keepset 0.36.0 → 0.37.0

package/dist/{bundled-BrTVG8j1.js → bundled-CRr7GPD3.js}

@@ -528,7 +528,17 @@ const InnerArchiveSchema = z.union([z.object({ kind: z.literal("gzip") }).strict
 	kind: FileArchiveKindSchema,
 	member: ArchiveMemberSchema
 }).strict()]);
-/** How the downloaded bytes unpack into the file the parser reads, or for a collection the files it reads one by one. A `members` selector gives every member it matches, at least one, in path order. `gzip` and `bzip2` are one compressed file, and `tar.gz` and `tar.bz2` a tar archive compressed with them. A member that is an archive in its turn (a contest bundle that ships its problems as a zip inside the zip) is unpacked by `inner`. A workbook (`xlsx`) is read as the CSV of its named sheet, with dates as ISO 8601 text, for the `csv-table` parser. A point shapefile (`shapefile`) is a zip whose `member` selects the `.shp`, read with the `.dbf`, `.prj` and `.cpg` of the same name beside it as the CSV of its attribute table with each point's stored coordinates and its WGS84 position, for the `csv-table` parser. */
+/** The name of a part of a document read from several files, which the parser gives its meaning: the split a triples file holds (`train`, `valid`, `test`), or which of an HGB dataset's files it is (`nodes`, `links`). */
+const PartNameSchema = z.string().regex(/^[a-z][a-z0-9]*$/, { message: "must be a lowercase name of letters and digits, such as train" });
+/** An archive one document is read from several files of: each part names the one member it is read from, and the parser reads the parts together in the order given. */
+const PartsArchiveSchema = z.object({
+	kind: FileArchiveKindSchema,
+	parts: z.array(z.object({
+		name: PartNameSchema,
+		member: ArchiveMemberSchema
+	}).strict()).min(1).refine((parts) => new Set(parts.map((part) => part.name)).size === parts.length, { message: "each part must have its own name" })
+}).strict();
+/** How the downloaded bytes unpack into the file the parser reads, the named files a parser reads together into one document (`parts`), or for a collection the files it reads one by one. A `members` selector gives every member it matches, at least one, in path order. `gzip` and `bzip2` are one compressed file, and `tar.gz` and `tar.bz2` a tar archive compressed with them. A member that is an archive in its turn (a contest bundle that ships its problems as a zip inside the zip) is unpacked by `inner`. A workbook (`xlsx`) is read as the CSV of its named sheet, with dates as ISO 8601 text, for the `csv-table` parser. A point shapefile (`shapefile`) is a zip whose `member` selects the `.shp`, read with the `.dbf`, `.prj` and `.cpg` of the same name beside it as the CSV of its attribute table with each point's stored coordinates and its WGS84 position, for the `csv-table` parser. */
 const ArchiveSchema = z.union([
 	z.object({ kind: z.literal("none") }).strict(),
 	z.object({ kind: z.literal("gzip") }).strict(),
@@ -539,6 +549,7 @@ const ArchiveSchema = z.union([
 		inner: InnerArchiveSchema.optional()
 	}).strict(),
 	CollectionArchiveSchema,
+	PartsArchiveSchema,
 	z.object({
 		kind: z.literal("xlsx"),
 		sheet: z.string().min(1)
@@ -601,6 +612,50 @@ const ParserSchema = z.discriminatedUnion("format", [
 		/** Whether the fourth column is read as the edge's time. */
 		timestamped: z.boolean()
 	}).strict(),
+	z.object({
+		format: z.literal("user-item-counts"),
+		encoding: z.literal("utf-8"),
+		/** What separates the fields of a line. */
+		delimiter: z.enum(["tab"]),
+		/** The header lines before the first count, counted as the file's lines. */
+		skipLines: z.int().min(0),
+		/** What the counted items are, which is their node type and fixes the prefix of their ids. */
+		itemType: z.enum(["artist"])
+	}).strict(),
+	z.object({
+		format: z.literal("kg-triples"),
+		encoding: EncodingSchema,
+		/** What separates the three fields of a line. */
+		delimiter: z.enum(["tab", "space"]),
+		/** Which field holds the relation: the middle one, or the last, as OpenKE's id files write it. */
+		order: z.enum(["head-relation-tail", "head-tail-relation"]),
+		/** What each file's first line is: a triple, a header naming the columns (skipped), or the number of triples the file holds (checked). */
+		header: z.enum([
+			"none",
+			"names",
+			"count"
+		]),
+		/** Whether every field is wrapped in double quotes, which are not part of the id. */
+		quoted: z.boolean()
+	}).strict(),
+	z.object({
+		format: z.literal("hgb"),
+		encoding: z.literal("utf-8"),
+		/** The names of the node types, in the order of their codes in `node.dat`, as the dataset's `info.dat` gives them. */
+		nodeTypes: z.array(z.string().min(1)).min(1),
+		/** The link types, in the order of their codes in `link.dat`, each with its name and the node types its links run from and to, as `info.dat` gives them. */
+		linkTypes: z.array(z.object({
+			name: z.string().min(1),
+			source: z.string().min(1),
+			target: z.string().min(1)
+		}).strict()).min(1)
+	}).strict().refine((parser) => new Set(parser.nodeTypes).size === parser.nodeTypes.length, {
+		message: "each node type must have its own name",
+		path: ["nodeTypes"]
+	}).refine((parser) => parser.linkTypes.every((link) => parser.nodeTypes.includes(link.source) && parser.nodeTypes.includes(link.target)), {
+		message: "a link type's source and target must be node types the recipe names",
+		path: ["linkTypes"]
+	}),
 	z.object({
 		format: z.literal("gml"),
 		encoding: EncodingSchema
@@ -719,13 +774,30 @@ const PagingSchema = z.discriminatedUnion("kind", [
 	}).strict()
 ]);
 /** How to obtain the dataset: a fetchable recipe naming the source (by index into `sources`), the archive layout and the parser, or a statement that no declarative recipe exists yet and why. */
+/** The parser formats that read a document from named parts, and the part names each requires (none for a format whose part names are its own, such as the splits of a triples download). */
+const PART_FORMATS = /* @__PURE__ */ new Map([["kg-triples", void 0], ["hgb", ["nodes", "links"]]]);
 const RecipeSchema = z.discriminatedUnion("status", [z.object({
 	status: z.literal("fetchable"),
 	source: z.int().min(0),
 	paging: PagingSchema.optional(),
 	archive: ArchiveSchema,
 	parser: ParserSchema
-}).strict(), z.object({
+}).strict().superRefine((recipe, ctx) => {
+	const required = PART_FORMATS.get(recipe.parser.format);
+	if ("parts" in recipe.archive && !PART_FORMATS.has(recipe.parser.format)) ctx.addIssue({
+		code: "custom",
+		path: ["archive", "parts"],
+		message: `the ${recipe.parser.format} parser reads one file, so its archive selects a member rather than parts`
+	});
+	if (required !== void 0) {
+		const names = "parts" in recipe.archive ? recipe.archive.parts.map((part) => part.name) : [];
+		if (names.length !== required.length || !required.every((name) => names.includes(name))) ctx.addIssue({
+			code: "custom",
+			path: ["archive"],
+			message: `the ${recipe.parser.format} parser reads the parts ${required.join(" and ")} from a parts archive`
+		});
+	}
+}), z.object({
 	status: z.literal("not-fetchable"),
 	reason: z.string().min(1)
 }).strict()]);
@@ -1108,7 +1180,7 @@ var catalogue_default = {
 		"id": "keepset",
 		"name": "keepset",
 		"description": "The keepset catalogue: datasets with machine-readable recipes that fetch each one and normalise it into its category's target shape.",
-		"version": "1.27.0",
+		"version": "1.28.0",
 		"repository": "https://github.com/Mearman/keepset",
 		"namespaceAliases": {
 			"networkrepository": ["nr"],
@@ -1163,6 +1235,75 @@ var catalogue_default = {
 			],
 			"oneWay": []
 		},
+		"crosswalks/graph-v1/relations-split.json": {
+			"source": "keepset/relations-split",
+			"target": "graph/v1",
+			"rules": [
+				{
+					"op": "move",
+					"from": "meta",
+					"to": "meta"
+				},
+				{
+					"op": "move",
+					"from": "nodes[].id",
+					"to": "nodes[].id"
+				},
+				{
+					"op": "move",
+					"from": "edges[].source",
+					"to": "edges[].source"
+				},
+				{
+					"op": "move",
+					"from": "edges[].target",
+					"to": "edges[].target"
+				},
+				{
+					"op": "move",
+					"from": "edges[].type",
+					"to": "edges[].type"
+				},
+				{
+					"op": "move",
+					"from": "edges[].split",
+					"to": "edges[].split"
+				}
+			],
+			"oneWay": []
+		},
+		"crosswalks/graph-v1/relations.json": {
+			"source": "keepset/relations",
+			"target": "graph/v1",
+			"rules": [
+				{
+					"op": "move",
+					"from": "meta",
+					"to": "meta"
+				},
+				{
+					"op": "move",
+					"from": "nodes[].id",
+					"to": "nodes[].id"
+				},
+				{
+					"op": "move",
+					"from": "edges[].source",
+					"to": "edges[].source"
+				},
+				{
+					"op": "move",
+					"from": "edges[].target",
+					"to": "edges[].target"
+				},
+				{
+					"op": "move",
+					"from": "edges[].type",
+					"to": "edges[].type"
+				}
+			],
+			"oneWay": []
+		},
 		"crosswalks/graph-v1/timestamped.json": {
 			"source": "keepset/timestamped",
 			"target": "graph/v1",
@@ -1222,6 +1363,85 @@ var catalogue_default = {
 			],
 			"oneWay": []
 		},
+		"crosswalks/graph-v1/typed-weighted.json": {
+			"source": "keepset/typed-weighted",
+			"target": "graph/v1",
+			"rules": [
+				{
+					"op": "move",
+					"from": "meta",
+					"to": "meta"
+				},
+				{
+					"op": "move",
+					"from": "nodes[].id",
+					"to": "nodes[].id"
+				},
+				{
+					"op": "move",
+					"from": "edges[].source",
+					"to": "edges[].source"
+				},
+				{
+					"op": "move",
+					"from": "edges[].target",
+					"to": "edges[].target"
+				},
+				{
+					"op": "move",
+					"from": "nodes[].type",
+					"to": "nodes[].type"
+				},
+				{
+					"op": "move",
+					"from": "edges[].type",
+					"to": "edges[].type"
+				},
+				{
+					"op": "move",
+					"from": "edges[].weight",
+					"to": "edges[].weight"
+				}
+			],
+			"oneWay": []
+		},
+		"crosswalks/graph-v1/typed.json": {
+			"source": "keepset/typed",
+			"target": "graph/v1",
+			"rules": [
+				{
+					"op": "move",
+					"from": "meta",
+					"to": "meta"
+				},
+				{
+					"op": "move",
+					"from": "nodes[].id",
+					"to": "nodes[].id"
+				},
+				{
+					"op": "move",
+					"from": "edges[].source",
+					"to": "edges[].source"
+				},
+				{
+					"op": "move",
+					"from": "edges[].target",
+					"to": "edges[].target"
+				},
+				{
+					"op": "move",
+					"from": "nodes[].type",
+					"to": "nodes[].type"
+				},
+				{
+					"op": "move",
+					"from": "edges[].type",
+					"to": "edges[].type"
+				}
+			],
+			"oneWay": []
+		},
 		"crosswalks/graph-v1/weighted-timestamped.json": {
 			"source": "keepset/weighted-timestamped",
 			"target": "graph/v1",
@@ -1358,7 +1578,7 @@ var catalogue_default = {
 		{
 			"id": "acm-hgb",
 			"name": "Heterogeneous Graph Benchmark - ACM",
-			"description": "Heterogeneous academic network from the Heterogeneous Graph Benchmark suite. Contains papers, authors, and subject categories as node types with 3 relation types: writes (author-paper), published-in (paper-subject), and related subject links. Moderate-size heterogeneous graph used for benchmarking heterogeneous graph neural networks on node classification.",
+			"description": "The ACM network as the Heterogeneous Graph Benchmark suite ships it, the same file as the acm entry. Contains papers, authors, subjects and terms, linked by citation, reference, authorship, subject and term relations, each listed in both directions. Used for benchmarking heterogeneous graph neural networks on node classification.",
 			"tags": [],
 			"category": "graph/v1",
 			"sources": [{
@@ -1369,10 +1589,72 @@ var catalogue_default = {
 			"licence": { "kind": "unrecorded" },
 			"citations": [{ "text": "Lv et al. (2021). Are We Really Making Much Progress? A Worrying Analysis of Recent Neural Relation Extraction Models." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (hgb-zip: format hgb-zip has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "zip",
+					"parts": [{
+						"name": "nodes",
+						"member": { "path": "ACM/node.dat" }
+					}, {
+						"name": "links",
+						"member": { "path": "ACM/link.dat" }
+					}]
+				},
+				"parser": {
+					"format": "hgb",
+					"encoding": "utf-8",
+					"nodeTypes": [
+						"paper",
+						"author",
+						"subject",
+						"term"
+					],
+					"linkTypes": [
+						{
+							"name": "paper-cite-paper",
+							"source": "paper",
+							"target": "paper"
+						},
+						{
+							"name": "paper-ref-paper",
+							"source": "paper",
+							"target": "paper"
+						},
+						{
+							"name": "paper-author",
+							"source": "paper",
+							"target": "author"
+						},
+						{
+							"name": "author-paper",
+							"source": "author",
+							"target": "paper"
+						},
+						{
+							"name": "paper-subject",
+							"source": "paper",
+							"target": "subject"
+						},
+						{
+							"name": "subject-paper",
+							"source": "subject",
+							"target": "paper"
+						},
+						{
+							"name": "paper-term",
+							"source": "paper",
+							"target": "term"
+						},
+						{
+							"name": "term-paper",
+							"source": "term",
+							"target": "paper"
+						}
+					]
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/typed.json" },
 			"expectedCounts": {
 				"edges": 547872,
 				"nodes": 10942
@@ -1383,10 +1665,10 @@ var catalogue_default = {
 				"cyclic": true,
 				"density": .004575930276299103,
 				"directed": true,
-				"edgeTypes": 1,
+				"edgeTypes": 8,
 				"heterogeneous": true,
 				"multigraph": true,
-				"nodeTypes": 1,
+				"nodeTypes": 4,
 				"rawBytes": 0,
 				"selfLoops": true,
 				"weighted": false
@@ -1395,7 +1677,7 @@ var catalogue_default = {
 		{
 			"id": "acm",
 			"name": "ACM",
-			"description": "Heterogeneous academic network from ACM digital library. Contains papers, authors, and subjects with authorship and subject-assignment relations. Used as a standard benchmark for heterogeneous graph neural networks.",
+			"description": "Heterogeneous academic network from the ACM digital library, as the Heterogeneous Graph Benchmark packages it. Contains papers, authors, subjects and terms, linked by citation, reference, authorship, subject and term relations, each listed in both directions. Used as a standard benchmark for heterogeneous graph neural networks.",
 			"domain": "Academic heterogeneous",
 			"tags": ["heterogeneous"],
 			"category": "graph/v1",
@@ -1407,10 +1689,72 @@ var catalogue_default = {
 			"licence": { "kind": "unrecorded" },
 			"citations": [{ "text": "Xinyu Lv, Tong Zhao, Yozen Liu, Neil Shah, Yong Li (2021). Are we really making much progress? Revisiting, benchmarking and refining heterogeneous graph neural networks. Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (hgb: format hgb has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "zip",
+					"parts": [{
+						"name": "nodes",
+						"member": { "path": "ACM/node.dat" }
+					}, {
+						"name": "links",
+						"member": { "path": "ACM/link.dat" }
+					}]
+				},
+				"parser": {
+					"format": "hgb",
+					"encoding": "utf-8",
+					"nodeTypes": [
+						"paper",
+						"author",
+						"subject",
+						"term"
+					],
+					"linkTypes": [
+						{
+							"name": "paper-cite-paper",
+							"source": "paper",
+							"target": "paper"
+						},
+						{
+							"name": "paper-ref-paper",
+							"source": "paper",
+							"target": "paper"
+						},
+						{
+							"name": "paper-author",
+							"source": "paper",
+							"target": "author"
+						},
+						{
+							"name": "author-paper",
+							"source": "author",
+							"target": "paper"
+						},
+						{
+							"name": "paper-subject",
+							"source": "paper",
+							"target": "subject"
+						},
+						{
+							"name": "subject-paper",
+							"source": "subject",
+							"target": "paper"
+						},
+						{
+							"name": "paper-term",
+							"source": "paper",
+							"target": "term"
+						},
+						{
+							"name": "term-paper",
+							"source": "term",
+							"target": "paper"
+						}
+					]
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/typed.json" },
 			"expectedCounts": {
 				"edges": 547872,
 				"nodes": 10942
@@ -1421,10 +1765,10 @@ var catalogue_default = {
 				"cyclic": true,
 				"density": .0045759971008822835,
 				"directed": false,
-				"edgeTypes": 1,
+				"edgeTypes": 8,
 				"heterogeneous": true,
 				"multigraph": true,
-				"nodeTypes": 2,
+				"nodeTypes": 4,
 				"rawBytes": 0,
 				"selfLoops": true,
 				"weighted": false
@@ -5769,10 +6113,62 @@ var catalogue_default = {
 			"licence": { "kind": "unrecorded" },
 			"citations": [{ "text": "Xinyu Lv, Tong Zhao, Yozen Liu, Neil Shah, Yong Li (2021). Are we really making much progress? Revisiting, benchmarking and refining heterogeneous graph neural networks. Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (hgb: format hgb has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "zip",
+					"parts": [{
+						"name": "nodes",
+						"member": { "path": "DBLP/node.dat" }
+					}, {
+						"name": "links",
+						"member": { "path": "DBLP/link.dat" }
+					}]
+				},
+				"parser": {
+					"format": "hgb",
+					"encoding": "utf-8",
+					"nodeTypes": [
+						"author",
+						"paper",
+						"term",
+						"venue"
+					],
+					"linkTypes": [
+						{
+							"name": "author-paper",
+							"source": "author",
+							"target": "paper"
+						},
+						{
+							"name": "paper-term",
+							"source": "paper",
+							"target": "term"
+						},
+						{
+							"name": "paper-venue",
+							"source": "paper",
+							"target": "venue"
+						},
+						{
+							"name": "paper-author",
+							"source": "paper",
+							"target": "author"
+						},
+						{
+							"name": "term-paper",
+							"source": "term",
+							"target": "paper"
+						},
+						{
+							"name": "venue-paper",
+							"source": "venue",
+							"target": "paper"
+						}
+					]
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/typed.json" },
 			"expectedCounts": {
 				"edges": 239566,
 				"nodes": 26128
@@ -5783,10 +6179,10 @@ var catalogue_default = {
 				"cyclic": true,
 				"density": .00035093725028794413,
 				"directed": false,
-				"edgeTypes": 1,
+				"edgeTypes": 6,
 				"heterogeneous": true,
 				"multigraph": true,
-				"nodeTypes": 2,
+				"nodeTypes": 4,
 				"rawBytes": 0,
 				"selfLoops": false,
 				"weighted": false
@@ -6923,23 +7319,48 @@ var catalogue_default = {
 			"licence": { "kind": "unrecorded" },
 			"citations": [{ "text": "Kristina Toutanova, Danqi Chen, Patrick Pantel, Hoifung Poon, Pallavi Choudhury, Michael Gamon (2015). Representing Text for Joint Embedding of Text and Knowledge Bases. Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing (EMNLP)." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (kge-triples-tar: format kge-triples-tar has a pipeline-specific parser with no declarative keepset form yet; landing-page: a landing page, not a download; landing-page: the download is a web page, not a data file)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "tar.gz",
+					"parts": [
+						{
+							"name": "train",
+							"member": { "path": "fb15k-237/train.txt" }
+						},
+						{
+							"name": "valid",
+							"member": { "path": "fb15k-237/valid.txt" }
+						},
+						{
+							"name": "test",
+							"member": { "path": "fb15k-237/test.txt" }
+						}
+					]
+				},
+				"parser": {
+					"format": "kg-triples",
+					"encoding": "utf-8",
+					"delimiter": "tab",
+					"order": "head-relation-tail",
+					"header": "none",
+					"quoted": false
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/relations-split.json" },
 			"expectedCounts": {
-				"edges": 283869,
-				"nodes": 14543
+				"edges": 310116,
+				"nodes": 14541
 			},
 			"properties": {
-				"components": 116,
+				"components": 6,
 				"connected": false,
-				"cyclic": false,
+				"cyclic": true,
 				"density": .004329004329004329,
 				"directed": true,
-				"edgeTypes": 1,
+				"edgeTypes": 237,
 				"heterogeneous": true,
-				"multigraph": false,
+				"multigraph": true,
 				"nodeTypes": 1,
 				"rawBytes": 0,
 				"selfLoops": true,
@@ -7430,24 +7851,230 @@ var catalogue_default = {
 			"licence": { "kind": "unrecorded" },
 			"citations": [{ "text": "Xinyu Lv, Tong Zhao, Yozen Liu, Neil Shah, Yong Li (2021). Are we really making much progress? Revisiting, benchmarking and refining heterogeneous graph neural networks. Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (hgb: format hgb has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "zip",
+					"parts": [{
+						"name": "nodes",
+						"member": { "path": "Freebase/node.dat" }
+					}, {
+						"name": "links",
+						"member": { "path": "Freebase/link.dat" }
+					}]
+				},
+				"parser": {
+					"format": "hgb",
+					"encoding": "utf-8",
+					"nodeTypes": [
+						"BOOK",
+						"FILM",
+						"MUSIC",
+						"SPORTS",
+						"PEOPLE",
+						"LOCATION",
+						"ORGANIZATION",
+						"BUSINESS"
+					],
+					"linkTypes": [
+						{
+							"name": "BOOK-and-BOOK",
+							"source": "BOOK",
+							"target": "BOOK"
+						},
+						{
+							"name": "BOOK-to-FILM",
+							"source": "BOOK",
+							"target": "FILM"
+						},
+						{
+							"name": "BOOK-on-SPORTS",
+							"source": "BOOK",
+							"target": "SPORTS"
+						},
+						{
+							"name": "BOOK-on-LOCATION",
+							"source": "BOOK",
+							"target": "LOCATION"
+						},
+						{
+							"name": "BOOK-about-ORGANIZATION",
+							"source": "BOOK",
+							"target": "ORGANIZATION"
+						},
+						{
+							"name": "FILM-and-FILM",
+							"source": "FILM",
+							"target": "FILM"
+						},
+						{
+							"name": "MUSIC-in-BOOK",
+							"source": "MUSIC",
+							"target": "BOOK"
+						},
+						{
+							"name": "MUSIC-in-FILM",
+							"source": "MUSIC",
+							"target": "FILM"
+						},
+						{
+							"name": "MUSIC-and-MUSIC",
+							"source": "MUSIC",
+							"target": "MUSIC"
+						},
+						{
+							"name": "MUSIC-for-SPORTS",
+							"source": "MUSIC",
+							"target": "SPORTS"
+						},
+						{
+							"name": "MUSIC-on-LOCATION",
+							"source": "MUSIC",
+							"target": "LOCATION"
+						},
+						{
+							"name": "SPORTS-in-FILM",
+							"source": "SPORTS",
+							"target": "FILM"
+						},
+						{
+							"name": "SPORTS-and-SPORTS",
+							"source": "SPORTS",
+							"target": "SPORTS"
+						},
+						{
+							"name": "SPORTS-on-LOCATION",
+							"source": "SPORTS",
+							"target": "LOCATION"
+						},
+						{
+							"name": "PEOPLE-to-BOOK",
+							"source": "PEOPLE",
+							"target": "BOOK"
+						},
+						{
+							"name": "PEOPLE-to-FILM",
+							"source": "PEOPLE",
+							"target": "FILM"
+						},
+						{
+							"name": "PEOPLE-to-MUSIC",
+							"source": "PEOPLE",
+							"target": "MUSIC"
+						},
+						{
+							"name": "PEOPLE-to-SPORTS",
+							"source": "PEOPLE",
+							"target": "SPORTS"
+						},
+						{
+							"name": "PEOPLE-and-PEOPLE",
+							"source": "PEOPLE",
+							"target": "PEOPLE"
+						},
+						{
+							"name": "PEOPLE-on-LOCATION",
+							"source": "PEOPLE",
+							"target": "LOCATION"
+						},
+						{
+							"name": "PEOPLE-in-ORGANIZATION",
+							"source": "PEOPLE",
+							"target": "ORGANIZATION"
+						},
+						{
+							"name": "PEOPLE-in-BUSINESS",
+							"source": "PEOPLE",
+							"target": "BUSINESS"
+						},
+						{
+							"name": "LOCATION-in-FILM",
+							"source": "LOCATION",
+							"target": "FILM"
+						},
+						{
+							"name": "LOCATION-and-LOCATION",
+							"source": "LOCATION",
+							"target": "LOCATION"
+						},
+						{
+							"name": "ORGANIZATION-in-FILM",
+							"source": "ORGANIZATION",
+							"target": "FILM"
+						},
+						{
+							"name": "ORGANIZATION-to-MUSIC",
+							"source": "ORGANIZATION",
+							"target": "MUSIC"
+						},
+						{
+							"name": "ORGANIZATION-to-SPORTS",
+							"source": "ORGANIZATION",
+							"target": "SPORTS"
+						},
+						{
+							"name": "ORGANIZATION-on-LOCATION",
+							"source": "ORGANIZATION",
+							"target": "LOCATION"
+						},
+						{
+							"name": "ORGANIZATION-and-ORGANIZATION",
+							"source": "ORGANIZATION",
+							"target": "ORGANIZATION"
+						},
+						{
+							"name": "ORGANIZATION-for-BUSINESS",
+							"source": "ORGANIZATION",
+							"target": "BUSINESS"
+						},
+						{
+							"name": "BUSINESS-about-BOOK",
+							"source": "BUSINESS",
+							"target": "BOOK"
+						},
+						{
+							"name": "BUSINESS-about-FILM",
+							"source": "BUSINESS",
+							"target": "FILM"
+						},
+						{
+							"name": "BUSINESS-about-MUSIC",
+							"source": "BUSINESS",
+							"target": "MUSIC"
+						},
+						{
+							"name": "BUSINESS-about-SPORTS",
+							"source": "BUSINESS",
+							"target": "SPORTS"
+						},
+						{
+							"name": "BUSINESS-on-LOCATION",
+							"source": "BUSINESS",
+							"target": "LOCATION"
+						},
+						{
+							"name": "BUSINESS-and-BUSINESS",
+							"source": "BUSINESS",
+							"target": "BUSINESS"
+						}
+					]
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/typed.json" },
 			"expectedCounts": {
 				"edges": 1057688,
 				"nodes": 180098
 			},
 			"properties": {
-				"components": 353,
+				"components": 378,
 				"connected": false,
 				"cyclic": true,
 				"density": 4721988619699187e-20,
 				"directed": true,
-				"edgeTypes": 1,
+				"edgeTypes": 36,
 				"heterogeneous": true,
 				"multigraph": false,
-				"nodeTypes": 2,
+				"nodeTypes": 8,
 				"rawBytes": 0,
 				"selfLoops": true,
 				"weighted": false
@@ -8197,30 +8824,29 @@ var catalogue_default = {
 				"source": 0,
 				"archive": { "kind": "gzip" },
 				"parser": {
-					"format": "edge-list",
+					"format": "kg-triples",
 					"encoding": "utf-8",
-					"delimiter": "detect",
-					"skipLines": 1,
-					"commentPrefixes": ["#", "%"],
-					"sourceColumn": 0,
-					"targetColumn": 2
+					"delimiter": "tab",
+					"order": "head-relation-tail",
+					"header": "names",
+					"quoted": false
 				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/relations.json" },
 			"expectedCounts": {
 				"edges": 2250197,
 				"nodes": 45158
 			},
 			"properties": {
-				"components": 6,
-				"connected": false,
+				"components": 1,
+				"connected": true,
 				"cyclic": true,
 				"density": .0003807640611381848,
 				"directed": true,
-				"edgeTypes": 1,
+				"edgeTypes": 24,
 				"heterogeneous": true,
 				"multigraph": true,
-				"nodeTypes": 2,
+				"nodeTypes": 1,
 				"rawBytes": 0,
 				"selfLoops": false,
 				"weighted": false
@@ -8675,7 +9301,7 @@ var catalogue_default = {
 		{
 			"id": "imdb",
 			"name": "IMDB",
-			"description": "Heterogeneous film network from the Internet Movie Database. Contains movies, directors, and actors with directed-by and acted-in relations. Used as a standard benchmark for heterogeneous graph neural networks.",
+			"description": "Heterogeneous film network from the Internet Movie Database, as the Heterogeneous Graph Benchmark packages it. Contains movies, directors, actors and plot keywords, with director, actor and keyword relations each listed in both directions. Used as a standard benchmark for heterogeneous graph neural networks.",
 			"domain": "Film heterogeneous",
 			"tags": ["heterogeneous"],
 			"category": "graph/v1",
@@ -8687,10 +9313,62 @@ var catalogue_default = {
 			"licence": { "kind": "unrecorded" },
 			"citations": [{ "text": "Xinyu Lv, Tong Zhao, Yozen Liu, Neil Shah, Yong Li (2021). Are we really making much progress? Revisiting, benchmarking and refining heterogeneous graph neural networks. Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (hgb: format hgb has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "zip",
+					"parts": [{
+						"name": "nodes",
+						"member": { "path": "IMDB/node.dat" }
+					}, {
+						"name": "links",
+						"member": { "path": "IMDB/link.dat" }
+					}]
+				},
+				"parser": {
+					"format": "hgb",
+					"encoding": "utf-8",
+					"nodeTypes": [
+						"movie",
+						"director",
+						"actor",
+						"keyword"
+					],
+					"linkTypes": [
+						{
+							"name": "movie->director",
+							"source": "movie",
+							"target": "director"
+						},
+						{
+							"name": "director->movie",
+							"source": "director",
+							"target": "movie"
+						},
+						{
+							"name": "movie->actorh",
+							"source": "movie",
+							"target": "actor"
+						},
+						{
+							"name": "actor->movie",
+							"source": "actor",
+							"target": "movie"
+						},
+						{
+							"name": "movie->keyword",
+							"source": "movie",
+							"target": "keyword"
+						},
+						{
+							"name": "keyword->movie",
+							"source": "keyword",
+							"target": "movie"
+						}
+					]
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/typed.json" },
 			"expectedCounts": {
 				"edges": 86642,
 				"nodes": 21420
@@ -8701,10 +9379,10 @@ var catalogue_default = {
 				"cyclic": true,
 				"density": .00018884687883899688,
 				"directed": false,
-				"edgeTypes": 1,
+				"edgeTypes": 6,
 				"heterogeneous": true,
 				"multigraph": true,
-				"nodeTypes": 2,
+				"nodeTypes": 4,
 				"rawBytes": 0,
 				"selfLoops": false,
 				"weighted": false
@@ -9102,16 +9780,15 @@ var catalogue_default = {
 				"source": 0,
 				"archive": { "kind": "none" },
 				"parser": {
-					"format": "edge-list",
+					"format": "kg-triples",
 					"encoding": "utf-8",
-					"delimiter": "detect",
-					"skipLines": 0,
-					"commentPrefixes": ["#", "%"],
-					"sourceColumn": 0,
-					"targetColumn": 2
+					"delimiter": "tab",
+					"order": "head-relation-tail",
+					"header": "none",
+					"quoted": false
 				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/relations.json" },
 			"expectedCounts": {
 				"edges": 32888,
 				"nodes": 2034
@@ -9142,10 +9819,35 @@ var catalogue_default = {
 			},
 			"citations": [{ "text": "C. Kemp, J. B. Tenenbaum, T. L. Griffiths, T. Yamada, N. Ueda (2006). Learning systems of concepts with an infinite relational model. Proceedings of the 21st National Conference on Artificial Intelligence, 381-388." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (tsv: format tsv is a triple file whose head and tail columns the catalogue does not record)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "tar.gz",
+					"parts": [
+						{
+							"name": "train",
+							"member": { "path": "train.txt" }
+						},
+						{
+							"name": "valid",
+							"member": { "path": "valid.txt" }
+						},
+						{
+							"name": "test",
+							"member": { "path": "test.txt" }
+						}
+					]
+				},
+				"parser": {
+					"format": "kg-triples",
+					"encoding": "utf-8",
+					"delimiter": "tab",
+					"order": "head-relation-tail",
+					"header": "none",
+					"quoted": false
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/relations-split.json" },
 			"expectedCounts": {
 				"edges": 10686,
 				"nodes": 104
@@ -9176,10 +9878,35 @@ var catalogue_default = {
 			},
 			"citations": [{ "text": "L. A. Zetterström (2021). Knowledge Graph Embeddings." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (tsv: format tsv is a triple file whose head and tail columns the catalogue does not record)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "tar.gz",
+					"parts": [
+						{
+							"name": "train",
+							"member": { "path": "train.txt" }
+						},
+						{
+							"name": "valid",
+							"member": { "path": "valid.txt" }
+						},
+						{
+							"name": "test",
+							"member": { "path": "test.txt" }
+						}
+					]
+				},
+				"parser": {
+					"format": "kg-triples",
+					"encoding": "utf-8",
+					"delimiter": "tab",
+					"order": "head-relation-tail",
+					"header": "none",
+					"quoted": false
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/relations-split.json" },
 			"expectedCounts": {
 				"edges": 1992,
 				"nodes": 14
@@ -9210,10 +9937,35 @@ var catalogue_default = {
 			},
 			"citations": [{ "text": "S. Kok, P. Domingos (2007). Statistical Predicate Invention. Proceedings of the 24th International Conference on Machine Learning, 433-440." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (tsv: format tsv is a triple file whose head and tail columns the catalogue does not record)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "tar.gz",
+					"parts": [
+						{
+							"name": "train",
+							"member": { "path": "train.txt" }
+						},
+						{
+							"name": "valid",
+							"member": { "path": "valid.txt" }
+						},
+						{
+							"name": "test",
+							"member": { "path": "test.txt" }
+						}
+					]
+				},
+				"parser": {
+					"format": "kg-triples",
+					"encoding": "utf-8",
+					"delimiter": "tab",
+					"order": "head-relation-tail",
+					"header": "none",
+					"quoted": false
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/relations-split.json" },
 			"expectedCounts": {
 				"edges": 6529,
 				"nodes": 135
@@ -9593,7 +10345,7 @@ var catalogue_default = {
 		{
 			"id": "lastfm-hgb",
 			"name": "LastFM (HGB)",
-			"description": "Heterogeneous music recommendation network from Last.fm. Contains users, artists, and tags with directed user-to-artist listening relationships, social user-user connections, and artist-tag associations. Used for link prediction and recommendation tasks in heterogeneous graph neural network benchmarks.",
+			"description": "Bipartite user-artist listening network from Last.fm, from the HetRec 2011 release. Users are linked to the artists they listened to, and each edge's weight is how many times the user listened to the artist.",
 			"domain": "Music recommendation network",
 			"tags": ["heterogeneous"],
 			"category": "graph/v1",
@@ -9609,18 +10361,29 @@ var catalogue_default = {
 			},
 			"citations": [{ "text": "Xinyu Lv, Tong Zhao, Yozen Liu, Neil Shah, Yong Li (2021). Are We Really Making Much Progress? Revisiting, Benchmarking and Refining Heterogeneous Graph Neural Networks. Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (two-mode-table: format two-mode-table has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "zip",
+					"member": { "path": "user_artists.dat" }
+				},
+				"parser": {
+					"format": "user-item-counts",
+					"encoding": "utf-8",
+					"delimiter": "tab",
+					"skipLines": 1,
+					"itemType": "artist"
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/typed-weighted.json" },
 			"expectedCounts": {
 				"edges": 92834,
 				"nodes": 19524
 			},
 			"properties": {
-				"components": 2,
+				"components": 8,
 				"connected": false,
-				"cyclic": true,
+				"cyclic": false,
 				"density": .0005933961402754534,
 				"directed": true,
 				"edgeTypes": 1,
@@ -9629,7 +10392,7 @@ var catalogue_default = {
 				"nodeTypes": 2,
 				"rawBytes": 0,
 				"selfLoops": false,
-				"weighted": false
+				"weighted": true
 			}
 		},
 		{
@@ -12222,16 +12985,15 @@ var catalogue_default = {
 				"source": 0,
 				"archive": { "kind": "none" },
 				"parser": {
-					"format": "edge-list",
+					"format": "kg-triples",
 					"encoding": "utf-8",
-					"delimiter": "detect",
-					"skipLines": 1,
-					"commentPrefixes": ["#", "%"],
-					"sourceColumn": 0,
-					"targetColumn": 1
+					"delimiter": "space",
+					"order": "head-tail-relation",
+					"header": "count",
+					"quoted": false
 				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/relations.json" },
 			"expectedCounts": {
 				"edges": 149678,
 				"nodes": 74432
@@ -12242,10 +13004,10 @@ var catalogue_default = {
 				"cyclic": true,
 				"density": 46516201857617345e-21,
 				"directed": true,
-				"edgeTypes": 1,
+				"edgeTypes": 200,
 				"heterogeneous": true,
 				"multigraph": true,
-				"nodeTypes": 2,
+				"nodeTypes": 1,
 				"rawBytes": 0,
 				"selfLoops": true,
 				"weighted": false
@@ -17844,7 +18606,7 @@ var catalogue_default = {
 			"citations": [{ "text": "Weihua Hu, Matthias Fey, Marinka Zitnik, Yuxiao Dong, Hongyu Ren, Bowen Liu, Michele Catasta, Jure Leskovec (2020). Open Graph Benchmark: Datasets for Machine Learning on Graphs. Advances in Neural Information Processing Systems (NeurIPS)." }],
 			"recipe": {
 				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (kge-triples-zip: format kge-triples-zip has a pipeline-specific parser with no declarative keepset form yet)"
+				"reason": "no download is in a form keepset reads yet (kge-triples-zip: the triples are more edges than one fetch holds: a whole fetch at Node's default heap aborts with an allocation failure, because the crosswalk step copies every edge (ExaDev/crossmap#57))"
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -17857,12 +18619,12 @@ var catalogue_default = {
 				"cyclic": true,
 				"density": 2692292964066484e-21,
 				"directed": true,
-				"edgeTypes": 1,
+				"edgeTypes": 535,
 				"heterogeneous": true,
 				"multigraph": true,
-				"nodeTypes": 2,
+				"nodeTypes": 1,
 				"rawBytes": 0,
-				"selfLoops": false,
+				"selfLoops": true,
 				"weighted": false
 			}
 		},
@@ -62516,12 +63278,12 @@ var catalogue_default = {
 			"citations": [{ "text": "Xiaozhi Wang, Tianyu Gao, Zhaochen Song, Zenan Li, Bing Qin, Tie-Yan Liu, Zhiyuan Liu, Jie Tang (2021). KEPLER: A Unified Model for Knowledge Embedding and Pre-trained Language Representation. Transactions of the Association for Computational Linguistics." }],
 			"recipe": {
 				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (kge-triples-tar: format kge-triples-tar has a pipeline-specific parser with no declarative keepset form yet)"
+				"reason": "no download is in a form keepset reads yet (kge-triples-tar: the triples are more edges than one fetch holds: a whole fetch at Node's default heap aborts with an allocation failure, because the crosswalk step copies every edge (ExaDev/crossmap#57))"
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
-				"edges": 21e6,
-				"nodes": 5e6
+				"edges": 20624575,
+				"nodes": 4594485
 			},
 			"properties": {
 				"components": 1491,
@@ -62529,7 +63291,7 @@ var catalogue_default = {
 				"cyclic": true,
 				"density": 9.770379358670848e-7,
 				"directed": true,
-				"edgeTypes": 1,
+				"edgeTypes": 822,
 				"heterogeneous": true,
 				"multigraph": false,
 				"nodeTypes": 1,
@@ -62556,24 +63318,49 @@ var catalogue_default = {
 			"licence": { "kind": "unrecorded" },
 			"citations": [{ "text": "Tim Dettmers, Pasquale Minervini, Pontus Stenetorp, Sebastian Riedel (2018). Convolutional 2D Knowledge Graph Embeddings. Proceedings of the AAAI Conference on Artificial Intelligence." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (kge-triples-tar: format kge-triples-tar has a pipeline-specific parser with no declarative keepset form yet; landing-page: a landing page, not a download)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "tar.gz",
+					"parts": [
+						{
+							"name": "train",
+							"member": { "path": "wnrr/train.txt" }
+						},
+						{
+							"name": "valid",
+							"member": { "path": "wnrr/valid.txt" }
+						},
+						{
+							"name": "test",
+							"member": { "path": "wnrr/test.txt" }
+						}
+					]
+				},
+				"parser": {
+					"format": "kg-triples",
+					"encoding": "utf-8",
+					"delimiter": "tab",
+					"order": "head-relation-tail",
+					"header": "none",
+					"quoted": false
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/relations-split.json" },
 			"expectedCounts": {
-				"edges": 92879,
+				"edges": 93003,
 				"nodes": 40943
 			},
 			"properties": {
-				"components": 2,
+				"components": 13,
 				"connected": false,
 				"cyclic": true,
 				"density": .0006858835096536629,
 				"directed": true,
-				"edgeTypes": 1,
+				"edgeTypes": 11,
 				"heterogeneous": true,
-				"multigraph": false,
-				"nodeTypes": 2,
+				"multigraph": true,
+				"nodeTypes": 1,
 				"rawBytes": 0,
 				"selfLoops": true,
 				"weighted": false
@@ -63043,24 +63830,49 @@ var catalogue_default = {
 			"licence": { "kind": "unrecorded" },
 			"citations": [{ "text": "Suchanek et al. (2007). YAGO: A Core of Semantic Knowledge Unifying DBpedia and Wikipedia. ACM International Conference on World Wide Web." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (kge-triples-tar: format kge-triples-tar has a pipeline-specific parser with no declarative keepset form yet; landing-page: the download is a directory listing, not a data file)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "tar.gz",
+					"parts": [
+						{
+							"name": "train",
+							"member": { "path": "yago3-10/train.txt" }
+						},
+						{
+							"name": "valid",
+							"member": { "path": "yago3-10/valid.txt" }
+						},
+						{
+							"name": "test",
+							"member": { "path": "yago3-10/test.txt" }
+						}
+					]
+				},
+				"parser": {
+					"format": "kg-triples",
+					"encoding": "utf-8",
+					"delimiter": "tab",
+					"order": "head-relation-tail",
+					"header": "none",
+					"quoted": false
+				}
 			},
-			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
+			"crosswalk": { "file": "crosswalks/graph-v1/relations-split.json" },
 			"expectedCounts": {
-				"edges": 797297,
+				"edges": 1089040,
 				"nodes": 123182
 			},
 			"properties": {
-				"components": 21,
+				"components": 11,
 				"connected": false,
 				"cyclic": true,
 				"density": .0004193789497930294,
 				"directed": true,
-				"edgeTypes": 1,
-				"heterogeneous": false,
-				"multigraph": false,
-				"nodeTypes": 2,
+				"edgeTypes": 37,
+				"heterogeneous": true,
+				"multigraph": true,
+				"nodeTypes": 1,
 				"rawBytes": 0,
 				"selfLoops": false,
 				"weighted": false