API JSON + Summary PDF for cert 9501-3059-8202-7356-0204. RR/Mid-
terrace flat, 4 building storeys, TFA 113.08 m², mains gas boiler
(PCDB idx 19007), age band B. Worksheet target unrounded SAP
**68.5252**.
Second boiler cert per the per-cert mapper validation workflow:
Summary path proves itself against the worksheet (Layer 2 1e-4 pin),
then the API path catches up (Layer 4 1e-4 pin) — mirrors the cert
0330 cycle.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Adds the (API JSON + Summary PDF) fixtures for cert
0380-2471-3250-2596-8761 — the Air Source Heat Pump pilot
identified in the handover. Property: 16 Beech Lea, WIGTON CA7 5JY
(semi-detached bungalow, ASHP PCDB idx 104568).
Source: API JSON fetched via EpcClientService. Summary PDF copied
from `sap worksheets/Additional data with api/
0380-2471-3250-2596-8761/Summary_000899.pdf`.
Worksheet target: SAP 88.5104 (continuous), from `dr87-0001-000899
.pdf`.
**This is the HP pilot, intentionally deferred.** Initial probe on
these fixtures (uncommitted before this slice):
- Summary mapper cascade SAP: 18.08 (Δ -70.43 vs worksheet)
- API mapper cascade SAP: 70.14 (Δ -18.37 vs worksheet)
Both paths are catastrophically RED. The mapper has never been
validated against an ASHP cert and there's substantial cascade
plumbing required:
- API mapper correctly identifies the HP (COP 2.3) but fabric HLC
is 104 W/K vs the ~50 W/K needed for SAP 88.51.
- Summary mapper misreads the HP as an 80%-efficient boiler
(catastrophic).
- 7 of 9 newly-staged certs are ASHPs (6 share PCDB idx 104568,
cert 9418 uses 102421), so a shared HP-cascade fix will likely
close most of them at once.
Stashed here so the next agent can pick up the HP workstream
without needing to refetch from the EPB API. Recommend not
attempting these slices until the boiler workflow (cert 0330) is
proven; the boiler cascade is the reference shape and HP work
should build on a known-good baseline. Handover §"Heat-pump
workstream sketch" outlines the likely 15-30 slice queue.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Adds the (API JSON + Summary PDF) fixtures for cert
0330-2249-8150-2326-4121 — the boiler pilot identified in the
handover. Property: 17 Summerfield Road, MANCHESTER M22 1AE
(mid-terrace house, mains gas boiler PCDB idx 10241, age D).
Source: API JSON fetched via EpcClientService from
https://api.get-energy-performance-data.communities.gov.uk
(OPEN_EPC_API_TOKEN). Summary PDF copied from
`sap worksheets/Additional data with api/0330-2249-8150-2326-4121/
Summary_000897.pdf` (where the user provided the triple).
Worksheet target: SAP 61.5993 (continuous), from `dr87-0001-000897
.pdf` in the same source directory.
Current state on these fixtures (uncommitted before this slice):
- Summary mapper cascade SAP: 62.0660 (Δ +0.4667 vs worksheet)
- API mapper cascade SAP: 63.7446 (Δ +2.1453 vs worksheet)
Both paths RED at 1e-4. Two specific cascade-component gaps
identified in the handover for follow-up slices:
1. Windows HLC +6.71 W/K (API vs Summary) — likely glazing_type=14
not in Slice 93's `_API_GLAZING_TYPE_TO_TRANSMISSION` (only
codes 3 and 13 mapped).
2. HW kWh +1060 (API 3172.65 vs Summary 2112.00) — §4 subsystem
gap; needs occupancy/shower/cylinder probe.
This commit stages the fixtures only — no tests added yet. The
follow-up slice should add a RED Layer 2 test (Summary path 1e-4
vs 61.5993) and proceed slice-by-slice.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
`from_elmhurst_site_notes` hard-coded `extensions_count=0` regardless of
how many extensions the survey lodged. The 6 cohort certs from Slices
47-53 all happened to have 0-2 extensions whose count nothing
load-bearing read, so this latent bug was invisible. Cert 001479
(Summary_001479.pdf, GOV.UK EPB cert 0535-9020-6509-0821-6222) has Main
+ Extension 1 + Extension 2 and is the first cohort cert with a real
API counterpart — accurate `extensions_count` becomes load-bearing the
moment the cross-mapper parity assertion compares API vs Elmhurst
EpcPropertyData side by side.
No SAP-cascade impact (the cascade iterates `sap_building_parts`, not
`extensions_count`) — but a real data-integrity bug surfaced by the
cross-mapper diff. Adds Summary_001479.pdf as a new chain-test fixture
and `_SUMMARY_001479_PDF` constant for follow-up slices that will
land per-bp ages, exposed floors, secondary-heating SAP codes, etc.
All 9 chain tests green; 321 mapper/site-notes/rdsap tests green;
pyright net-zero (35-error baseline preserved on mapper.py).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The §11 Windows table in the Summary PDF doesn't lay out identically
across the cohort. Three new quirks added to the layout-style parser
so the remaining 5 certs can be debugged with windows actually
extracted:
1. `Wood 0.70` combined frame_type+frame_factor line — previously the
parser expected them on separate lines (data+1 / data+2) and
rejected the window when the joined form appeared.
2. Trailing glazing-type on the data line — `1.22 1.76 2.15 Double
pre 2002` is the joined-cell variant in 000516; the W/H/Area
anchor now captures the trailing phrase as an optional 4th group
and feeds it through as `inline_glazing_type`, bypassing the
separate-line glazing-prefix scan.
3. Cross-window gap with no glazing marker — `_partition_after_manuf`
now falls back to "second orientation token in gap" when no
glazing-type-prefix word appears. Covers the 000516 layout where
each window has prefix+suffix orient tokens (no inline orient)
and the glazing-type is joined-to-data.
The 5 remaining Summary PDFs are copied into
`backend/documents_parser/tests/fixtures/` ready for per-cert mapper
work. Mirror pin tests deferred — each cert still has its own diff
to close (handover in NEXT_AGENT_PROMPT.md documents the per-cert
state, e.g. 000477 needs secondary-heating extraction, 000516 needs
roof-window separation).
Current cohort SAP deltas vs the U985 worksheet PDFs (target 1e-4):
000474 0.0000 ✓
000477 +6.3655 secondary heating + lighting
000480 +8.2695 diagnosis pending
000487 +8.1433 extractor still drops windows
000490 +5.6551 diagnosis pending
000516 +5.9812 roof-window separation
Wider regression stays green (754 pass). Pyright net-zero on
touched files.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The 6 worksheet fixtures build EpcPropertyData by hand, validating the cascade in isolation from the mapper. This commit lands the first half of the OTHER validation: Summary_000474.pdf → ElmhurstSiteNotesExtractor → from_elmhurst_site_notes → EpcPropertyData, asserting it produces the same shape as the hand-built fixture. Test is strict-xfail on sap_building_parts count (mapper produces 1, cert lodges 3). Includes a pdftotext-layout preprocessor that converts spatial label/value layout into the Textract-style sequence the existing extractor expects (test-only). Full punch list of 28 mapper-output diffs captured in project memory.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
