fix(destinations): school rows and the England reference are different grains

The annual DAG died with a BrokenPipeError from Meltano's log writer, which
is several frames from the cause: target-postgres exited first and the tap
saw its stdout close.

The tap declared primary_keys = [urn, ...] while emitting urn=None for the
national rows, and target-postgres turns primary_keys into a NOT NULL
constraint. The first national row of the run failed the insert and took
the loader with it. Every other tap in this repo keys on non-null columns.

Carrying two grains in one stream was the actual mistake, so the fix is to
separate them rather than paper over the null: four streams now, with
ees_ks4/ks5_destinations_national carrying no urn column at all — a school
identifier that is null in every row is a grain mismatch, not a column.
The staging models split the same way and the national mart reads the new
pair instead of filtering `where urn is null`.

Verified against the live API: the school stream yields 135,240 rows over
4,508 schools with no duplicate keys, no null key columns and all 31,382
suppression sentinels intact; the national streams yield 30 and 33 rows
with no urn column.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvdDKvFFSZuMVDH5fEyTob

pipeline/transform/models/marts/fact_destination_national.sql

@@ -22,8 +22,7 @@ select
     pupils,
     percentage,
     status
-from {{ ref('stg_ees_ks4_destinations') }}
-where urn is null
+from {{ ref('stg_ees_ks4_destinations_national') }}
 
 union all
 
@@ -36,5 +35,4 @@ select
     pupils,
     percentage,
     status
-from {{ ref('stg_ees_ks5_destinations') }}
-where urn is null
+from {{ ref('stg_ees_ks5_destinations_national') }}

pipeline/transform/models/staging/_stg_sources.yml

@@ -47,6 +47,17 @@ sources:
           16-18 study leavers destinations. Same grain, same suppression
           caveat, and only institutions with post-16 provision appear.
 
+      - name: ees_ks4_destinations_national
+        description: >
+          England KS4 destination measures by pupil group. A separate table
+          from ees_ks4_destinations because it is a separate grain — no school,
+          so no urn column. Same 'c' suppression caveat.
+
+      - name: ees_ks5_destinations_national
+        description: >
+          England 16-18 destination measures by pupil group. Same grain and
+          caveat as ees_ks4_destinations_national.
+
       - name: ees_ks4_performance
         description: KS4 performance tables (long format — one row per school × breakdown × sex)
 

pipeline/transform/models/staging/stg_ees_ks4_destinations.sql

@@ -1,7 +1,9 @@
 {{ config(materialized='table') }}
 
--- Staging model: KS4 leavers destinations, school level plus the England
--- reference (which carries a null urn).
+-- Staging model: KS4 leavers destinations, school level.
+--
+-- School rows only. The England reference is a different grain and lives in
+-- stg_ees_ks4_destinations_national.
 --
 -- DELIBERATELY DOES NOT USE safe_numeric. That macro maps every EES sentinel
 -- (z, c, x, q, u) to NULL, which is right for attainment — there, "suppressed"
@@ -14,13 +16,12 @@
 
 with source as (
     select * from {{ source('raw', 'ees_ks4_destinations') }}
-    -- National rows carry a null urn and feed fact_destination_national.
-    where (urn is null or urn = '' or urn ~ '^[0-9]+$')
+    where urn ~ '^[0-9]+$'
       and time_period ~ '^[0-9]+$'
 )
 
 select
-    case when urn ~ '^[0-9]+$' then cast(trim(urn) as integer) end as urn,
+    cast(trim(urn) as integer) as urn,
     cast(trim(time_period) as integer)                            as year,
     trim(pupil_group)                                             as pupil_group,
     trim(destination_measure)                                     as destination_measure,

pipeline/transform/models/staging/stg_ees_ks4_destinations_national.sql

@@ -0,0 +1,38 @@
+{{ config(materialized='table') }}
+
+-- Staging model: England KS4 destination measures — the national
+-- reference the school sections compare against.
+--
+-- A separate model because it is a separate grain: there is no school here, and
+-- carrying these rows in the school table meant a null urn inside the primary
+-- key, which the Postgres loader rejects.
+--
+-- DELIBERATELY DOES NOT USE safe_numeric, for the same reason as the school
+-- model: 'suppressed' and 'not applicable' are different claims.
+
+with source as (
+    select * from {{ source('raw', 'ees_ks4_destinations_national') }}
+    where time_period ~ '^[0-9]+$'
+)
+
+select
+    cast(trim(time_period) as integer) as year,
+    trim(pupil_group)                  as pupil_group,
+    trim(destination_measure)          as destination_measure,
+
+    case when cohort_pupils ~ '^[0-9]+$'
+         then cast(cohort_pupils as integer) end as cohort_pupils,
+
+    case when pupils_raw ~ '^[0-9]+$'
+         then cast(pupils_raw as integer) end as pupils,
+
+    case when percentage_raw ~ '^-?[0-9]+(\.[0-9]+)?$'
+         then cast(percentage_raw as numeric) end as percentage,
+
+    case
+        when pupils_raw ~ '^[0-9]+$'        then 'published'
+        when lower(trim(pupils_raw)) = 'c'  then 'suppressed'
+        else 'not_applicable'
+    end as status
+
+from source

pipeline/transform/models/staging/stg_ees_ks5_destinations.sql

@@ -1,8 +1,10 @@
 {{ config(materialized='table') }}
 
--- Staging model: 16-18 study leavers destinations, institution level plus the
--- England reference (which carries a null urn). Only sixth forms and colleges
--- appear here, so a secondary with no post-16 provision has no rows at all.
+-- Staging model: 16-18 study leavers destinations, institution level.
+--
+-- Only sixth forms and colleges appear, so a secondary with no post-16
+-- provision has no rows at all. The England reference is a different grain and
+-- lives in stg_ees_ks5_destinations_national.
 --
 -- DELIBERATELY DOES NOT USE safe_numeric. That macro maps every EES sentinel
 -- (z, c, x, q, u) to NULL, which is right for attainment — there, "suppressed"
@@ -15,13 +17,12 @@
 
 with source as (
     select * from {{ source('raw', 'ees_ks5_destinations') }}
-    -- National rows carry a null urn and feed fact_destination_national.
-    where (urn is null or urn = '' or urn ~ '^[0-9]+$')
+    where urn ~ '^[0-9]+$'
       and time_period ~ '^[0-9]+$'
 )
 
 select
-    case when urn ~ '^[0-9]+$' then cast(trim(urn) as integer) end as urn,
+    cast(trim(urn) as integer) as urn,
     cast(trim(time_period) as integer)                            as year,
     trim(pupil_group)                                             as pupil_group,
     trim(destination_measure)                                     as destination_measure,

pipeline/transform/models/staging/stg_ees_ks5_destinations_national.sql

@@ -0,0 +1,38 @@
+{{ config(materialized='table') }}
+
+-- Staging model: England KS5 destination measures — the national
+-- reference the school sections compare against.
+--
+-- A separate model because it is a separate grain: there is no school here, and
+-- carrying these rows in the school table meant a null urn inside the primary
+-- key, which the Postgres loader rejects.
+--
+-- DELIBERATELY DOES NOT USE safe_numeric, for the same reason as the school
+-- model: 'suppressed' and 'not applicable' are different claims.
+
+with source as (
+    select * from {{ source('raw', 'ees_ks5_destinations_national') }}
+    where time_period ~ '^[0-9]+$'
+)
+
+select
+    cast(trim(time_period) as integer) as year,
+    trim(pupil_group)                  as pupil_group,
+    trim(destination_measure)          as destination_measure,
+
+    case when cohort_pupils ~ '^[0-9]+$'
+         then cast(cohort_pupils as integer) end as cohort_pupils,
+
+    case when pupils_raw ~ '^[0-9]+$'
+         then cast(pupils_raw as integer) end as pupils,
+
+    case when percentage_raw ~ '^-?[0-9]+(\.[0-9]+)?$'
+         then cast(percentage_raw as numeric) end as percentage,
+
+    case
+        when pupils_raw ~ '^[0-9]+$'        then 'published'
+        when lower(trim(pupils_raw)) = 'c'  then 'suppressed'
+        else 'not_applicable'
+    end as status
+
+from source