoot":"","sources":["../src/stage.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,IAAI,EAAE,MAAM,cAAc,CAAC;AACzC,OAAO,KAAK,EAAE,iBAAiB,EAAE,QAAQ,EAAE,MAAM,sBAAsB,CAAC;AACxE,OAAO,EAAE,YAAY,EAAE,MAAM,sBAAsB,CAAC;AAEpD,MAAM,WAAW,YAAY;IAC3B,IAAI,EAAE,MAAM,CAAC;IACb,SAAS,EAAE,QAAQ,GAAG,QAAQ,EAAE,CAAC;CAClC;AAED,qBAAa,KAAK;IAChB,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC;IACtB,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAa;gBAE3B,OAAO,EAAE,YAAY;IAO3B,GAAG,CACP,OAAO,EAAE,iBAAiB,GACzB,OAAO,CAAC,aAAa,CAAC,IAAI,CAAC,GAAG,YAAY,CAAC;CAgB/C"}
+{"version":3,"file":"stage.d.ts","sourceRoot":"","sources":["../src/stage.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,SAAS,EAAE,IAAI,EAAE,MAAM,cAAc,CAAC;AACpD,OAAO,KAAK,EAAE,iBAAiB,EAAE,QAAQ,EAAE,MAAM,sBAAsB,CAAC;AACxE,OAAO,EAAE,YAAY,EAAE,MAAM,sBAAsB,CAAC;AAEpD,MAAM,WAAW,YAAY;IAC3B,IAAI,EAAE,MAAM,CAAC;IACb,SAAS,EAAE,QAAQ,GAAG,QAAQ,EAAE,CAAC;CAClC;AAED,qBAAa,KAAK;IAChB,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC;IACtB,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAa;gBAE3B,OAAO,EAAE,YAAY;IAO3B,GAAG,CACP,OAAO,EAAE,iBAAiB,GACzB,OAAO,CAAC,aAAa,CAAC,IAAI,CAAC,GAAG,YAAY,CAAC;CAgB/C;AAUD,+DAA+D;AAC/D,MAAM,MAAM,qBAAqB,GAAG,MAAM,CAAC,MAAM,EAAE,SAAS,CAAC,CAAC;AAE9D,8HAA8H;AAE9H,MAAM,WAAW,aAAc,SAAQ,aAAa,CAAC,qBAAqB,CAAC;CAAG"}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@lde/pipeline",
-  "version": "0.6.11",
+  "version": "0.6.12",
   "repository": {
     "url": "https://github.com/ldengine/lde",
     "directory": "packages/pipeline"
@@ -35,15 +35,19 @@
     "!**/*.tsbuildinfo"
   ],
   "dependencies": {
-    "@lde/dataset": "0.6.5",
-    "@lde/dataset-registry-client": "0.6.11",
-    "@lde/sparql-importer": "0.2.5",
-    "@lde/sparql-server": "0.4.5",
+    "@lde/dataset": "0.6.6",
+    "@lde/dataset-registry-client": "0.6.12",
+    "@lde/sparql-importer": "0.2.6",
+    "@lde/sparql-server": "0.4.6",
     "@rdfjs/types": "^2.0.1",
     "c12": "^3.0.2",
     "fetch-sparql-endpoint": "^6.0.0",
     "filenamify-url": "^3.0.0",
     "n3": "^1.17.0",
+    "sparqljs": "^3.7.3",
     "tslib": "^2.3.0"
+  },
+  "devDependencies": {
+    "@types/sparqljs": "^3.1.12"
   }
 }