Gramscii-IT/open-data-catalogue

Open Data catalogue

A catalogue of 15,990 open datasets of official statistics and Italian public finance, from ISTAT, Eurostat, OECD, ILO, DoveVannoINostriSoldi and Cruscotto Italia. For every dataset: what it is, the dimensions it is cut by and their codes, the words those codes stand for in each language the provider publishes, the notes it is documented with, and one searchable document per dataset and language, in Italian and in English.

Reading the six providers' catalogues, structures and codelists takes a day at the pace they allow; this archive holds the result, so a reader starts from it and asks the providers only for what changes.

One file, seven tables, 15,990 datasets, 31,673 documents, 55 MB. Taken on 7 September 2026 at 22:36 UTC.

Three repositories, used together

Where	What it holds	What it is for
This dataset	the catalogue: what each dataset is, its dimensions and codes, the words for the codes, its notes and its searchable documents	knowing what the six providers publish and asking them the right question, without reading their catalogues yourself
Gramscii-Git/open-data-catalogue on GitHub	the script that produces this archive	keeping the catalogue current: run on a deployment that holds it, the script asks the providers for what moved, rebuilds the documents, exports the tables, uploads the archive here and prints the pin to set
Gramscii-Git/boundaries on GitHub	the administrative boundaries of Italy, Europe and the world as SVG paths, each shape with the ISTAT, NUTS and ISO identifiers it is known by	drawing the numbers of a dataset on a map

How they fit: a dataset in this catalogue is cut by a territorial dimension whose codes are the identifiers the shapes of boundaries carry as aliases. ISTAT's dataset 164_279_DF_DCIS_RICPOPRES1991_14 below has a REF_AREA dimension with the code ITE4 for Lazio; the shape of Lazio in italy-regions.geo.json carries ITE4 among its aliases, so a row of that dataset colours that shape with no lookup table in between. Eurostat's geo codes (IT, EL, UK) land on europe.geo.json, OECD's and ILO's ISO alpha-3 codes (AFG) on world.geo.json, and the ISTAT municipality codes of DoveVannoINostriSoldi and Cruscotto Italia on italy-municipalities.geo.json.

Table	Rows	One row is	Key	Newer by
opendata_catalog	15,990	a dataset as its provider lists it: title, description, category, structure it points at, last update, licence and attribution	provider, dataset_id	synced_at
opendata_structures	15,854	the dimensions of one dataset, each with its codes and their names, the time coverage and the key order	provider, dataset_id	harvested_at
opendata_structure_dims	0	which concept scheme and codelist each dimension of a structure refers to	provider, structure_id, dimension_id	replaced
opendata_terms	1,414,937	one code of one codelist, with its name in one language	provider, language, scope, code	replaced
opendata_labels	8	the labels a dataset publishes for its codes, per language	provider, dataset_id, language	fetched_at
opendata_meta_reports	1,169	one ISTAT metadata report: the notes a dataset is documented with	provider, report_key	fetched_at
opendata_documents	31,673	the text a search reads, one per dataset and language: title, summary, dimensions and their codes	provider, dataset_id, language	built_at

"Newer by" is the rule the reader applies row by row: a row is inserted where its key is new, replaces the held row where the archive's is newer by that column, and changes nothing otherwise. Tables without such a column are replaced as they arrive.

Providers

Provider	Host	Active datasets	Structures read	Languages	Terms
ISTAT	esploradati.istat.it (SDMX)	4,897	4,773, plus 123 with a recorded error	it, en	CC BY 4.0, with exceptions recorded per dataset
Eurostat	ec.europa.eu (SDMX)	8,154	8,150	en	Eurostat reuse policy
OECD	sdmx.oecd.org (SDMX)	1,544	1,214, plus 330 with a recorded error	en	OECD terms and conditions
ILO	sdmx.ilo.org (SDMX)	1,212	1,212	en	ILO rights and permissions
DoveVannoINostriSoldi	www.dovevannoinostrisoldi.com	50	27	it	per dataset, as the source publishes it: CC BY 4.0, CC BY 3.0, CC BY-SA 4.0, IODL 2.0, or not declared
Cruscotto Italia	cruscotto-italia-mcp.agid.workers.dev	25	25	it	not recorded in this release

DoveVannoINostriSoldi aggregates Italian public-finance sources (OpenBDAP of the Ragioneria Generale dello Stato, SIOPE, ANAC, INPS, the Indice PA and others); each of its datasets names its own source, owner and licence in the catalogue row. Cruscotto Italia serves AgID's municipal indicators (demography, air quality, broadband coverage, public contracts and more).

Every catalogue row carries licence and attribution as the provider publishes them for that dataset, where it publishes them: 3,284 of ISTAT's 4,900 rows, 8,154 of Eurostat's 8,256, all 1,212 of ILO's, and 27 of DoveVannoINostriSoldi's 50 carry a licence or a note on it; the rest carry none because the provider states none for that dataset.

What each provider gives

One dataset per provider, exactly as the archive holds it, table by table: the catalogue row with every column, the structure, the document, and the rows of the other tables that belong to it. Only the lists of codes are cut, at four per dimension, with values_total saying how many there are.

The columns of a catalogue row

Column	What it holds
provider, dataset_id	the key: the provider's identifier and the dataset's own
title, title_en, names	the title as published, and per language in names
description, descriptions	the description as published, and per language; null where the provider publishes none
category_path, category_paths	the provider's own classification path, and per language
version, structure_id, agency	the dataflow's version, the data structure it points at, the agency maintaining it
last_update	what the provider publishes as the last update; null where it publishes none (OECD, DoveVannoINostriSoldi, Cruscotto Italia)
period_start, period_end	the coverage the provider declares in its listing, where it does
territorial_level	the level a dataset is served at, where the provider fixes one (municipality for Cruscotto Italia)
active, retired	still listed by the provider; no longer listed, and kept without being served
searchable	active, not retired, not a dead title (ISTAT's dataflow eliminato), and not a saved view of a searchable parent
served	whether the reader answers questions from this dataset's data; a listed dataset the reader does not serve keeps its row
freshness, caveat	how current the data is where the source says (live: read at question time), and a note the source attaches
filters	the parameters the dataset accepts, for sources that are searched rather than cut by dimensions
sources	the addresses behind the row: the dataset's own (role: dataset), the terms it is published under (role: terms, with their scope), and for DoveVannoINostriSoldi the upstream source with its owner and cadence
licence, attribution	as published for the dataset
keywords	keywords per language, where published
report_key	the metadata report the dataset is documented with, joining opendata_meta_reports
in_hub, seen_hub_at, seen_registry_at	ISTAT lists datasets in two places, the data browser hub and the SDMX registry: whether the hub lists it, and when each last did
match_text, text_hash	the text a search matches (title, category and description in every language) and its digest, which decides when the document is rebuilt
synced_at	when a sync last wrote the row

ISTAT: 164_279_DF_DCIS_RICPOPRES1991_14, Lazio

The catalogue row:

{"provider": "istat", "dataset_id": "164_279_DF_DCIS_RICPOPRES1991_14",
 "title": "Lazio", "title_en": "Lazio", "names": {"en": "Lazio", "it": "Lazio"},
 "description": null, "descriptions": {},
 "category_path": "Popolazione e famiglie › Popolazione intercensuaria › Popolazione residente ricostruita - Anni 1982-1991 › Regioni e comuni",
 "category_paths": {"en": "Population and Households › Inter censuses population › Estimated resident population - Years 1982-1991 › Regions and municipalities",
                    "it": "Popolazione e famiglie › Popolazione intercensuaria › Popolazione residente ricostruita - Anni 1982-1991 › Regioni e comuni"},
 "version": "1.0", "structure_id": "DCIS_RICPOPRES1991", "agency": "IT1",
 "last_update": "2022-12-09T21:13:35.232Z", "period_start": null, "period_end": null, "territorial_level": null,
 "active": true, "retired": false, "searchable": true, "served": true,
 "freshness": null, "caveat": null, "filters": [],
 "sources": [{"url": "https://esploradati.istat.it/SDMXWS/rest/dataflow/IT1/164_279_DF_DCIS_RICPOPRES1991_14", "role": "dataset"},
             {"url": "https://www.istat.it/dati/open-data/", "role": "terms", "scope": "provider-with-exceptions"}],
 "licence": "CC BY 4.0", "attribution": "Istituto nazionale di statistica (ISTAT)", "keywords": {},
 "report_key": "DCIS_RICPOPRES1991",
 "in_hub": true, "seen_hub_at": "2026-09-08T00:39:00.290175+02:00", "seen_registry_at": "2026-08-24T23:39:33.624249+02:00",
 "match_text": "Lazio\nPopolazione e famiglie › Popolazione intercensuaria › Popolazione residente ricostruita - Anni 1982-1991 › Regioni e comuni\nLazio\nPopulation and Households › Inter censuses population › Estimated resident population - Years 1982-1991 › Regions and municipalities",
 "text_hash": "9693162fa8707a612f2b34c7e1c8200e20cc7644963fee96b6e3880cc5caebdd",
 "synced_at": "2026-08-22T20:22:21.460782+02:00"}

The structure. time is the coverage the provider declares; key_order says where the order of the key came from (hub, dsd or constraints); error would hold the provider's refusal instead of dimensions:

{"provider": "istat", "dataset_id": "164_279_DF_DCIS_RICPOPRES1991_14",
 "harvested_at": "2026-08-27T17:00:18.17008+02:00", "obs_count": null, "error": null, "key_order": "hub",
 "time": {"min": "1982-01-01", "max": "1991-12-31"},
 "dimensions": [
   {"id": "FREQ", "name": "Frequency", "values_total": 1, "values": [{"code": "A", "name": "annual"}]},
   {"id": "REF_AREA", "name": "Territory", "values_total": 383, "values": [{"code": "IT", "name": "Italy"}, {"code": "ITE", "name": "Centro (I)"}, {"code": "ITE4", "name": "Lazio"}, {"code": "ITE41", "name": "Viterbo"}]},
   {"id": "DATA_TYPE", "name": "Data type", "values_total": 1, "values": [{"code": "JAN", "name": "population on 1st January"}]},
   {"id": "AGE", "name": "Age class", "values_total": 102, "values": [{"code": "Y0", "name": "0 years"}, {"code": "Y1", "name": "1 years"}, {"code": "Y2", "name": "2 years"}, {"code": "Y3", "name": "3 years"}]},
   {"id": "SEX", "name": "Gender", "values_total": 3, "values": [{"code": "1", "name": "males"}, {"code": "2", "name": "females"}, {"code": "9", "name": "total"}]}]}

The Italian document, language: it, built_at: 2026-08-29T22:33:54.480539+02:00, 2,032 characters. Its sections are Titolo, Copertura, Sintesi, Unità di misura and Periodo coperto; the dimensions are listed with the names of their codes, a long list as a count:

Titolo
Lazio

Copertura
Popolazione e famiglie › Popolazione intercensuaria › Popolazione residente ricostruita - Anni 1982-1991 › Regioni e comuni

Sintesi
Ricostruzione intercensuaria della popolazione residente: FONTI: Periodo 1952-1971: Ricostruzione della popolazione residente per sesso, età e regione - Anni 1952-1972 - Università degli studi di Roma La Sapienza, Dipartimento di Scienze Demografiche – Fonti e strumenti, 1983; Periodo 1972-1981: “Popolazione e bilanci demografici per sesso, età e regione – ricostruzione per gli anni 1972-1981” – Istat, supplemento al bollettino di statistica n.14 – anno 1985; Periodo 1982-1991: Istat 1996, “Ricostruzione della popolazione residente per età e sesso delle province italiane – Anni 1982-1991”, Informazioni speciale n. 17; Periodo 1992-2001: Istat 2006, “Ricostruzione della popolazione residente per età e sesso nei comuni italiani – Anni 1992-2001”, Informazioni n. 13.

Unità di misura
Frequenza: annuale
Territorio di riferimento (383 values)
Indicatore: popolazione al 1º gennaio
Età: 0 anni, 1 anni, 2 anni, 3 anni, 4 anni, 5 anni, 6 anni, 7 anni, 8 anni, 9 anni, 10 anni, 11 anni, 12 anni, 13 anni, 14 anni, 15 anni, 16 anni, 17 anni, 18 anni, 19 anni, 20 anni, 21 anni, 22 anni, 23 anni, 24 anni, 25 anni, 26 anni, 27 anni, 28 anni, 29 anni, 30 anni, 31 anni, 32 anni, 33 anni, 34 anni, 35 anni, 36 anni, 37 anni, 38 anni, 39 anni, 40 anni, 41 anni, 42 anni, 43 anni, 44 anni, 45 anni, 46 anni, 47 anni, 48 anni, 49 anni, 50 anni, 51 anni, 52 anni, 53 anni, 54 anni, 55 anni, 56 anni, 57 anni, 58 anni, 59 anni, 60 anni, 61 anni, 62 anni, 63 anni, 64 anni, 65 anni, 66 anni, 67 anni, 68 anni, 69 anni, 70 anni, 71 anni, 72 anni, 73 anni, 74 anni, 75 anni, 76 anni, 77 anni, 78 anni, 79 anni, 80 anni, 81 anni, 82 anni, 83 anni, 84 anni, 85 anni, 86 anni, 87 anni, 88 anni, 89 anni, 90 anni, 91 anni, 92 anni, 93 anni, 94 anni, 95 anni, 96 anni, 97 anni, 98 anni, 99 anni, 100 anni e più, totale
Sesso: maschi, femmine, totale

Periodo coperto
1982-01-01–1991-12-31

The English document, language: en, 2,082 characters, has the same sections as Title, Coverage, Summary, Unit and Period covered, with the summary and the codes' names in English (Intercensal register population estimates …, Age: 0 years, 1 years …, Sex: males, females, total).

The metadata report the row's report_key joins, one of ISTAT's 1,169, here DCAR_ATT_NOTAR with its texts cut: a report holds one text per language the provider writes it in.

{"provider": "istat", "report_key": "DCAR_ATT_NOTAR", "metadata_set_id": "MS_ISTAT_TOPMETA2",
 "source_url": "https://esploradati.istat.it/METADATA_API/api/getMetadata?metadataSetId=MS_ISTAT_TOPMETA2&reportId=DCAR_ATT_NOTAR",
 "texts": {"en": "Notary activity: The survey on notarial activity provides the number and the characteristics of the notarial deeds and of the agreements contained within them. The survey provides a comprehensive overview of notarial activity and a set of socio-economic information that is significant for the collectivity as a whole. …", "it": "…"},
 "fetched_at": "2026-08-24T23:39:44.901304+02:00", "last_error": null, "last_attempt_at": "2026-08-24T23:39:44.901304+02:00", "permanent": false}

A labels row, one of the 8, all of ISTAT: the names a dataset itself gives its codes, per language, keyed by dimension:

{"provider": "istat", "dataset_id": "25_74", "language": "it", "fetched_at": "2026-08-28T15:48:01.028412+02:00",
 "labels": {"FREQ": {"name": "Frequenza", "values": {"A": "annuale"}},
            "DATA_TYPE": {"name": "Tipo dato", "values": {"LBIRTH": "nati vivi", "LBIRTH_MALE": "nati vivi maschi", "LBIRTH_FEMALE": "nati vive femmine", "LBIRTH_FOREIGN": "nati vivi stranieri", "…": "…"}}, "…": "…"}}

Terms of ISTAT, in Italian and in English:

{"provider": "istat", "language": "en", "scope": "", "code": "ABSTRACT", "name": "Abstract"}
{"provider": "istat", "language": "it", "scope": "", "code": "ABSTRACT", "name": "Abstract"}
{"provider": "istat", "language": "en", "scope": "", "code": "ACADEMIC_DISCIPL_AREAS", "name": "Academic disciplinary areas for university research and teaching"}

Eurostat: CEI_PC035, Food waste

The catalogue row. Eurostat publishes a description, a coverage (period_start, period_end) and an ESMS report key, in English only:

{"provider": "eurostat", "dataset_id": "CEI_PC035",
 "title": "Food waste", "title_en": "Food waste", "names": {"en": "Food waste"},
 "description": "The indicator is defined as the amount of food waste generated per year divided by the average population of the country. Food waste is measured as fresh mass all along the food value chain, including production sites, processing and manufacture, retail and distribution, restaurants and food services and households.",
 "descriptions": {"en": "The indicator is defined as the amount of food waste generated per year divided by the average population of the country. …"},
 "category_path": "Environment › Waste › Waste streams", "category_paths": {"en": "Environment › Waste › Waste streams"},
 "version": "1.0", "structure_id": "CEI_PC035", "agency": "ESTAT",
 "last_update": "2026-02-18T23:00:00+0100", "period_start": "2020", "period_end": "2023", "territorial_level": null,
 "active": true, "retired": false, "searchable": true, "served": true,
 "freshness": null, "caveat": null, "filters": [],
 "sources": [{"url": "https://ec.europa.eu/eurostat/api/dissemination/sdmx/2.1/dataflow/ESTAT/CEI_PC035", "role": "dataset"},
             {"url": "https://ec.europa.eu/eurostat/help/copyright-notice", "role": "terms", "scope": "dataset-specific"}],
 "licence": "Eurostat reuse policy", "attribution": "European Commission, Eurostat", "keywords": {},
 "report_key": "cei_pc035_esmsip2",
 "in_hub": null, "seen_hub_at": null, "seen_registry_at": "2026-09-08T00:39:33.813663+02:00",
 "match_text": "Food waste\nThe indicator is defined as the amount of food waste generated per year divided by the average population of the country. …\nEnvironment › Waste › Waste streams",
 "text_hash": "3b3cd6d2c17e3f7594ca6926e5c193d7c3f0ca22082ae5cfd388f3a549472561",
 "synced_at": "2026-08-19T20:22:33.024373+02:00"}

The structure, six dimensions, the key in the order of the data structure definition; the coverage is in the catalogue row, not here:

{"provider": "eurostat", "dataset_id": "CEI_PC035",
 "harvested_at": "2026-08-28T13:35:24.974127+02:00", "obs_count": null, "error": null, "key_order": "dsd", "time": {},
 "dimensions": [
   {"id": "freq", "name": "Time frequency", "values_total": 1, "values": [{"code": "A", "name": "Annual"}]},
   {"id": "wst_oper", "name": "Waste management operations", "values_total": 1, "values": [{"code": "COL", "name": "Waste collected"}]},
   {"id": "waste", "name": "Waste categories", "values_total": 1, "values": [{"code": "W091_092_101_FD", "name": "Food waste - bio, household and similar waste"}]},
   {"id": "nace_r2", "name": "Statistical classification of economic activities in the European Community (NACE Rev. 2)", "values_total": 1, "values": [{"code": "TOT", "name": "Total (aggregate changing according to the context)"}]},
   {"id": "unit", "name": "Unit of measure", "values_total": 1, "values": [{"code": "KG_HAB", "name": "Kilograms per capita"}]},
   {"id": "geo", "name": "Geopolitical entity (reporting)", "values_total": 41, "values": [{"code": "EU27_2020", "name": "European Union - 27 countries (from 2020)"}, {"code": "EA21", "name": "Euro area – 21 countries (from 2026)"}, {"code": "EA20", "name": "Euro area – 20 countries (2023-2025)"}, {"code": "EA19", "name": "Euro area - 19 countries  (2015-2022)"}]}]}

The English document, built_at: 2026-08-29T15:37:03.124835+02:00, 3,083 characters. Its summary carries the description and the notes of the ESMS report:

Title
Food waste

Coverage
Environment › Waste › Waste streams

Summary
The indicator is defined as the amount of food waste generated per year divided by the average population of the country. Food waste is measured as fresh mass all along the food value chain, including production sites, processing and manufacture, retail and distribution, restaurants and food services and households.

The indicator measures food waste generated in the production, manufacturing, distribution, food services and households stages of the food supply chain (kg per capita). Currently food waste monitoring is hampered by lack of time series data, since reporting obligations and methodology have just entered into force for the reference year 2020. The indicator differentiates between food waste generated at primary production sectors (agriculture, fisheries), food processing, retail, restaurants and hospitality, and households. The greatest proportion of food waste is generated by households, where most of the consumption happens, complemented by restaurants and hospitality services, while primary production, processign and manufacturing and retail have the possibility of having by products to avoid food waste generation. The definition of food waste in the Waste Framework Directive (2008/98/EC) and the Commission Delegated Decision (EU) 2019/1597 excludes food losses in primary production from food waste, which reduces comparability with previous studies and UN initiatives. Data are available only for one reference year (2020). According to Commission delegated decision (EU) 2019/1597, countries may choose between a set of measurement methodologies defined in Annex III, based on a mandatory study to be conducted at least once every 4 years; moreover, for the first reporting period, Member States may use data already collected under existing arrangements for the year 2017 or later. Therefore, even if data seems robust, there are measurements to be improved in the years to come in order to ensure better comparability between countries. Time series are not present and cannot therefore be analysed. Kg per capita use would benefit from considering resident and non-resident population

Unit
Time frequency: Annual
Waste management operations: Waste collected
Waste categories: Food waste - bio, household and similar waste
Statistical classification of economic activities in the European Community (NACE Rev. 2): Total (aggregate changing according to the context)
Unit of measure: Kilograms per capita
Geopolitical entity (reporting): European Union - 27 countries (from 2020), Euro area – 21 countries (from 2026), Euro area – 20 countries (2023-2025), Euro area - 19 countries  (2015-2022), Belgium, Bulgaria, Czechia, Denmark, Germany, Estonia, Ireland, Greece, Spain, France, Croatia, Italy, Cyprus, Latvia, Lithuania, Luxembourg, Hungary, Malta, Netherlands, Austria, Poland, Portugal, Romania, Slovenia, Slovakia, Finland, Sweden, Iceland, Liechtenstein, Norway, Switzerland, United Kingdom, Montenegro, North Macedonia, Albania, Serbia, Türkiye

Period covered
–

The Italian document has the section headings in Italian (Titolo, Copertura, Sintesi, Unità di misura, Periodo coperto) and the same English text inside them, because Eurostat publishes in English only.

Terms of Eurostat, the GEO codelist in the three languages Eurostat publishes:

{"provider": "eurostat", "language": "de", "scope": "GEO", "code": "DE", "name": "Deutschland"}
{"provider": "eurostat", "language": "en", "scope": "GEO", "code": "DE", "name": "Germany"}
{"provider": "eurostat", "language": "fr", "scope": "GEO", "code": "DE", "name": "Allemagne"}
{"provider": "eurostat", "language": "en", "scope": "GEO", "code": "IT", "name": "Italy"}

OECD: DSD_LAND_USE@DF_LAND_USE, Land use

The catalogue row. OECD publishes a description but no last update, period or report:

{"provider": "oecd", "dataset_id": "DSD_LAND_USE@DF_LAND_USE",
 "title": "Land use", "title_en": "Land use", "names": {"en": "Land use"},
 "description": "Land resources are one of the four components of the natural environment: water, air, land and living resources. In this context land is both: a physical \"milieu\" necessary for the development of natural vegetation as well as cultivated vegetation; a resource for human activities. The data presented here give information concerning land use state and changes (e.g. agricultural land, forest land). Data source(s): Food and Agriculture Organization of the U.N., (FAOSTAT) Contact: ENV.Stat@oecd.org Dataset release date: August 5, 2026 Database documentation Countries metadata",
 "descriptions": {"en": "Land resources are one of the four components of the natural environment: …"},
 "category_path": "Environment and climate change", "category_paths": {"en": "Environment and climate change"},
 "version": "1.0", "structure_id": "DSD_LAND_USE", "agency": "OECD.ENV.EPI",
 "last_update": null, "period_start": null, "period_end": null, "territorial_level": null,
 "active": true, "retired": false, "searchable": true, "served": true,
 "freshness": null, "caveat": null, "filters": [],
 "sources": [{"url": "https://sdmx.oecd.org/public/rest/dataflow/all/DSD_LAND_USE@DF_LAND_USE/latest", "role": "dataset"},
             {"url": "https://www.oecd.org/en/about/terms-conditions.html", "role": "terms", "scope": "dataset-specific"}],
 "licence": "OECD Terms and Conditions", "attribution": "Organisation for Economic Co-operation and Development (OECD)", "keywords": {},
 "report_key": null, "in_hub": null, "seen_hub_at": null, "seen_registry_at": "2026-09-08T00:39:47.275164+02:00",
 "match_text": "Land use\nLand resources are one of the four components of the natural environment: …\nEnvironment and climate change",
 "text_hash": "3288a9ac87ddae042c3814d754e79ab13a0cdc4192d2bacacb8f47b3beae67b5",
 "synced_at": "2026-08-19T20:22:40.791892+02:00"}

The structure. OECD declares the coverage and the number of observations; the key follows the dataset's constraints. In this release the codes stand for themselves as names, because OECD's words come from the vocabulary, which is read again at the first refresh:

{"provider": "oecd", "dataset_id": "DSD_LAND_USE@DF_LAND_USE",
 "harvested_at": "2026-08-28T18:17:38.6943+02:00", "obs_count": 139412, "error": null, "key_order": "constraints",
 "time": {"min": "1961", "max": "2025"},
 "dimensions": [
   {"id": "REF_AREA", "name": "REF_AREA", "values_total": 230, "values": [{"code": "ABW", "name": "ABW"}, {"code": "AFG", "name": "AFG"}, {"code": "AGO", "name": "AGO"}, {"code": "AIA", "name": "AIA"}]},
   {"id": "FREQ", "name": "FREQ", "values_total": 1, "values": [{"code": "A", "name": "A"}]},
   {"id": "MEASURE", "name": "MEASURE", "values_total": 10, "values": [{"code": "AREA", "name": "AREA"}, {"code": "CROP", "name": "CROP"}, {"code": "CROP_TILL", "name": "CROP_TILL"}, {"code": "FOREST", "name": "FOREST"}]},
   {"id": "UNIT_MEASURE", "name": "UNIT_MEASURE", "values_total": 2, "values": [{"code": "KM2", "name": "KM2"}, {"code": "PT_LAR", "name": "PT_LAR"}]}]}

The English document, built_at: 2026-08-28T21:33:04.488921+02:00, 819 characters:

Title
Land use

Coverage
Environment and climate change

Summary
Land resources are one of the four components of the natural environment: water, air, land and living resources. In this context land is both: a physical "milieu" necessary for the development of natural vegetation as well as cultivated vegetation; a resource for human activities. The data presented here give information concerning land use state and changes (e.g. agricultural land, forest land). Data source(s): Food and Agriculture Organization of the U.N., (FAOSTAT) Contact: ENV.Stat@oecd.org Dataset release date: August 5, 2026 Database documentation Countries metadata

Unit
REF_AREA (230 values)
FREQ: A
MEASURE: AREA, CROP, CROP_TILL, FOREST, FOREST_N, FOREST_PL, LAND, MEAD, ORGANIC, OTHER
UNIT_MEASURE: KM2, PT_LAR

Period covered
1961–2025

Terms of OECD, in English and French:

{"provider": "oecd", "language": "en", "scope": "", "code": "ACCOMMODATION_TYPE", "name": "Accommodation type"}
{"provider": "oecd", "language": "fr", "scope": "", "code": "ACCOMMODATION_TYPE", "name": "Type d'hébergement"}

ILO: DF_LUU_XLUX_SEX_NB, Jobs gap

The catalogue row:

{"provider": "ilo", "dataset_id": "DF_LUU_XLUX_SEX_NB",
 "title": "Jobs gap", "title_en": "Jobs gap", "names": {"en": "Jobs gap"},
 "description": "With the aim of promoting international comparability, statistics presented on ILOSTAT are based on standard international definitions wherever feasible and may differ from official national figures. This series is based on the 13th ICLS definitions. For time series comparability, it includes countries that have implemented the 19th ICLS standards, for which data are also available in the Work Statistics -- 19th ICLS (WORK) database. The jobs gap is defined as the sum of the unemployed, the potential labour force and willing non-jobseekers. For more information, refer to the Labour Force Statistics (LFS and STLFS) database description.",
 "descriptions": {"en": "With the aim of promoting international comparability, …"},
 "category_path": "Unemployment and labour underutilization", "category_paths": {"en": "Unemployment and labour underutilization"},
 "version": "1.0", "structure_id": "LUU_XLUX_SEX_NB", "agency": "ILO",
 "last_update": "2026-08-27T07:07:14", "period_start": null, "period_end": null, "territorial_level": null,
 "active": true, "retired": false, "searchable": true, "served": true,
 "freshness": null, "caveat": null, "filters": [],
 "sources": [{"url": "https://sdmx.ilo.org/rest/dataflow/ILO/DF_LUU_XLUX_SEX_NB/latest", "role": "dataset"},
             {"url": "https://www.ilo.org/rights-and-permissions", "role": "terms", "scope": "dataset-specific"}],
 "licence": "ILO rights and permissions", "attribution": "International Labour Organization (ILO)", "keywords": {},
 "report_key": null, "in_hub": null, "seen_hub_at": null, "seen_registry_at": "2026-09-08T00:39:48.771941+02:00",
 "match_text": "Jobs gap\nWith the aim of promoting international comparability, …\nUnemployment and labour underutilization",
 "text_hash": "4683b3bb8abf5c35b1f75662eccd260f416366b4efeb0c83cc204b212744ba49",
 "synced_at": "2026-08-18T20:22:19.038213+02:00"}

The structure, four dimensions; the dimension names are what ILO's structure carries (Classification: REF_AREA), and the codes of SEX already have their names while those of REF_AREA wait for the vocabulary:

{"provider": "ilo", "dataset_id": "DF_LUU_XLUX_SEX_NB",
 "harvested_at": "2026-08-28T19:00:03.52707+02:00", "obs_count": null, "error": null, "key_order": "constraints", "time": {},
 "dimensions": [
   {"id": "REF_AREA", "name": "Classification: REF_AREA", "values_total": 210, "values": [{"code": "AFG", "name": "AFG"}, {"code": "AGO", "name": "AGO"}, {"code": "ALB", "name": "ALB"}, {"code": "ARM", "name": "ARM"}]},
   {"id": "FREQ", "name": "Classification: FREQ", "values_total": 2, "values": [{"code": "A", "name": "A"}, {"code": "Q", "name": "Q"}]},
   {"id": "MEASURE", "name": "Classification: MEASURE", "values_total": 1, "values": [{"code": "LUU_XLUX_NB", "name": "LUU_XLUX_NB"}]},
   {"id": "SEX", "name": "Classification: SEX", "values_total": 4, "values": [{"code": "SEX_T", "name": "Total"}, {"code": "SEX_M", "name": "Male"}, {"code": "SEX_F", "name": "Female"}, {"code": "SEX_O", "name": "Other"}]}]}

The English document, built_at: 2026-08-29T18:14:35.125189+02:00, 892 characters:

Title
Jobs gap

Coverage
Unemployment and labour underutilization

Summary
With the aim of promoting international comparability, statistics presented on ILOSTAT are based on standard international definitions wherever feasible and may differ from official national figures. This series is based on the 13th ICLS definitions. For time series comparability, it includes countries that have implemented the 19th ICLS standards, for which data are also available in the Work Statistics -- 19th ICLS (WORK) database. The jobs gap is defined as the sum of the unemployed, the potential labour force and willing non-jobseekers. For more information, refer to the Labour Force Statistics (LFS and STLFS) database description.

Unit
Classification: REF_AREA (210 values)
Classification: FREQ: A, Q
Classification: MEASURE: LUU_XLUX_NB
Classification: SEX: Total, Male, Female, Other

Period covered
–

Terms of ILO, in the three languages ILO publishes:

{"provider": "ilo", "language": "en", "scope": "", "code": "AGE", "name": "Age"}
{"provider": "ilo", "language": "es", "scope": "", "code": "AGE", "name": "Edad"}
{"provider": "ilo", "language": "fr", "scope": "", "code": "AGE", "name": "Âge"}

DoveVannoINostriSoldi: ipa_enti, Enti pubblici IPA

The catalogue row. A dataset of this provider names its upstream source with owner and cadence, says its data are read live at question time (freshness: live), and lists the filters it accepts; this one is a search over the Indice PA, which the reader does not serve as data (served: false):

{"provider": "dvns", "dataset_id": "ipa_enti",
 "title": "Enti pubblici IPA", "title_en": null, "names": {"it": "Enti pubblici IPA"},
 "description": "Ricerca e scheda degli enti nell’Indice PA.", "descriptions": {"it": "Ricerca e scheda degli enti nell’Indice PA."},
 "category_path": null, "category_paths": {},
 "version": null, "structure_id": null, "agency": "AgID",
 "last_update": null, "period_start": null, "period_end": null, "territorial_level": null,
 "active": true, "retired": false, "searchable": true, "served": false,
 "freshness": "live", "caveat": null, "filters": ["query", "code", "limit", "offset"],
 "sources": [{"id": "ipa", "url": "https://www.indicepa.gov.it/ipa-dati/dataset/enti", "name": "Indice PA (IPA)", "owner": "AgID", "cadence": "Giornaliera"}],
 "licence": null, "attribution": null, "keywords": {},
 "report_key": null, "in_hub": null, "seen_hub_at": null, "seen_registry_at": "2026-09-08T00:39:58.575676+02:00",
 "match_text": "Enti pubblici IPA\n\nRicerca e scheda degli enti nell’Indice PA.",
 "text_hash": "d0038abcc847934d0f21af63974bca336969488dd584bb14facf12b0a8cb5bc0",
 "synced_at": "2026-08-28T21:08:56.180752+02:00"}

The structure: the four filters, as dimensions without codes:

{"provider": "dvns", "dataset_id": "ipa_enti",
 "harvested_at": "2026-08-28T21:09:55.868008+02:00", "obs_count": null, "error": null, "key_order": "constraints", "time": {},
 "dimensions": [
   {"id": "query", "name": "query", "values_total": 0, "values": []},
   {"id": "code", "name": "code", "values_total": 0, "values": []},
   {"id": "limit", "name": "limit", "values_total": 0, "values": []},
   {"id": "offset", "name": "offset", "values_total": 0, "values": []}]}

Both documents in full. Italian, built_at: 2026-08-29T18:27:26.52153+02:00, 226 characters:

Titolo
Enti pubblici IPA

Sintesi
Ricerca e scheda degli enti nell’Indice PA.

Filtri accettati da questo dataset
ricerca libera, codice, dimensione della pagina, scarto della pagina

Fonti
Indice PA (IPA) · AgID · Giornaliera

English, built_at: 2026-08-28T21:11:10.055396+02:00, 220 characters:

Title
IPA public bodies

Summary
Search and profile of bodies in the Public Administration Index.

Filters this dataset accepts
free-text search, code, page size, page offset

Sources
Indice PA (IPA) · AgID · Giornaliera

This provider publishes no codelists: it has no rows in opendata_terms.

Cruscotto Italia: ambiente, Ambiente e rifiuti

The catalogue row. Every dataset of this provider is an indicator served per municipality (territorial_level: municipality), maintained by the agency the row names (here ISPRA), with no source addresses recorded:

{"provider": "cruscotto", "dataset_id": "ambiente",
 "title": "Ambiente e rifiuti", "title_en": null, "names": {"it": "Ambiente e rifiuti"},
 "description": "Ambiente e rifiuti", "descriptions": {"it": "Ambiente e rifiuti"},
 "category_path": null, "category_paths": {},
 "version": null, "structure_id": null, "agency": "ISPRA",
 "last_update": null, "period_start": null, "period_end": null, "territorial_level": "municipality",
 "active": true, "retired": false, "searchable": true, "served": true,
 "freshness": null, "caveat": null, "filters": [], "sources": [],
 "licence": null, "attribution": null, "keywords": {},
 "report_key": null, "in_hub": null, "seen_hub_at": null, "seen_registry_at": "2026-09-08T00:39:59.021289+02:00",
 "match_text": "Ambiente e rifiuti",
 "text_hash": "1b2211169cc305bd09762c97b6b1cfb2cde3f633f9829638c81bee6a7f5ef83b",
 "synced_at": "2026-08-28T21:08:57.164596+02:00"}

The structure, one dimension, the municipality, whose codes are the ISTAT municipality codes and are not listed:

{"provider": "cruscotto", "dataset_id": "ambiente",
 "harvested_at": "2026-08-28T21:12:45.719974+02:00", "obs_count": null, "error": null, "key_order": "constraints", "time": {},
 "dimensions": [{"id": "comune", "name": "comune", "values_total": 0, "values": []}]}

The document, Italian only, built_at: 2026-08-28T22:06:01.568814+02:00:

Titolo
Ambiente e rifiuti

This provider publishes no codelists either.

Territories and maps

The territory codes in these structures are the identifiers the shapes of Gramscii-Git/boundaries carry as aliases, so a table keyed by them is drawn on a map without a lookup step: ISTAT's REF_AREA uses the NUTS codes ISTAT publishes under (ITE, ITE4, ITE41 above) and ITTER107 the ISTAT codes of regions, provinces and municipalities; Eurostat's geo uses NUTS 2024 codes and Eurostat's country codes (EL, UK); OECD's and ILO's REF_AREA use ISO 3166-1 alpha-3 (AFG, ABW); DoveVannoINostriSoldi and Cruscotto Italia use ISTAT municipality codes.

Format

A gzip-compressed tar holding one file per table plus a manifest:

manifest.json
opendata_catalog.jsonl
opendata_structures.jsonl
opendata_structure_dims.jsonl
opendata_terms.jsonl
opendata_labels.jsonl
opendata_meta_reports.jsonl
opendata_documents.jsonl

Each .jsonl file holds one JSON object per line, one line per row, with the table's columns as keys, ordered by the table's key. Values are written as the database holds them: timestamps in ISO 8601 with their offset, JSON columns nested, absent values as null. The rows shown above are such lines, pretty-printed.

manifest.json:

{
  "schema_version": 1,
  "taken_at": "2026-09-07T22:36:26.908691+00:00",
  "tables": {
    "opendata_catalog": 15990,
    "opendata_structures": 15854,
    "opendata_structure_dims": 0,
    "opendata_terms": 1414937,
    "opendata_labels": 8,
    "opendata_meta_reports": 1169,
    "opendata_documents": 31673
  }
}

Reading it

Unpack the tar, read each .jsonl line by line, and apply every row by its table's key: insert where the key is new, replace where the row is newer by the table's "newer by" column, leave the rest. The archive is identified by the address it is served at, its SHA-256 and its size:

url:    https://huggingface.co/datasets/Gramscii-IT/open-data-catalogue/resolve/main/open-data-catalogue.tar.gz
sha256: 1e9eafeb4e691dc810d0de73fe387098b5029fcb740934d2fd4eb4ada56864e9
bytes:  55034326

A reader that starts from it holds what the six providers published on the date above, and needs to ask them only for what moved since.

What the archive does not carry

• Observations. The numbers themselves are read from the providers at question time, per dataset and period.
• A search index. Whoever searches opendata_documents builds the vectors over them with their own embedder.
• Credentials. Nothing in the archive is a secret; every provider here is a public, unauthenticated API.

Known gaps in this release

• 453 structures carry a recorded error instead of dimensions: 123 of ISTAT and 330 of OECD, where the provider answered with an error or not at all. The next release retries them.
• opendata_structure_dims is empty and the scope of every row in opendata_terms is the bare codelist identifier, as an earlier reader wrote it. The reader that produces this archive now writes codelist:GEO and conceptscheme:…; the next release carries the vocabulary in that form, read again from the providers.
• The OECD and ILO structures carry their codes as names (REF_AREA named REF_AREA, AFG named AFG): the words come from that vocabulary and arrive with it.

A later release replaces this file; the digest tells them apart.

Keeping it current

The archive is written and uploaded by the script in Gramscii-Git/open-data-catalogue, run on the deployment that holds the catalogue: it asks the providers for what moved, rebuilds the documents, exports the tables, uploads the archive with its manifest and digest, and prints the new address, digest and size.

Integrity

1e9eafeb4e691dc810d0de73fe387098b5029fcb740934d2fd4eb4ada56864e9  open-data-catalogue.tar.gz

shasum -a 256 -c SHA256SUMS

The digest identifies this archive, never the providers' own files.

Licences

Every row of opendata_catalog names the licence and the attribution its provider publishes for that dataset; those terms follow the row wherever it goes, and where the provider's terms are more restrictive they decide what may be done with that dataset's metadata and documents.

Gramscii's assembly, the archive as a compilation and the documents built from the providers' metadata, is released under Creative Commons Attribution 4.0 International. That covers what Gramscii added, never the source data beneath it.

• ISTAT: CC BY 4.0, with the exceptions ISTAT states per dataset, recorded in the row. Attribution: Istituto nazionale di statistica (ISTAT).
• Eurostat: the Eurostat reuse policy. Attribution: European Commission, Eurostat.
• OECD: the OECD terms and conditions. Attribution: Organisation for Economic Co-operation and Development (OECD).
• ILO: ILO rights and permissions. Attribution: International Labour Organization (ILO).
• DoveVannoINostriSoldi: the licence and attribution of each dataset's own source, as recorded in its row; four datasets state none.
• Cruscotto Italia: AgID's terms; not recorded in this release.