cryptorugmunch
8d25702eca
Squashed from chore/license-relicense. Full message preserved in the original branch commitbb77eb5. See ADR-0002 for the decision rationale. Refs: ADR-0002, commitbb77eb5
86 lines
2.6 KiB
Python
86 lines
2.6 KiB
Python
# SPDX-License-Identifier: MIT
|
|
# Copyright (c) 2026 Rug Munch Media LLC
|
|
#
|
|
# Part of Pry — https://git.rugmunch.io/RugMunchMedia/pryscraper
|
|
# Licensed under MIT. See LICENSE.
|
|
"""Tests for entity reconciliation."""
|
|
|
|
from reconciliation import (
|
|
VERTICAL_SCHEMAS,
|
|
_build_field_map,
|
|
_normalize_record,
|
|
compute_similarity,
|
|
match_entities,
|
|
)
|
|
|
|
|
|
def test_compute_similarity_exact() -> None:
|
|
assert compute_similarity("iPhone 15", "iPhone 15") == 1.0
|
|
|
|
|
|
def test_compute_similarity_partial() -> None:
|
|
sim = compute_similarity("iPhone 15 Pro Max", "iPhone 15 Pro")
|
|
assert sim > 0.5
|
|
|
|
|
|
def test_compute_similarity_different() -> None:
|
|
sim = compute_similarity("iPhone 15", "Samsung Galaxy S24")
|
|
assert sim < 0.5
|
|
|
|
|
|
def test_compute_similarity_empty() -> None:
|
|
assert compute_similarity("", "test") == 0.0
|
|
assert compute_similarity("test", "") == 0.0
|
|
|
|
|
|
def test_match_entities_products() -> None:
|
|
records = [
|
|
{
|
|
"name": "iPhone 15 Pro",
|
|
"price": 999,
|
|
"url": "https://amazon.com/iphone",
|
|
"_source": "amazon",
|
|
},
|
|
{
|
|
"name": "iPhone 15 Pro",
|
|
"price": 999.99,
|
|
"url": "https://walmart.com/iphone",
|
|
"_source": "walmart",
|
|
},
|
|
{"name": "Samsung TV", "price": 499, "url": "https://bestbuy.com/tv", "_source": "bestbuy"},
|
|
]
|
|
entities = match_entities(records, "product", threshold=0.7)
|
|
assert len(entities) >= 2
|
|
iphone_entity = [e for e in entities if "iphone" in str(e).lower()]
|
|
assert len(iphone_entity) > 0
|
|
assert iphone_entity[0]["record_count"] >= 2
|
|
|
|
|
|
def test_build_field_map() -> None:
|
|
schema = VERTICAL_SCHEMAS["product"]
|
|
field_map = _build_field_map(schema)
|
|
assert field_map["name"] == "name"
|
|
assert field_map["title"] == "name"
|
|
assert field_map["cost"] == "price"
|
|
|
|
|
|
def test_normalize_record() -> None:
|
|
schema = VERTICAL_SCHEMAS["product"]
|
|
field_map = _build_field_map(schema)
|
|
record = {"title": "iPhone", "cost": "$999.99", "link": "https://example.com"}
|
|
normalized = _normalize_record(record, field_map, schema)
|
|
assert normalized.get("name") == "iPhone"
|
|
assert normalized.get("price") == 999.99
|
|
assert normalized.get("url") == "https://example.com"
|
|
|
|
|
|
def test_list_schemas_keys() -> None:
|
|
assert "product" in VERTICAL_SCHEMAS
|
|
assert "job" in VERTICAL_SCHEMAS
|
|
assert "real_estate" in VERTICAL_SCHEMAS
|
|
assert "review" in VERTICAL_SCHEMAS
|
|
|
|
|
|
def test_identity_fields_present() -> None:
|
|
for _key, schema in VERTICAL_SCHEMAS.items():
|
|
assert len(schema["identity_fields"]) > 0
|