keepset 0.38.0 → 0.39.0

package/README.md

@@ -63,7 +63,7 @@ The catalogue itself is a manifest (`catalogue/manifest.json`: id, version, repo
 
 ## Categories and normalisation targets
 
-A category names a normalisation target: the shape every entry in that category resolves to, whatever its native shape. `graph/v1` is `{ meta, nodes, edges }` with unique node ids and edges whose endpoints name nodes; its entries are graph datasets taken from upstream collections such as SuiteSparse, Netzschleuder, SNAP, KONECT and OGB, read by one parser per format (delimited edge lists, which can also be a sheet of a spreadsheet, Matrix Market, GML, Pajek, UCINet DL full matrices, Graph Drawing contest files, Witvliet connectome synapse files, KONECT edge files, rating and count files, knowledge graph triples and Heterogeneous Graph Benchmark node and link files), each giving the graph's topology and, where the format states them, node and edge types, weights, times and the split each edge belongs to. `tasks/v1` is `{ meta, nodes }`, a set of tasks as the places or cases it offers to visit or serve (a routing benchmark, or an open data register of permits and complaints), each flat with optional coordinates, prize, prize probability, duration, travel time, skill, territory, lifecycle status and times, and deadline, read from tabular sources by the `csv-table` parser (a delimited file, a member of an archive, a sheet of an Excel workbook, a legacy binary xls workbook or an OpenDocument spreadsheet, or an open data portal read a page at a time) and from ArcGIS feature layers by the `geojson` parser. Its entries come from a survey of open operations-research datasets, imported with a curation that records, for each source, what its download gives and how its columns map onto the node; a dataset whose source has no usable table is imported without a recipe and says why. Any category can be fetched as a collection: a recipe that selects every matching archive member with `members` (by path, by the extensions its name ends with, or by a regular expression its whole path matches) resolves to `{ meta, instances }`, where each instance is a document of the entry's category under the path of the member it came from, the schema derived from the category's own, so a category needs nothing of its own to be collected. The parser reads each member, the crosswalk is applied to each, and a member that does not parse or does not validate fails the fetch and is named. Instance files are read by one parser per format (Solomon and TOPTW time window files, Chao team orienteering files, TSPLIB style orienteering files, VRP-REP XML, VeRoLog 2019 files and customer point lists), which gives each node the columns the file has and each instance a header of its own; a curation says which of them fill the `tasks/v1` node's fields, which pass through under their own names and which are dropped, so what an entry leaves out is written down, and a source whose time quantities are not seconds keeps them under the source's names with the unit recorded as the entry's `timeUnit`. A new category starts by defining its target schema in `src/targets/` and registering it in the category registry; an entry cannot name a category that does not exist. The statically generated site renders every entry with schema.org Dataset JSON-LD, so the catalogue surfaces in the dataset discovery layers.
+A category names a normalisation target: the shape every entry in that category resolves to, whatever its native shape. `graph/v1` is `{ meta, nodes, edges }` with unique node ids and edges whose endpoints name nodes; its entries are graph datasets taken from upstream collections such as SuiteSparse, Netzschleuder, SNAP, KONECT and OGB, read by one parser per format (delimited edge lists, which can also be a sheet of a spreadsheet, Matrix Market, GML, Pajek, UCINet DL full matrices, Graph Drawing contest files, Witvliet connectome synapse files, Open Graph Benchmark raw directories read together from the several files of a dataset's zip, KONECT edge files, rating and count files, knowledge graph triples and Heterogeneous Graph Benchmark node and link files), each giving the graph's topology and, where the format states them, node and edge types, weights, times and the split each edge belongs to. `tasks/v1` is `{ meta, nodes }`, a set of tasks as the places or cases it offers to visit or serve (a routing benchmark, or an open data register of permits and complaints), each flat with optional coordinates, prize, prize probability, duration, travel time, skill, territory, lifecycle status and times, and deadline, read from tabular sources by the `csv-table` parser (a delimited file, a member of an archive, a sheet of an Excel workbook, a legacy binary xls workbook or an OpenDocument spreadsheet, or an open data portal read a page at a time) and from ArcGIS feature layers by the `geojson` parser. Its entries come from a survey of open operations-research datasets, imported with a curation that records, for each source, what its download gives and how its columns map onto the node; a dataset whose source has no usable table is imported without a recipe and says why. Any category can be fetched as a collection: a recipe that selects every matching archive member with `members` (by path, by the extensions its name ends with, or by a regular expression its whole path matches) resolves to `{ meta, instances }`, where each instance is a document of the entry's category under the path of the member it came from, the schema derived from the category's own, so a category needs nothing of its own to be collected. The parser reads each member, the crosswalk is applied to each, and a member that does not parse or does not validate fails the fetch and is named. Instance files are read by one parser per format (Solomon and TOPTW time window files, Chao team orienteering files, TSPLIB style orienteering files, VRP-REP XML, VeRoLog 2019 files and customer point lists), which gives each node the columns the file has and each instance a header of its own; a curation says which of them fill the `tasks/v1` node's fields, which pass through under their own names and which are dropped, so what an entry leaves out is written down, and a source whose time quantities are not seconds keeps them under the source's names with the unit recorded as the entry's `timeUnit`. A new category starts by defining its target schema in `src/targets/` and registering it in the category registry; an entry cannot name a category that does not exist. The statically generated site renders every entry with schema.org Dataset JSON-LD, so the catalogue surfaces in the dataset discovery layers.
 
 ## Licensing
 

package/dist/{bundled-U2l4WT5T.js → bundled-6dYCPVD-.js}

@@ -523,6 +523,12 @@ const CollectionArchiveSchema = z.object({
 		sheets: MemberPatternSchema.optional()
 	}).strict().optional()
 }).strict();
+/** An archive whose dataset is spread over several of its files, which a parser of a multi-file format (`ogb`) reads together into one document: every file the `files` selector matches, at least one, in path order, each gunzipped first when `inner` says so (an OGB dataset zip holds its tables as `.csv.gz` files). */
+const FilesArchiveSchema = z.object({
+	kind: FileArchiveKindSchema,
+	files: ArchiveMemberSchema,
+	inner: z.object({ kind: z.literal("gzip") }).strict().optional()
+}).strict();
 /** A spreadsheet read as the CSV of one of its sheets (the sheet's name, as the workbook gives it): an Excel workbook (`xlsx`), a legacy binary Excel 97 to 2003 workbook (`xls`, BIFF8) or an OpenDocument spreadsheet (`ods`, whose sheets are its tables). The three give the same text for the same cells. */
 const SheetArchiveSchema = z.object({
 	kind: z.enum([
@@ -551,7 +557,7 @@ const PartsArchiveSchema = z.object({
 		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, a spreadsheet inside a zip) is unpacked by `inner`. A spreadsheet (`xlsx`, `xls` or `ods`) is read as the CSV of its named sheet, with dates as ISO 8601 text, for the `csv-table` or `edge-list` 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. */
+/** How the downloaded bytes unpack into the file the parser reads, the named files a parser of a parts format reads together into one document (`parts`, such as the split files of a triples download or an HGB dataset's nodes and links), the files a parser of a multi-file format reads together into one document (`files`, such as an OGB raw directory), 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, a spreadsheet inside a zip) is unpacked by `inner`. A spreadsheet (`xlsx`, `xls` or `ods`) is read as the CSV of its named sheet, with dates as ISO 8601 text, for the `csv-table` or `edge-list` 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(),
@@ -563,6 +569,7 @@ const ArchiveSchema = z.union([
 	}).strict(),
 	CollectionArchiveSchema,
 	PartsArchiveSchema,
+	FilesArchiveSchema,
 	SheetArchiveSchema,
 	z.object({
 		kind: z.literal("shapefile"),
@@ -598,12 +605,21 @@ const ParserSchema = z.discriminatedUnion("format", [
 		/** The column holding how many edges a line stands for, a whole number, for a table that lists each pair once with its count; a line that stands for none adds neither an edge nor a node. */
 		multiplicityColumn: ColumnSchema.optional(),
 		/** The table lists every undirected edge twice, once in each direction and with the same multiplicity: each pair's edges are emitted once, in the direction it is first listed, a self-loop's once, and a line whose mirror is missing is an error. graph/v1 records direction in `meta` and never by duplicating edges, so an undirected graph read from such a table needs this. */
-		mirrored: z.literal(true).optional()
+		mirrored: z.literal(true).optional(),
+		/** The conditions a data line must meet to be an edge, each a column whose trimmed field equals a value; a file whose rows are of several kinds (an interaction file listing genetic beside physical interactions, or interactions with another organism's proteins) is read for the rows the dataset is made of. Without it every data line is an edge. */
+		where: z.array(z.object({
+			column: ColumnSchema,
+			equals: z.string().min(1)
+		}).strict()).min(1).optional()
 	}).strict().refine((parser) => parser.sourceColumn !== parser.targetColumn, {
 		message: "sourceColumn and targetColumn must differ",
 		path: ["targetColumn"]
-	}).refine((parser) => [parser.keep?.column, parser.multiplicityColumn].every((column) => column === void 0 || column !== parser.sourceColumn && column !== parser.targetColumn), {
-		message: "the keep and multiplicity columns must not be an endpoint column",
+	}).refine((parser) => [
+		parser.keep?.column,
+		parser.multiplicityColumn,
+		...(parser.where ?? []).map((condition) => condition.column)
+	].every((column) => column === void 0 || column !== parser.sourceColumn && column !== parser.targetColumn), {
+		message: "the keep, multiplicity and where columns must not be an endpoint column",
 		path: ["keep"]
 	}),
 	z.object({
@@ -700,6 +716,10 @@ const ParserSchema = z.discriminatedUnion("format", [
 		format: z.literal("witvliet-synapses"),
 		encoding: z.literal("utf-8")
 	}).strict(),
+	z.object({
+		format: z.literal("ogb"),
+		encoding: z.literal("utf-8")
+	}).strict(),
 	z.object({
 		format: z.literal("customer-points"),
 		encoding: EncodingSchema
@@ -799,6 +819,8 @@ const PagingSchema = z.discriminatedUnion("kind", [
 		pageSize: z.int().min(1)
 	}).strict()
 ]);
+/** The parser formats that read a dataset spread over several files of one archive, and so the only ones a `files` archive is given to. */
+const MULTI_FILE_FORMATS = /* @__PURE__ */ new Set(["ogb"]);
 /** 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"]]]);
@@ -823,6 +845,11 @@ const RecipeSchema = z.discriminatedUnion("status", [z.object({
 			message: `the ${recipe.parser.format} parser reads the parts ${required.join(" and ")} from a parts archive`
 		});
 	}
+	if ("files" in recipe.archive !== MULTI_FILE_FORMATS.has(recipe.parser.format)) ctx.addIssue({
+		code: "custom",
+		path: ["archive"],
+		message: `a files archive is read by a multi-file parser (${[...MULTI_FILE_FORMATS].join(", ")}), and such a parser reads only a files archive`
+	});
 }), z.object({
 	status: z.literal("not-fetchable"),
 	reason: z.string().min(1)
@@ -1206,7 +1233,7 @@ 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": "1.29.0",
+		"version": "1.30.0",
 		"repository": "https://github.com/Mearman/keepset",
 		"namespaceAliases": {
 			"networkrepository": ["nr"],
@@ -2463,7 +2490,7 @@ var catalogue_default = {
 			"citations": [{ "text": "Jie Tang, Jing Zhang, Limin Yao, Juanzi Li, Li Zhang, Zhong Su (2008). ArnetMiner: Extraction and Mining of Academic Social Networks. Proceedings of the 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining." }],
 			"recipe": {
 				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (aminer-7z: 7z compression has no keepset unpacker yet)"
+				"reason": "no download is in a form keepset reads yet (aminer-7z: the archive is a 7z (LZMA2) file holding one JSON array of paper records (dblp.v12.json) that unpacks to over twelve gigabytes, more than a fetch can hold in memory, so no 7z reader is added for it)"
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -2556,8 +2583,21 @@ var catalogue_default = {
 			"licence": { "kind": "unrecorded" },
 			"citations": [{ "text": "R. Guimera, L. Danon, A. Diaz-Guilera, F. Giralt, A. Arenas (2003). Self-similar community structure in a network of human interactions. Physical Review E 68, 065103(R)." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (various: an archive in format various, whose member and format the catalogue does not record)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "zip",
+					"member": { "path": "email.txt" }
+				},
+				"parser": {
+					"format": "edge-list",
+					"encoding": "utf-8",
+					"delimiter": "whitespace",
+					"skipLines": 0,
+					"commentPrefixes": ["#", "%"],
+					"sourceColumn": 0,
+					"targetColumn": 1
+				}
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -9407,7 +9447,7 @@ var catalogue_default = {
 			"citations": [{ "text": "Junheng Lin, Yibin Shen, Yifan Wang, Fangzhao Wu, Xiang Wang (2024). H2GB: A Hierarchical Heterogeneous Graph Benchmark for Node Property Prediction. Proceedings of the AAAI Conference on Artificial Intelligence." }],
 			"recipe": {
 				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (ieee-cis-zip: format ieee-cis-zip has a pipeline-specific parser with no declarative keepset form yet)"
+				"reason": "no download is in a form keepset reads yet (ieee-cis-zip: the archive holds the IEEE-CIS fraud detection competition's raw transaction and identity tables, not a graph: H2GB builds its graph from shared card, address, email and device fields with its own construction code)"
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -18616,7 +18656,7 @@ var catalogue_default = {
 			"citations": [{ "text": "Junheng Lin, Yibin Shen, Yifan Wang, Fangzhao Wu, Xiang Wang (2024). H2GB: A Hierarchical Heterogeneous Graph Benchmark for Node Property Prediction. Proceedings of the AAAI Conference on Artificial Intelligence." }],
 			"recipe": {
 				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (pytorch-hetero: format pytorch-hetero has a pipeline-specific parser with no declarative keepset form yet)"
+				"reason": "no download is in a form keepset reads yet (pytorch-hetero: the file is a PyTorch torch.save archive whose graph is a pickled Python dictionary (pyHGT's OAG layout of node_forward, node_feature and edge_list by target type, source type and relation), which only a Python runtime with PyTorch can unpickle, and it is several gigabytes)"
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -18648,6 +18688,10 @@ var catalogue_default = {
 				"url": "https://data.pyg.org/datasets/ogbl_biokg.zip",
 				"retrieved": "2026-03-26",
 				"status": "live"
+			}, {
+				"url": "https://snap.stanford.edu/ogb/data/linkproppred/biokg.zip",
+				"retrieved": "2026-10-21",
+				"status": "live"
 			}],
 			"licence": {
 				"kind": "spdx",
@@ -18655,8 +18699,17 @@ var catalogue_default = {
 			},
 			"citations": [{ "text": "Hu et al. (2020). Open Graph Benchmark: Datasets for Machine Learning on Graphs. NeurIPS." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (ogb-zip: format ogb-zip has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 1,
+				"archive": {
+					"kind": "zip",
+					"files": { "pattern": "[^/]+/raw/(?:num-node-list|num-node-dict|edge|num-edge-list|triplet-type-list|relations/[^/]+/(?:edge|num-edge-list))\\.csv\\.gz" },
+					"inner": { "kind": "gzip" }
+				},
+				"parser": {
+					"format": "ogb",
+					"encoding": "utf-8"
+				}
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -18696,8 +18749,17 @@ 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. NeurIPS." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (ogb: format ogb has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "zip",
+					"files": { "pattern": "[^/]+/raw/(?:num-node-list|num-node-dict|edge|num-edge-list|triplet-type-list|relations/[^/]+/(?:edge|num-edge-list))\\.csv\\.gz" },
+					"inner": { "kind": "gzip" }
+				},
+				"parser": {
+					"format": "ogb",
+					"encoding": "utf-8"
+				}
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -18769,6 +18831,10 @@ var catalogue_default = {
 				"url": "https://data.pyg.org/datasets/ogbn_mag.zip",
 				"retrieved": "2026-03-26",
 				"status": "live"
+			}, {
+				"url": "https://snap.stanford.edu/ogb/data/nodeproppred/mag.zip",
+				"retrieved": "2026-10-21",
+				"status": "live"
 			}],
 			"licence": {
 				"kind": "spdx",
@@ -18776,8 +18842,17 @@ var catalogue_default = {
 			},
 			"citations": [{ "text": "Hu et al. (2020). Open Graph Benchmark: Datasets for Machine Learning on Graphs. NeurIPS." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (ogb-zip: format ogb-zip has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 1,
+				"archive": {
+					"kind": "zip",
+					"files": { "pattern": "[^/]+/raw/(?:num-node-list|num-node-dict|edge|num-edge-list|triplet-type-list|relations/[^/]+/(?:edge|num-edge-list))\\.csv\\.gz" },
+					"inner": { "kind": "gzip" }
+				},
+				"parser": {
+					"format": "ogb",
+					"encoding": "utf-8"
+				}
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -18821,7 +18896,7 @@ var catalogue_default = {
 			}],
 			"recipe": {
 				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (parquet: format parquet has a pipeline-specific parser with no declarative keepset form yet)"
+				"reason": "no download is in a form keepset reads yet (parquet: the download is a Hugging Face dataset repository whose citation table is thousands of Parquet files holding billions of rows, tens of gigabytes in all, more than a fetch can hold in memory, so no Parquet reader is added for it)"
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -23458,8 +23533,35 @@ var catalogue_default = {
 			},
 			"citations": [{ "text": "Szklarczyk et al. (2019). STRING: functional protein association networks. Nature Methods." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (biogrid-zip: format biogrid-zip has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": {
+					"kind": "zip",
+					"member": { "pattern": "BIOGRID-ORGANISM-Homo_sapiens-[0-9.]+\\.tab3\\.txt" }
+				},
+				"parser": {
+					"format": "edge-list",
+					"encoding": "utf-8",
+					"delimiter": "tab",
+					"skipLines": 0,
+					"commentPrefixes": ["#"],
+					"sourceColumn": 3,
+					"targetColumn": 4,
+					"where": [
+						{
+							"column": 12,
+							"equals": "physical"
+						},
+						{
+							"column": 15,
+							"equals": "9606"
+						},
+						{
+							"column": 16,
+							"equals": "9606"
+						}
+					]
+				}
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -23887,8 +23989,18 @@ var catalogue_default = {
 			"licence": { "kind": "unrecorded" },
 			"citations": [{ "text": "Leskovec, Jure, Huttenlocher, Daniel, Kleinberg, Jon (2010). Signed Networks in Social Media. Proceedings of the SIGCHI Conference on Human Factors in Computing Systems." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (snap-signed: format snap-signed has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": { "kind": "gzip" },
+				"parser": {
+					"format": "edge-list",
+					"encoding": "utf-8",
+					"delimiter": "tab",
+					"skipLines": 0,
+					"commentPrefixes": ["#"],
+					"sourceColumn": 0,
+					"targetColumn": 1
+				}
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -60469,8 +60581,18 @@ var catalogue_default = {
 			},
 			"citations": [{ "text": "Szklarczyk, Damian, Kirsch, Rebecca, Koutrouli, Mikaela, Nastou, Katerina, Mehryary, Farrokh, Hachilif, Radja, Annika, Gable, Fang, Tao, Doncheva, Nadezhda T., Letunic, Ivica, Bork, Peer, Jensen, Lars J., von Mering, Christian (2023). The STRING database in 2023: protein–protein association networks and functional enrichment analyses for any of 12 535 organisms. Nucleic Acids Research." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (string-ppi: format string-ppi has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": { "kind": "gzip" },
+				"parser": {
+					"format": "edge-list",
+					"encoding": "utf-8",
+					"delimiter": "space",
+					"skipLines": 1,
+					"commentPrefixes": [],
+					"sourceColumn": 0,
+					"targetColumn": 1
+				}
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {
@@ -60957,8 +61079,18 @@ var catalogue_default = {
 			},
 			"citations": [{ "text": "Barrat et al. (2004). The architecture of complex weighted networks. Proceedings of the National Academy of Sciences." }],
 			"recipe": {
-				"status": "not-fetchable",
-				"reason": "no download is in a form keepset reads yet (csv: format csv has a pipeline-specific parser with no declarative keepset form yet)"
+				"status": "fetchable",
+				"source": 0,
+				"archive": { "kind": "none" },
+				"parser": {
+					"format": "edge-list",
+					"encoding": "utf-8",
+					"delimiter": "detect",
+					"skipLines": 0,
+					"commentPrefixes": ["#", "%"],
+					"sourceColumn": 2,
+					"targetColumn": 4
+				}
 			},
 			"crosswalk": { "file": "crosswalks/graph-v1/topology.json" },
 			"expectedCounts": {