Delete the dead geopandas runtime mapper (C-39) by Polichinel · Pull Request #42 · views-platform/views-postprocessing

Polichinel · 2026-06-24T02:40:23Z

Delete the dead geopandas runtime mapper (C-39)

ADR-011 replaced the runtime PriogridCountryMapper with the precomputed GAUL lookup (GaulLookupEnricher). The expert-code-review verified the mapper is orphaned in production — the live manager unfao/managers/unfao.py never references it, nothing re-exports it. This deletes the whole dead cluster (CCP: it exists only to support the mapper).

What's removed (46 files, ~5.1k deletions)

views_postprocessing/unfao/mapping/ — the 3,171-line geopandas/shapely module
views_postprocessing/shapefiles/ — the 1.3 GB bundle (GAUL L1/L2, Natural Earth, PRIO-GRID)
mapper tests (test_mapping, test_regression, test_integration) + conftest.py (the gpd.read_file intercept + mapper fixtures)
the two ADR-011 diff scripts (diff_enrichment.py, plot_enrichment_diff.py)
docs/CICs/PriogridCountryMapper.md
cachetools from pyproject.toml (mapper-only)

Why it matters

geopandas + shapely are now gone from the codebase (zero references) — they survived solely to feed this cluster.
Dissolves the old-mapper risk cluster — C-02/05/06/11/12/14/16/17/19/20/21, D-01/D-02 describe code that no longer exists. C-39 marked resolved in the register.
The package now screams its real responsibility: lookup-based enrichment, not runtime geometry.

Safety

Pure deletion — no production logic touched. The manager, enricher, schema, frames adapters, and all of reconciliation/ are untouched and clean of the mapper.
Deletion broke nothing: 101 passed, 43 xfailed, 0 failed, ruff clean. The keep-tests run with no conftest.

Second commit (unrelated, included for a green suite)

Re-xfails test_datafactory_deploy_readiness::test_version_bumped_past_latest_tag — it flipped red because datafactory tagged v1.4.0 (the over-strict cross-repo probe), nothing to do with the mapper.

🤖 Generated with Claude Code

ADR-011 replaced the runtime PriogridCountryMapper with the GAUL lookup (GaulLookupEnricher); the mapper was verified orphaned in production (the live manager never references it). This removes the whole dead cluster (CCP): the 3,171-line geopandas module, the 1.3 GB shapefile bundle, the mapper tests + conftest, the two ADR-011 diff scripts, and the mapper CIC -- 46 files, ~5.1k deletions. geopandas + shapely are now gone from the codebase; cachetools dropped from pyproject (mapper-only). Dissolves the old-mapper concern cluster (C-02/05/06/11/12/14/16/17/19/20/21, D-01/D-02); C-39 marked resolved. Deletion broke nothing -- keep-tests green. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…v1.4.0) datafactory cut + tagged v1.4.0, so pyproject version == latest tag and this over-strict cross-repo deploy probe flips red. Re-mark it xfail(strict): the gate flaps with datafactory's release cadence and re-promotes (fails) when datafactory bumps past the tag. Unrelated to the mapper deletion. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…9 follow-up) Register-curation pass after the mapper deletion (C-39). Move the 11 now-dead mapper concerns (C-02, C-04, C-05, C-06, C-11, C-12, C-14, C-16, C-17, C-20, C-21) and 2 disagreements (D-01, D-02) from Open to Resolved — the code they describe no longer exists. C-19 narrowed (the mapper raises are gone; 3 unfao.py raises remain, issue #13); C-08 kept open (its datafactory area-math dimension stays under C-31). Counts: open 36 -> 25, resolved 4 -> 15. Bookkeeping only. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…eletion The two doc-accuracy leftovers from the C-39 curation: - Causal Clusters: replace the stale "Defer until ADR-011 is executed" contingency notes on Clusters A/C/D with resolved stamps (ADR-011 is executed, mapper deleted in PR #42); mark Cluster E executed and Cluster F partially resolved (the PriogridCountryMapper CIC is gone, the manager CIC stays). - C-03: narrow to the manager validation + enrich->validate path. Drop the deleted tests/test_mapping.py from Location and the gone mapper find_* methods from the trigger; the two live manager-side gaps remain. Bookkeeping only. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…-diff catch review-diff on the curation found the Cluster C resolved-stamp falsely claimed C-10 "was already resolved". C-10 (manager↔mapper coupling via the module-level _DEFAULT_MAPPER global) is itself a dead Cluster C concern: mapping.py:3088-3122 and the unfao.py get_default_mapper() call are both deleted (C-39, PR #42). Move C-10 to Resolved and correct the note to "C-02 and C-10 are both resolved". Counts: open 25 -> 24, resolved 15 -> 16. Bookkeeping only. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…valence proof (Tier 2) The ADR-011 enrichment swap (runtime mapper -> GAUL lookup enricher) went live without the output-equivalence proof its own plan required (umbrella #20: Stage 0 baseline + Stage 2 diff, "zero unexplained differences"). That proof was never produced, and the old mapper + both diff scripts were since deleted (eba1df8, PR #42), so it is now unrecoverable. Accepting option A (a smoke-test delivery) verifies the path runs, not that it produces the same/correct values; _validate only checks the 9 columns are non-null, not correct, so a lookup/merge bug ships wrong-but-non-null geo metadata to FAO silently. Tier 2 (not 1: the lookup sources from datafactory's authoritative GAUL parquets, D-07); trigger is the go-global flip to land_gaul (64k cells). Header 42/22/20 -> 43/23/20. Cross-refs C-03, C-22, C-23, C-30/32/34, C-39, D-08. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

… build (#141) ADR-011 swapped runtime geopandas mapping for a precomputed lookup, and the plan's required old-vs-new value diff (#20 Stage 2 / #23) was signed off without being produced. The comparator was then deleted in PR #42, so the backward check is unrecoverable — and run-0 shipped 64,742 cells to FAO on 2026-07-27 with no gate anywhere asserting that a *present* GAUL value was the *right* one. The check that remains available runs forwards, against the producer. Run this session, zero mismatches on all four links: gid->lat/lon formula vs PRIO-GRID priogrid_cell.dbf, all 259,200 cells -> max abs error 0.00e+00 lookup values vs the 7 datafactory gaul_admin parquets -> 0 mismatches / 64,742 cells lookup gid set vs land_gaul_pgids.json -> exactly equal, all unique delivered run-0 sidecar bytes vs the lookup -> 0 mismatches, sha256 matches the manifest The leading hypothesis going in — a flipped or off-by-one coordinate formula putting wrong-but-non-null coordinates on every cell, invisible to every gate — was falsified. The formula is exact across the entire global grid. This commit turns that session result into a standing guarantee. 18 tests split by dependency: always-on checks on the committed artifact (key uniqueness, no nulls, no -1 sentinels, region count, coordinate formula against a committed 219-cell PRIO-GRID ground-truth sample) and skipif checks against the views-datafactory sibling (all 7 columns, full 259,200-cell formula sweep). build_gaul_lookup.py: the three bare `assert`s become explicit LookupBuildError raises (python -O silently strips asserts, writing an unvalidated artifact that looks identical), plus a new index-uniqueness guard. Rebuild verified value-identical to the committed artifact. Two entries I registered earlier today are corrected here on mutation evidence rather than left overstated: - C-59 tier 2 -> 3. A duplicate cannot reach the artifact on the production path: Index.intersection de-duplicates first. The guard is reachable only via --region all. Kept anyway — that de-duplication is accidental pandas behaviour, not a declared guard, and silently absorbing upstream duplication hides a producer defect. - C-61 exposure corrected. The primary protection against -1 is the completeness filter, which is plain code and survives -O; the strippable-assert argument never applied to the -1 case. The real defect was stating invariants in a form the interpreter can delete. C-43 scope stated explicitly and deliberately: this proves transcription fidelity, not assignment correctness. If views-datafactory's area-majority join puts a cell in the wrong country, every test here still passes. That is filed as views-datafactory#387 — it had been resolved here on 2026-06-24 as "tracked there" and was never actually opened there, leaving it untracked platform-wide for five weeks. Register: 61 concerns, 24 open. Cluster K added. ruff clean; 296 passed (5 known local pyarrow byte-parity failures, CI authoritative). Co-authored-by: Claude Fable 5 <noreply@anthropic.com>

Polichinel and others added 5 commits June 24, 2026 04:39

Polichinel merged commit 26e476a into development Jun 24, 2026
4 checks passed

Polichinel deleted the chore/remove-dead-geopandas-mapper branch June 24, 2026 03:15

This was referenced Jun 24, 2026

docs(register): add C-41 — vestigial Git LFS config breaks routine git ops #43

Merged

chore: retire vestigial Git LFS config (resolves C-41) #44

Merged

Add shapefile version and error count to upload metadata (C-15) #12

Closed

This was referenced Jul 31, 2026

[UMBRELLA] FAO global delivery: lookup-based enrichment (ADR-011) + global coverage #20

Closed

test(gaul): C-43 forward-check as a standing guard + fail-loud lookup build #141

Merged

Navigation Menu

Delete the dead geopandas runtime mapper (C-39) - #42

Delete the dead geopandas runtime mapper (C-39)#42
Polichinel merged 5 commits into
developmentfrom
chore/remove-dead-geopandas-mapper

Polichinel commented Jun 24, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

Polichinel commented Jun 24, 2026