FA-99671 / Chunk citation alignment / Open access

Overlapping chunk citations include the already stitched prefix · case 01

Overlapping chunk citations include the already stitched prefix.

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

ROOT CAUSE

Chunk-local ranges are copied into output coordinates without removing source text covered by an earlier chunk.

THE FAILURE

Chunk-local ranges are copied into output coordinates without removing source text covered by an earlier chunk.

Unsuccessful approach: Clipping away the duplicated prefix still leaves citation coordinates in document space instead of output-local space.

Case contract

Chunks are source intervals [start,end] in ascending order, each next interval overlaps or touches the prior covered frontier; stitching retains the first chunk in full and discards duplicate prefixes. Citation [chunk_index,local_start,local_end] maps to the nonempty retained piece in output-local coordinates, or None if fully overlapped.

Why this case matters

A deterministic citation bookkeeping model isolates source attribution from rendering and retrieval.

1 / The failure

Exit 1

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

N = 1
observations = []
def solve(chunks, citation):
    i,lo,hi=citation
    return [lo,hi]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
chunks=[[N,N+5],[N+3,N+8]]
check('citation straddles duplicated prefix',solve(chunks,[1,1,4]),[5,7])
check('entire citation already stitched',solve(chunks,[1,0,2]),None)
check('citation starts after overlap',solve(chunks,[1,2,5]),[5,8])
check('first chunk unchanged',solve(chunks,[0,1,3]),[1,3])
check('touching chunk boundary',solve([[N,N+2],[N+2,N+4]],[1,0,2]),[2,4])
check('empty citation',solve(chunks,[1,2,2]),None)
check('no chunks',solve([],[0,0,1]),None)
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
citation straddles duplicated prefix	`[1, 4]`	`[5, 7]`	Failed
entire citation already stitched	`[0, 2]`	`None`	Failed
citation starts after overlap	`[2, 5]`	`[5, 8]`	Failed
first chunk unchanged	`[1, 3]`	`[1, 3]`	Passed
touching chunk boundary	`[0, 2]`	`[2, 4]`	Failed
empty citation	`[2, 2]`	`None`	Failed
no chunks	`[0, 1]`	`None`	Failed

SHA-256 / abf21de34a2de3f527d506233a6bf2788f93ce0d4bef4b3d539782c772717716

2 / The unsuccessful fix

Exit 1

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

N = 1
observations = []
def solve(chunks, citation):
    if not chunks: return None
    i,lo,hi=citation
    a,b=chunks[i]
    frontier=max((end for start,end in chunks[:i]),default=a)
    left=max(a+lo,frontier)
    right=a+hi
    return [left,right] if left<right else None
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
chunks=[[N,N+5],[N+3,N+8]]
check('citation straddles duplicated prefix',solve(chunks,[1,1,4]),[5,7])
check('entire citation already stitched',solve(chunks,[1,0,2]),None)
check('citation starts after overlap',solve(chunks,[1,2,5]),[5,8])
check('first chunk unchanged',solve(chunks,[0,1,3]),[1,3])
check('touching chunk boundary',solve([[N,N+2],[N+2,N+4]],[1,0,2]),[2,4])
check('empty citation',solve(chunks,[1,2,2]),None)
check('no chunks',solve([],[0,0,1]),None)
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
citation straddles duplicated prefix	`[6, 8]`	`[5, 7]`	Failed
entire citation already stitched	`None`	`None`	Passed
citation starts after overlap	`[6, 9]`	`[5, 8]`	Failed
first chunk unchanged	`[2, 4]`	`[1, 3]`	Failed
touching chunk boundary	`[3, 5]`	`[2, 4]`	Failed
empty citation	`None`	`None`	Passed
no chunks	`None`	`None`	Passed

SHA-256 / 1a12d9bcd2dc41b36fe16be094a5daacfef1ec7ff34ab5ee216e4dfacf1ddf79

HELD IN THE MEMBER ARCHIVE

The verified repair and its recorded checks are member-only.

This mechanism has 7 recorded checks per implementation. The open-access tier publishes the failure and the unsuccessful fix; the repaired source that passes every check, and the observations that prove it, are available to members.

Every case sharing this mechanism uses the same contract and the same repair, so this one record is held back for all of them.

Member access is invitation-based. Sign in with your invited account to inspect the repair.

Verification & scope

Inputs are small integer offsets and identifiers under the stated toy contract; this does not define a document format or production citation API. 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 2026-09-30T16:48:26.712466+00:00.

Case digest / 3b1fb4101529e0c734b61893c52a084bad160f2bd26a6b57ba306d14364a9c69