FA-11856 / Search retrieval semantics / Open access

Required terms in different fields produce a same-field hit · case 01

Required terms in different fields produce a same-field hit.

Verified by executionVariant 1 · 7 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

Flattened document fields erase the field boundary required by the query.

VERIFIED REPAIR

Test the entire query against each field independently.

Unsuccessful approach: Restricting to the first field drops legitimate matches in later fields.

Case contract

Return whether any single field contains all query terms. Empty query matches any existing field; no fields never matches.

Why this case matters

An offline deterministic retrieval model isolates this search contract from tokenization, storage, and network behavior.

1 / The failure

Exit 1

"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(fields, query):
    return set(query) <= set(t for field in fields for t in field)
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
a='a'+str(N)
check('cross field leakage', solve([[a],['b']],[a,'b']),False)
check('later field complete', solve([[],[a,'b']],[a,'b']),True)
check('first field complete', solve([[a,'b'],[]],[a,'b']),True)
check('no fields', solve([],[]),False)
check('empty query existing field', solve([[]],[]),True)
check('repeated query term', solve([[a]],[a,a]),True)
check('absent term', solve([[a]],['z']),False)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
cross field leakage	`True`	`False`	Failed
later field complete	`True`	`True`	Passed
first field complete	`True`	`True`	Passed
no fields	`True`	`False`	Failed
empty query existing field	`True`	`True`	Passed
repeated query term	`True`	`True`	Passed
absent term	`False`	`False`	Passed

SHA-256 / 351912e278d151c43c278b38c43b54f1974fd5a0237faab6fb20ae7c7b445b44

2 / The unsuccessful fix

Exit 1

"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(fields, query):
    return bool(fields) and set(query) <= set(fields[0])
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
a='a'+str(N)
check('cross field leakage', solve([[a],['b']],[a,'b']),False)
check('later field complete', solve([[],[a,'b']],[a,'b']),True)
check('first field complete', solve([[a,'b'],[]],[a,'b']),True)
check('no fields', solve([],[]),False)
check('empty query existing field', solve([[]],[]),True)
check('repeated query term', solve([[a]],[a,a]),True)
check('absent term', solve([[a]],['z']),False)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
cross field leakage	`False`	`False`	Passed
later field complete	`False`	`True`	Failed
first field complete	`True`	`True`	Passed
no fields	`False`	`False`	Passed
empty query existing field	`True`	`True`	Passed
repeated query term	`True`	`True`	Passed
absent term	`False`	`False`	Passed

SHA-256 / 4cf09ebcb033a4cabf22ed60b695dfff4bd03d4d8b01ec75a48e659a0c19d2b1

3 / The verified repair

Exit 0

"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(fields, query):
    return any(set(query) <= set(field) for field in fields)
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
a='a'+str(N)
check('cross field leakage', solve([[a],['b']],[a,'b']),False)
check('later field complete', solve([[],[a,'b']],[a,'b']),True)
check('first field complete', solve([[a,'b'],[]],[a,'b']),True)
check('no fields', solve([],[]),False)
check('empty query existing field', solve([[]],[]),True)
check('repeated query term', solve([[a]],[a,a]),True)
check('absent term', solve([[a]],['z']),False)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
cross field leakage	`False`	`False`	Passed
later field complete	`True`	`True`	Passed
first field complete	`True`	`True`	Passed
no fields	`False`	`False`	Passed
empty query existing field	`True`	`True`	Passed
repeated query term	`True`	`True`	Passed
absent term	`False`	`False`	Passed

SHA-256 / 340bcd8c16520f7c60ffbd873624c923ba26f3fb1b44fdc682fbc482f4fbf112

Verification & scope

Inputs are already tokenized or scored; this model makes no claim about production engine performance or linguistic analysis. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.

Observations recorded using Python 3.12.14 at 2026-09-29T14:38:51.674344+00:00.

Case digest / 3635220e85a9996c855d072c3289cd2e2e9308379187027e61b3df0439233f44