pc_base¶
ordinary postcode: validated small-user split suffix removed; can be ambiguous
| Table | Type | Nullable | Category | In the CSV export | SQL Server type |
|---|---|---|---|---|---|
history table (postcode_simd_history, SPD) |
string | no | derived from the NRS record | yes | nvarchar(200) |
Provenance in the history table (postcode_simd_history, SPD)¶
Where it comes from
| Input | Field | Condition | Reaches the row by |
|---|---|---|---|
SPD/spd_postcodeindex_cut_26_2_csv/SmallUser.csv (pinned, SHA256 fc82f1091ac1…) |
Postcode |
the row itself | |
SPD/spd_postcodeindex_cut_26_2_csv/LargeUser.csv (pinned, SHA256 dcb060cbce01…) |
Postcode |
the row itself | |
SPD/spd_postcodeindex_cut_26_2_csv/SmallUser.csv (pinned, SHA256 fc82f1091ac1…) |
SplitIndicator |
Y marks a split part, whose A, B or C suffix is validated and removed; large users are never split | the row itself |
Scope: the row itself.
What was done to it: normalised. The ordinary postcode: pc_norm with the split suffix removed where SplitIndicator is Y on a small-user record. (code: simd_ingest/core/spd.py)
How it is checked (what each check tests; whether a build passed it is in that build's manifest.json["checks"])
spd.primary_key_unique: the key (pc_norm, introduced_on) is unique and never nullspd.primary_key_nonnull: the key (pc_norm, introduced_on) is unique and never nullreadback.history.index_columns: the saved file's NRS columns equal the source, re-read