feat: add hospital beds by NUTS2 region (HLTH_RS_BDSRG2) (#75)

datasets/eurostat-hospital-beds-nuts2/dataset.yml

@@ -0,0 +1,49 @@
+root: "../../out"
+schema_version: 1
+
+dataset:
+  name: "eurostat_hospital_beds_nuts2"
+  source_id: "eurostat"
+  years: [2026]
+  time_coverage:
+    start_year: 1993
+    end_year: 2024
+
+raw:
+  output_policy: overwrite
+  sources:
+    - name: "eurostat_hlth_rs_bdsrg2"
+      type: "script"
+      args:
+        command: "../../connectors/tsv_normalize.py --flow HLTH_RS_BDSRG2 --output hlth_rs_bdsrg2_normalized.parquet"
+        output: "hlth_rs_bdsrg2_normalized.parquet"
+        filename: "hlth_rs_bdsrg2_normalized.parquet"
+      primary: true
+
+clean:
+  sql: "sql/clean.sql"
+  read:
+    source: auto
+    mode: latest
+    delim: ","
+    encoding: utf-8
+    header: true
+  validate:
+    min_rows: 100
+
+mart:
+  tables:
+    - name: "mart_eurostat_hospital_beds_nuts2"
+      sql: "sql/mart.sql"
+  required_tables:
+    - "mart_eurostat_hospital_beds_nuts2"
+
+validation:
+  fail_on_error: true
+
+registry:
+  dataflow: "HLTH_RS_BDSRG2"
+  theme: "Available beds in hospitals by NUTS 2 region"
+  nuts_level: 2
+  dimensions: ["freq", "unit", "geo"]
+  description: "Available beds in hospitals by NUTS 2 region"

datasets/eurostat-hospital-beds-nuts2/sql/clean.sql

@@ -0,0 +1,21 @@
+-- clean.sql: HLTH_RS_BDSRG2 — posti letto ospedalieri per regione NUTS2
+SELECT
+    r.freq,
+    r.unit,
+    r.geo,
+    r.year,
+    f.label_en AS freq_label_en,
+    u.label_en AS unit_label_en,
+    g.label_en AS geo_label_en,
+    g.nuts_level,
+    g.parent_code AS nuts_parent_code,
+    gp.label_en AS nuts_parent_label_en,
+    r.value,
+    r.flag,
+    fl.description_en AS flag_desc_en
+FROM raw_input r
+LEFT JOIN read_csv('codelists/freq.csv', auto_detect=true, delim=',', header=true) f ON r.freq = f.freq
+LEFT JOIN read_csv('codelists/units.csv', auto_detect=true, delim=',', header=true) u ON r.unit = u.unit
+LEFT JOIN read_csv('codelists/geo.csv', auto_detect=true, delim=',', header=true) g ON r.geo = g.code
+LEFT JOIN read_csv('codelists/flags.csv', auto_detect=true, delim=',', header=true) fl ON r.flag = fl.flag
+LEFT JOIN read_csv('codelists/geo.csv', auto_detect=true, delim=',', header=true) gp ON g.parent_code = gp.code

datasets/eurostat-hospital-beds-nuts2/sql/mart.sql

@@ -0,0 +1,20 @@
+-- mart.sql: Posti letto ospedalieri NUTS2 — Italia
+SELECT
+    year,
+    geo,
+    geo_label_en,
+    nuts_level,
+    nuts_parent_code,
+    nuts_parent_label_en,
+    freq,
+    freq_label_en,
+    unit,
+    unit_label_en,
+    value,
+    flag,
+    flag_desc_en
+FROM clean_input
+WHERE geo LIKE 'IT%'
+  AND nuts_level = 'NUTS2'
+  AND value IS NOT NULL
+ORDER BY year DESC, geo

docs/dataset-registry.md

@@ -24,6 +24,7 @@
 | `eurostat-soil-erosion-nuts3` | `AEI_PR_SOILER` | Environment / Estimated soil erosion by water | 5 (freq, levels, clc18, unit, geo) | 523,200 | ✅ |
 | `eurostat-area-nuts3` | `REG_AREA3` | Geography / Area by land use category | 4 (freq, landuse, unit, geo) | 65,240 | ✅ |
 | `eurostat-physicians-nuts2` | `HLTH_RS_PHYSREG` | Health / Physicians by NUTS2 region | 3 (freq, unit, geo) | 27,360 | ✅ |
+| `eurostat-hospital-beds-nuts2` | `HLTH_RS_BDSRG2` | Health / Available beds in hospitals by NUTS2 region | 3 (freq, unit, geo) | 25,888 | ✅ |
 | `eurostat-fertility-nuts3` | `DEMO_R_FIND3` | Demography / Fertility indicators | 4 (freq, indic_de, unit, geo) | 76,392 | ✅ |
 | `eurostat-business-demography-nuts3` | `BD_SIZE_R3` | Business / Business demography by size class | 5 (freq, indic_sb, sizeclas, nace_r2, geo) | 120,955 | ✅ |