feat(address): AddressGenerator(dataset='europe') region-group support#195
Merged
Conversation
dataset= can now be a region-group alias (REGION_GROUPS in the new region_groups.py, starting with "europe") that draws a concrete country per address instead of one fixed country. The naive approach - resolve the country once in AddressGenerator.__init__ - is wrong: a single AddressGenerator instance is reused across every row of a run (BaseDomainService.generate() builds a fresh Address per row but keeps the same underlying generator, confirmed by tracing generate_worker.py). That would give every row in the run the same country, not variety. Instead: AddressGenerator.resolve_row() draws a fresh country per call for a region group (unchanged, cached single row for a concrete dataset - zero behavior change on the common path), lazily building and caching city/ country/phone/street sub-generators per country encountered so a repeated draw reuses its already-loaded data instead of reloading it. Address resolves its row once in __init__ and reads every property through it, so city/ phone/street/country_code all agree on the same drawn country. Verified the naive single-resolve bug is exactly what the new discriminating test (test_europe_draws_vary_per_row_and_stay_in_the_group) catches - patched resolve_row() to cache one row instead of per-call and confirmed it fails. Updated a hand-rolled AddressGenerator test double (_SentinelAddressGenerator in test_person_gender_alignment.py) to the new resolve_row() interface. Full suite 1432 passed/15 skipped, mypy/ruff clean.
|
ake2l
added a commit
that referenced
this pull request
Jul 9, 2026
…memstore aggregation Three small Benerator-migration fixes surfaced by converting a real enterprise project, each with DSL happy/edge/error coverage. 1. **CSV headers are whitespace-trimmed on read** (`FileUtil.read_csv_to_dict_list` - the single reader all CSV paths funnel through). Padded/aligned CSVs (Benerator entity-CSV style: `ean_code ,name ,...`) used to produce keys like `"name "` that script field access (`this.name`) could not resolve. Only the KEYS are trimmed - cell VALUES keep their whitespace untouched (trimming values would corrupt data; only the keys are structural). 2. **Sub-region dataset groups** for `AddressGenerator`: `western_europe`, `central_europe`, `southern_europe`, `eastern_europe`, `north_america`, `oceania`, `french`, `iberia` - pure data additions over the existing region-group mechanism (#195), no engine change. Every listed code verified to have city/street data files, enforced by a new data gate test (a group member without data would otherwise only fail when the seeded draw happens to pick it - a flaky runtime crash). 3. **`Memstore.sumEntityColumn` skips non-numeric cells** (Benerator-lenient aggregation): a CSV-sourced column carrying a stray placeholder (`'n/a'`, empty) no longer aborts the whole sum. Leniency is strictly per-CELL - a wrong column NAME still raises `KeyError` instead of silently summing to 0. ## Test plan - [x] Region groups (DSL): `iberia` draws vary per row, stay inside the pool, and city/country stay consistent (happy); uppercase `NORTH_AMERICA` resolves like lowercase (edge); an unknown dataset fails loudly (error). Data gate: every code in every group has city/street CSVs. - [x] Header trim (DSL, end to end): `this.name` resolves on a padded header; record keys are clean; padded cell values stay untouched (contract boundary). Unit test kept. - [x] Lenient sum: DSL end-to-end (`5, n/a, 7` sums to 12) + unit edge cases (`None`/`'n/a'` skipped) + missing-column-stays-fatal regression. - [x] All three behavior tests verified discriminating (fixes reverted locally → exactly those tests fail). - [x] Full suite 1464 passed / 15 skipped, mypy (424 files) and ruff clean.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Test plan