FA-80406 / Bidirectional text layout / Open access

Weak type resolution: leftover separator neutralisation · case 01

Unattached currency signs keep a weak type and block neutral resolution.

Verified by executionVariant 1 · 8 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

W6 omits ET from the set converted to ON.

VERIFIED REPAIR

Convert all remaining ES, ET and CS to ON.

Unsuccessful approach: Omitting ES instead leaves stray plus signs weak.

Case contract

Input [classes of one isolating run sequence, sos]. W1 NSM takes the previous type (sos at start; ON after LRI/RLI/FSI/PDI). W2 EN after the last strong AL becomes AN. W3 AL->R. W4 a single ES between EN and EN -> EN; a single CS between two numbers of the same type takes that type. W5 a run of ET adjacent to EN -> EN. W6 remaining ES/ET/CS -> ON. W7 EN whose last strong (sos at start) is L -> L. Return resolved classes.

Why this case matters

Mixed right-to-left and left-to-right text must resolve levels and visual order exactly, or words, numbers and carets land in the wrong place.

1 / The failure

Exit 1

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

N = 1
observations = []
def solve(x):
    types, sos = x
    t = list(types)
    n = len(t)
    prev = sos
    for i in range(n):
        if t[i] == 'NSM':
            t[i] = 'ON' if prev in ('LRI', 'RLI', 'FSI', 'PDI') else prev
        prev = t[i]
    last_strong = sos
    for i in range(n):
        if t[i] in ('L', 'R', 'AL'):
            last_strong = t[i]
        elif t[i] == 'EN' and last_strong == 'AL':
            t[i] = 'AN'
    t = ['R' if c == 'AL' else c for c in t]
    for i in range(1, n - 1):
        if t[i] == 'ES' and t[i - 1] == 'EN' and t[i + 1] == 'EN':
            t[i] = 'EN'
        elif t[i] == 'CS' and t[i - 1] == t[i + 1] and t[i - 1] in ('EN', 'AN'):
            t[i] = t[i - 1]
    i = 0
    while i < n:
        if t[i] == 'ET':
            j = i
            while j < n and t[j] == 'ET':
                j += 1
            if (i > 0 and t[i - 1] == 'EN') or (j < n and t[j] == 'EN'):
                for k in range(i, j):
                    t[k] = 'EN'
            i = j
        else:
            i += 1
    t = ['ON' if c in ('ES', 'CS') else c for c in t]
    strong = sos
    for i in range(n):
        if t[i] in ('L', 'R'):
            strong = t[i]
        elif t[i] == 'EN' and strong == 'L':
            t[i] = 'L'
    return t
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[('regression: leftover separator neutralisation', [['AL', 'PDI', 'ET', 'CS', 'ET'], 'R'], ['R', 'PDI', 'ON', 'ON', 'ON']), ('regression: leftover separator neutralisation', [['R', 'ET', 'NSM', 'NSM', 'LRI', 'CS', 'AL', 'R', 'ON', 'AL', 'ES'], 'L'], ['R', 'ON', 'ON', 'ON', 'LRI', 'ON', 'R', 'R', 'ON', 'R', 'ON']), ('partial-repair probe', [['AL', 'LRI', 'WS', 'ES', 'L', 'CS', 'LRI', 'AN', 'NSM'], 'R'], ['R', 'LRI', 'WS', 'ON', 'L', 'ON', 'LRI', 'AN', 'AN']), ('partial-repair probe', [['ES', 'L', 'CS', 'AL', 'NSM', 'L', 'ES'], 'R'], ['ON', 'L', 'ON', 'R', 'R', 'L', 'ON']), ('percent after number', [['EN', 'ET', 'ET', 'L'], 'R'], ['EN', 'EN', 'EN', 'L']), ('NSM chain after R', [['R', 'NSM', 'NSM', 'EN'], 'L'], ['R', 'R', 'R', 'EN']), ('control layout', [['AN', 'AN', 'NSM'], 'L'], ['AN', 'AN', 'AN']), ('control layout', [['NSM', 'R', 'EN', 'ON'], 'L'], ['L', 'R', 'EN', 'ON'])], [('regression: leftover separator neutralisation', [['NSM', 'PDI', 'L', 'R', 'WS', 'ET', 'CS', 'AL'], 'L'], ['L', 'PDI', 'L', 'R', 'WS', 'ON', 'ON', 'R']), ('regression: leftover separator neutralisation', [['EN', 'R', 'ET', 'ET', 'NSM'], 'R'], ['EN', 'R', 'ON', 'ON', 'ON']), ('partial-repair probe', [['EN', 'WS', 'ES', 'ES', 'L', 'LRI'], 'R'], ['EN', 'WS', 'ON', 'ON', 'L', 'LRI']), ('partial-repair probe', [['NSM', 'WS', 'CS', 'ES'], 'R'], ['R', 'WS', 'ON', 'ON']), ('Arabic number context', [['AL', 'EN', 'CS', 'EN'], 'L'], ['R', 'AN', 'AN', 'AN']), ('plus between numbers', [['EN', 'ES', 'EN', 'R'], 'R'], ['EN', 'EN', 'EN', 'R']), ('control layout', [['LRI', 'EN', 'R'], 'L'], ['LRI', 'L', 'R']), ('control layout', [['LRI', 'AL', 'WS', 'EN', 'EN', 'EN', 'LRI', 'EN', 'PDI', 'EN', 'R'], 'R'], ['LRI', 'R', 'WS', 'AN', 'AN', 'AN', 'LRI', 'AN', 'PDI', 'AN', 'R'])], [('regression: leftover separator neutralisation', [['PDI', 'AL', 'PDI', 'AN', 'ON', 'ET', 'EN', 'ES'], 'R'], ['PDI', 'R', 'PDI', 'AN', 'ON', 'ON', 'AN', 'ON']), ('regression: leftover separator neutralisation', [['ES', 'PDI', 'CS', 'AL', 'ET', 'L', 'CS'], 'R'], ['ON', 'PDI', 'ON', 'R', 'ON', 'L', 'ON']), ('regression: leftover separator neutralisation', [['ES', 'R', 'AL', 'CS', 'R', 'ET', 'AN', 'CS', 'R', 'EN', 'AL'], 'R'], ['ON', 'R', 'R', 'ON', 'R', 'ON', 'AN', 'ON', 'R', 'EN', 'R']), ('regression: leftover separator neutralisation', [['NSM', 'ET', 'ES', 'CS', 'EN'], 'R'], ['R', 'ON', 'ON', 'ON', 'EN']), ('NSM after isolate initiator', [['LRI', 'NSM', 'L', 'PDI'], 'R'], ['LRI', 'ON', 'L', 'PDI']), ('Arabic number context', [['AL', 'EN', 'CS', 'EN'], 'L'], ['R', 'AN', 'AN', 'AN']), ('control layout', [['AN', 'ON', 'CS', 'L'], 'L'], ['AN', 'ON', 'ON', 'L']), ('control layout', [['ET', 'EN'], 'L'], ['L', 'L'])], [('regression: leftover separator neutralisation', [['AN', 'PDI', 'EN', 'CS', 'R', 'NSM', 'AN', 'ET'], 'R'], ['AN', 'PDI', 'EN', 'ON', 'R', 'R', 'AN', 'ON']), ('regression: leftover separator neutralisation', [['ON', 'AN', 'LRI', 'ET'], 'R'], ['ON', 'AN', 'LRI', 'ON']), ('regression: leftover separator neutralisation', [['ES', 'ET', 'R', 'R', 'ET', 'ET', 'ON', 'LRI'], 'L'], ['ON', 'ON', 'R', 'R', 'ON', 'ON', 'ON', 'LRI']), ('partial-repair probe', [['ES', 'EN', 'PDI', 'PDI', 'EN', 'L', 'ES'], 'R'], ['ON', 'EN', 'PDI', 'PDI', 'EN', 'L', 'ON']), ('percent after number', [['EN', 'ET', 'ET', 'L'], 'R'], ['EN', 'EN', 'EN', 'L']), ('NSM after isolate initiator', [['LRI', 'NSM', 'L', 'PDI'], 'R'], ['LRI', 'ON', 'L', 'PDI']), ('control layout', [['LRI', 'AN', 'L', 'AL', 'ON'], 'R'], ['LRI', 'AN', 'L', 'R', 'ON']), ('control layout', [['EN', 'EN'], 'L'], ['L', 'L'])], [('regression: leftover separator neutralisation', [['AL', 'ET', 'LRI', 'ES', 'NSM', 'R', 'LRI', 'ES', 'R', 'EN'], 'L'], ['R', 'ON', 'LRI', 'ON', 'ON', 'R', 'LRI', 'ON', 'R', 'EN']), ('regression: leftover separator neutralisation', [['ET', 'ON', 'EN', 'AL', 'EN', 'R', 'ON', 'AN', 'EN'], 'R'], ['ON', 'ON', 'EN', 'R', 'AN', 'R', 'ON', 'AN', 'EN']), ('partial-repair probe', [['EN', 'NSM', 'ES', 'NSM', 'EN', 'LRI', 'NSM', 'WS', 'L'], 'R'], ['EN', 'EN', 'ON', 'ON', 'EN', 'LRI', 'ON', 'WS', 'L']), ('partial-repair probe', [['ES', 'EN'], 'L'], ['ON', 'L']), ('NSM chain after R', [['R', 'NSM', 'NSM', 'EN'], 'L'], ['R', 'R', 'R', 'EN']), ('Arabic number context', [['AL', 'EN', 'CS', 'EN'], 'L'], ['R', 'AN', 'AN', 'AN']), ('control layout', [['L', 'AL', 'PDI', 'LRI', 'AL', 'EN', 'CS', 'EN', 'WS', 'CS'], 'R'], ['L', 'R', 'PDI', 'LRI', 'R', 'AN', 'AN', 'AN', 'WS', 'ON']), ('control layout', [['AL', 'CS', 'PDI'], 'L'], ['R', 'ON', 'PDI'])]]
for label, args, expected in fixtures[N - 1]:
    check(label, solve(args), expected)
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
regression: leftover separator neutralisation	`['R', 'PDI', 'ET', 'ON', 'ET']`	`['R', 'PDI', 'ON', 'ON', 'ON']`	Failed
regression: leftover separator neutralisation	`['R', 'ET', 'ET', 'ET', 'LRI', 'ON', 'R', 'R', 'ON', 'R', 'ON']`	`['R', 'ON', 'ON', 'ON', 'LRI', 'ON', 'R', 'R', 'ON', 'R', 'ON']`	Failed
partial-repair probe	`['R', 'LRI', 'WS', 'ON', 'L', 'ON', 'LRI', 'AN', 'AN']`	`['R', 'LRI', 'WS', 'ON', 'L', 'ON', 'LRI', 'AN', 'AN']`	Passed
partial-repair probe	`['ON', 'L', 'ON', 'R', 'R', 'L', 'ON']`	`['ON', 'L', 'ON', 'R', 'R', 'L', 'ON']`	Passed
percent after number	`['EN', 'EN', 'EN', 'L']`	`['EN', 'EN', 'EN', 'L']`	Passed
NSM chain after R	`['R', 'R', 'R', 'EN']`	`['R', 'R', 'R', 'EN']`	Passed
control layout	`['AN', 'AN', 'AN']`	`['AN', 'AN', 'AN']`	Passed
control layout	`['L', 'R', 'EN', 'ON']`	`['L', 'R', 'EN', 'ON']`	Passed

SHA-256 / 7b9aa2d40127769ecc3ef39c2eb410552bb5b65ec5678e3e0c008f7409fdf99d

2 / The unsuccessful fix

Exit 1

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

N = 1
observations = []
def solve(x):
    types, sos = x
    t = list(types)
    n = len(t)
    prev = sos
    for i in range(n):
        if t[i] == 'NSM':
            t[i] = 'ON' if prev in ('LRI', 'RLI', 'FSI', 'PDI') else prev
        prev = t[i]
    last_strong = sos
    for i in range(n):
        if t[i] in ('L', 'R', 'AL'):
            last_strong = t[i]
        elif t[i] == 'EN' and last_strong == 'AL':
            t[i] = 'AN'
    t = ['R' if c == 'AL' else c for c in t]
    for i in range(1, n - 1):
        if t[i] == 'ES' and t[i - 1] == 'EN' and t[i + 1] == 'EN':
            t[i] = 'EN'
        elif t[i] == 'CS' and t[i - 1] == t[i + 1] and t[i - 1] in ('EN', 'AN'):
            t[i] = t[i - 1]
    i = 0
    while i < n:
        if t[i] == 'ET':
            j = i
            while j < n and t[j] == 'ET':
                j += 1
            if (i > 0 and t[i - 1] == 'EN') or (j < n and t[j] == 'EN'):
                for k in range(i, j):
                    t[k] = 'EN'
            i = j
        else:
            i += 1
    t = ['ON' if c in ('ET', 'CS') else c for c in t]
    strong = sos
    for i in range(n):
        if t[i] in ('L', 'R'):
            strong = t[i]
        elif t[i] == 'EN' and strong == 'L':
            t[i] = 'L'
    return t
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[('regression: leftover separator neutralisation', [['AL', 'PDI', 'ET', 'CS', 'ET'], 'R'], ['R', 'PDI', 'ON', 'ON', 'ON']), ('regression: leftover separator neutralisation', [['R', 'ET', 'NSM', 'NSM', 'LRI', 'CS', 'AL', 'R', 'ON', 'AL', 'ES'], 'L'], ['R', 'ON', 'ON', 'ON', 'LRI', 'ON', 'R', 'R', 'ON', 'R', 'ON']), ('partial-repair probe', [['AL', 'LRI', 'WS', 'ES', 'L', 'CS', 'LRI', 'AN', 'NSM'], 'R'], ['R', 'LRI', 'WS', 'ON', 'L', 'ON', 'LRI', 'AN', 'AN']), ('partial-repair probe', [['ES', 'L', 'CS', 'AL', 'NSM', 'L', 'ES'], 'R'], ['ON', 'L', 'ON', 'R', 'R', 'L', 'ON']), ('percent after number', [['EN', 'ET', 'ET', 'L'], 'R'], ['EN', 'EN', 'EN', 'L']), ('NSM chain after R', [['R', 'NSM', 'NSM', 'EN'], 'L'], ['R', 'R', 'R', 'EN']), ('control layout', [['AN', 'AN', 'NSM'], 'L'], ['AN', 'AN', 'AN']), ('control layout', [['NSM', 'R', 'EN', 'ON'], 'L'], ['L', 'R', 'EN', 'ON'])], [('regression: leftover separator neutralisation', [['NSM', 'PDI', 'L', 'R', 'WS', 'ET', 'CS', 'AL'], 'L'], ['L', 'PDI', 'L', 'R', 'WS', 'ON', 'ON', 'R']), ('regression: leftover separator neutralisation', [['EN', 'R', 'ET', 'ET', 'NSM'], 'R'], ['EN', 'R', 'ON', 'ON', 'ON']), ('partial-repair probe', [['EN', 'WS', 'ES', 'ES', 'L', 'LRI'], 'R'], ['EN', 'WS', 'ON', 'ON', 'L', 'LRI']), ('partial-repair probe', [['NSM', 'WS', 'CS', 'ES'], 'R'], ['R', 'WS', 'ON', 'ON']), ('Arabic number context', [['AL', 'EN', 'CS', 'EN'], 'L'], ['R', 'AN', 'AN', 'AN']), ('plus between numbers', [['EN', 'ES', 'EN', 'R'], 'R'], ['EN', 'EN', 'EN', 'R']), ('control layout', [['LRI', 'EN', 'R'], 'L'], ['LRI', 'L', 'R']), ('control layout', [['LRI', 'AL', 'WS', 'EN', 'EN', 'EN', 'LRI', 'EN', 'PDI', 'EN', 'R'], 'R'], ['LRI', 'R', 'WS', 'AN', 'AN', 'AN', 'LRI', 'AN', 'PDI', 'AN', 'R'])], [('regression: leftover separator neutralisation', [['PDI', 'AL', 'PDI', 'AN', 'ON', 'ET', 'EN', 'ES'], 'R'], ['PDI', 'R', 'PDI', 'AN', 'ON', 'ON', 'AN', 'ON']), ('regression: leftover separator neutralisation', [['ES', 'PDI', 'CS', 'AL', 'ET', 'L', 'CS'], 'R'], ['ON', 'PDI', 'ON', 'R', 'ON', 'L', 'ON']), ('regression: leftover separator neutralisation', [['ES', 'R', 'AL', 'CS', 'R', 'ET', 'AN', 'CS', 'R', 'EN', 'AL'], 'R'], ['ON', 'R', 'R', 'ON', 'R', 'ON', 'AN', 'ON', 'R', 'EN', 'R']), ('regression: leftover separator neutralisation', [['NSM', 'ET', 'ES', 'CS', 'EN'], 'R'], ['R', 'ON', 'ON', 'ON', 'EN']), ('NSM after isolate initiator', [['LRI', 'NSM', 'L', 'PDI'], 'R'], ['LRI', 'ON', 'L', 'PDI']), ('Arabic number context', [['AL', 'EN', 'CS', 'EN'], 'L'], ['R', 'AN', 'AN', 'AN']), ('control layout', [['AN', 'ON', 'CS', 'L'], 'L'], ['AN', 'ON', 'ON', 'L']), ('control layout', [['ET', 'EN'], 'L'], ['L', 'L'])], [('regression: leftover separator neutralisation', [['AN', 'PDI', 'EN', 'CS', 'R', 'NSM', 'AN', 'ET'], 'R'], ['AN', 'PDI', 'EN', 'ON', 'R', 'R', 'AN', 'ON']), ('regression: leftover separator neutralisation', [['ON', 'AN', 'LRI', 'ET'], 'R'], ['ON', 'AN', 'LRI', 'ON']), ('regression: leftover separator neutralisation', [['ES', 'ET', 'R', 'R', 'ET', 'ET', 'ON', 'LRI'], 'L'], ['ON', 'ON', 'R', 'R', 'ON', 'ON', 'ON', 'LRI']), ('partial-repair probe', [['ES', 'EN', 'PDI', 'PDI', 'EN', 'L', 'ES'], 'R'], ['ON', 'EN', 'PDI', 'PDI', 'EN', 'L', 'ON']), ('percent after number', [['EN', 'ET', 'ET', 'L'], 'R'], ['EN', 'EN', 'EN', 'L']), ('NSM after isolate initiator', [['LRI', 'NSM', 'L', 'PDI'], 'R'], ['LRI', 'ON', 'L', 'PDI']), ('control layout', [['LRI', 'AN', 'L', 'AL', 'ON'], 'R'], ['LRI', 'AN', 'L', 'R', 'ON']), ('control layout', [['EN', 'EN'], 'L'], ['L', 'L'])], [('regression: leftover separator neutralisation', [['AL', 'ET', 'LRI', 'ES', 'NSM', 'R', 'LRI', 'ES', 'R', 'EN'], 'L'], ['R', 'ON', 'LRI', 'ON', 'ON', 'R', 'LRI', 'ON', 'R', 'EN']), ('regression: leftover separator neutralisation', [['ET', 'ON', 'EN', 'AL', 'EN', 'R', 'ON', 'AN', 'EN'], 'R'], ['ON', 'ON', 'EN', 'R', 'AN', 'R', 'ON', 'AN', 'EN']), ('partial-repair probe', [['EN', 'NSM', 'ES', 'NSM', 'EN', 'LRI', 'NSM', 'WS', 'L'], 'R'], ['EN', 'EN', 'ON', 'ON', 'EN', 'LRI', 'ON', 'WS', 'L']), ('partial-repair probe', [['ES', 'EN'], 'L'], ['ON', 'L']), ('NSM chain after R', [['R', 'NSM', 'NSM', 'EN'], 'L'], ['R', 'R', 'R', 'EN']), ('Arabic number context', [['AL', 'EN', 'CS', 'EN'], 'L'], ['R', 'AN', 'AN', 'AN']), ('control layout', [['L', 'AL', 'PDI', 'LRI', 'AL', 'EN', 'CS', 'EN', 'WS', 'CS'], 'R'], ['L', 'R', 'PDI', 'LRI', 'R', 'AN', 'AN', 'AN', 'WS', 'ON']), ('control layout', [['AL', 'CS', 'PDI'], 'L'], ['R', 'ON', 'PDI'])]]
for label, args, expected in fixtures[N - 1]:
    check(label, solve(args), expected)
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
regression: leftover separator neutralisation	`['R', 'PDI', 'ON', 'ON', 'ON']`	`['R', 'PDI', 'ON', 'ON', 'ON']`	Passed
regression: leftover separator neutralisation	`['R', 'ON', 'ON', 'ON', 'LRI', 'ON', 'R', 'R', 'ON', 'R', 'ES']`	`['R', 'ON', 'ON', 'ON', 'LRI', 'ON', 'R', 'R', 'ON', 'R', 'ON']`	Failed
partial-repair probe	`['R', 'LRI', 'WS', 'ES', 'L', 'ON', 'LRI', 'AN', 'AN']`	`['R', 'LRI', 'WS', 'ON', 'L', 'ON', 'LRI', 'AN', 'AN']`	Failed
partial-repair probe	`['ES', 'L', 'ON', 'R', 'R', 'L', 'ES']`	`['ON', 'L', 'ON', 'R', 'R', 'L', 'ON']`	Failed
percent after number	`['EN', 'EN', 'EN', 'L']`	`['EN', 'EN', 'EN', 'L']`	Passed
NSM chain after R	`['R', 'R', 'R', 'EN']`	`['R', 'R', 'R', 'EN']`	Passed
control layout	`['AN', 'AN', 'AN']`	`['AN', 'AN', 'AN']`	Passed
control layout	`['L', 'R', 'EN', 'ON']`	`['L', 'R', 'EN', 'ON']`	Passed

SHA-256 / 80dadf55f8d96622327273a23eb1e53a80fa36d21c6417d3d16f27349d1a351f

3 / The verified repair

Exit 0

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

N = 1
observations = []
def solve(x):
    types, sos = x
    t = list(types)
    n = len(t)
    prev = sos
    for i in range(n):
        if t[i] == 'NSM':
            t[i] = 'ON' if prev in ('LRI', 'RLI', 'FSI', 'PDI') else prev
        prev = t[i]
    last_strong = sos
    for i in range(n):
        if t[i] in ('L', 'R', 'AL'):
            last_strong = t[i]
        elif t[i] == 'EN' and last_strong == 'AL':
            t[i] = 'AN'
    t = ['R' if c == 'AL' else c for c in t]
    for i in range(1, n - 1):
        if t[i] == 'ES' and t[i - 1] == 'EN' and t[i + 1] == 'EN':
            t[i] = 'EN'
        elif t[i] == 'CS' and t[i - 1] == t[i + 1] and t[i - 1] in ('EN', 'AN'):
            t[i] = t[i - 1]
    i = 0
    while i < n:
        if t[i] == 'ET':
            j = i
            while j < n and t[j] == 'ET':
                j += 1
            if (i > 0 and t[i - 1] == 'EN') or (j < n and t[j] == 'EN'):
                for k in range(i, j):
                    t[k] = 'EN'
            i = j
        else:
            i += 1
    t = ['ON' if c in ('ES', 'ET', 'CS') else c for c in t]
    strong = sos
    for i in range(n):
        if t[i] in ('L', 'R'):
            strong = t[i]
        elif t[i] == 'EN' and strong == 'L':
            t[i] = 'L'
    return t
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[('regression: leftover separator neutralisation', [['AL', 'PDI', 'ET', 'CS', 'ET'], 'R'], ['R', 'PDI', 'ON', 'ON', 'ON']), ('regression: leftover separator neutralisation', [['R', 'ET', 'NSM', 'NSM', 'LRI', 'CS', 'AL', 'R', 'ON', 'AL', 'ES'], 'L'], ['R', 'ON', 'ON', 'ON', 'LRI', 'ON', 'R', 'R', 'ON', 'R', 'ON']), ('partial-repair probe', [['AL', 'LRI', 'WS', 'ES', 'L', 'CS', 'LRI', 'AN', 'NSM'], 'R'], ['R', 'LRI', 'WS', 'ON', 'L', 'ON', 'LRI', 'AN', 'AN']), ('partial-repair probe', [['ES', 'L', 'CS', 'AL', 'NSM', 'L', 'ES'], 'R'], ['ON', 'L', 'ON', 'R', 'R', 'L', 'ON']), ('percent after number', [['EN', 'ET', 'ET', 'L'], 'R'], ['EN', 'EN', 'EN', 'L']), ('NSM chain after R', [['R', 'NSM', 'NSM', 'EN'], 'L'], ['R', 'R', 'R', 'EN']), ('control layout', [['AN', 'AN', 'NSM'], 'L'], ['AN', 'AN', 'AN']), ('control layout', [['NSM', 'R', 'EN', 'ON'], 'L'], ['L', 'R', 'EN', 'ON'])], [('regression: leftover separator neutralisation', [['NSM', 'PDI', 'L', 'R', 'WS', 'ET', 'CS', 'AL'], 'L'], ['L', 'PDI', 'L', 'R', 'WS', 'ON', 'ON', 'R']), ('regression: leftover separator neutralisation', [['EN', 'R', 'ET', 'ET', 'NSM'], 'R'], ['EN', 'R', 'ON', 'ON', 'ON']), ('partial-repair probe', [['EN', 'WS', 'ES', 'ES', 'L', 'LRI'], 'R'], ['EN', 'WS', 'ON', 'ON', 'L', 'LRI']), ('partial-repair probe', [['NSM', 'WS', 'CS', 'ES'], 'R'], ['R', 'WS', 'ON', 'ON']), ('Arabic number context', [['AL', 'EN', 'CS', 'EN'], 'L'], ['R', 'AN', 'AN', 'AN']), ('plus between numbers', [['EN', 'ES', 'EN', 'R'], 'R'], ['EN', 'EN', 'EN', 'R']), ('control layout', [['LRI', 'EN', 'R'], 'L'], ['LRI', 'L', 'R']), ('control layout', [['LRI', 'AL', 'WS', 'EN', 'EN', 'EN', 'LRI', 'EN', 'PDI', 'EN', 'R'], 'R'], ['LRI', 'R', 'WS', 'AN', 'AN', 'AN', 'LRI', 'AN', 'PDI', 'AN', 'R'])], [('regression: leftover separator neutralisation', [['PDI', 'AL', 'PDI', 'AN', 'ON', 'ET', 'EN', 'ES'], 'R'], ['PDI', 'R', 'PDI', 'AN', 'ON', 'ON', 'AN', 'ON']), ('regression: leftover separator neutralisation', [['ES', 'PDI', 'CS', 'AL', 'ET', 'L', 'CS'], 'R'], ['ON', 'PDI', 'ON', 'R', 'ON', 'L', 'ON']), ('regression: leftover separator neutralisation', [['ES', 'R', 'AL', 'CS', 'R', 'ET', 'AN', 'CS', 'R', 'EN', 'AL'], 'R'], ['ON', 'R', 'R', 'ON', 'R', 'ON', 'AN', 'ON', 'R', 'EN', 'R']), ('regression: leftover separator neutralisation', [['NSM', 'ET', 'ES', 'CS', 'EN'], 'R'], ['R', 'ON', 'ON', 'ON', 'EN']), ('NSM after isolate initiator', [['LRI', 'NSM', 'L', 'PDI'], 'R'], ['LRI', 'ON', 'L', 'PDI']), ('Arabic number context', [['AL', 'EN', 'CS', 'EN'], 'L'], ['R', 'AN', 'AN', 'AN']), ('control layout', [['AN', 'ON', 'CS', 'L'], 'L'], ['AN', 'ON', 'ON', 'L']), ('control layout', [['ET', 'EN'], 'L'], ['L', 'L'])], [('regression: leftover separator neutralisation', [['AN', 'PDI', 'EN', 'CS', 'R', 'NSM', 'AN', 'ET'], 'R'], ['AN', 'PDI', 'EN', 'ON', 'R', 'R', 'AN', 'ON']), ('regression: leftover separator neutralisation', [['ON', 'AN', 'LRI', 'ET'], 'R'], ['ON', 'AN', 'LRI', 'ON']), ('regression: leftover separator neutralisation', [['ES', 'ET', 'R', 'R', 'ET', 'ET', 'ON', 'LRI'], 'L'], ['ON', 'ON', 'R', 'R', 'ON', 'ON', 'ON', 'LRI']), ('partial-repair probe', [['ES', 'EN', 'PDI', 'PDI', 'EN', 'L', 'ES'], 'R'], ['ON', 'EN', 'PDI', 'PDI', 'EN', 'L', 'ON']), ('percent after number', [['EN', 'ET', 'ET', 'L'], 'R'], ['EN', 'EN', 'EN', 'L']), ('NSM after isolate initiator', [['LRI', 'NSM', 'L', 'PDI'], 'R'], ['LRI', 'ON', 'L', 'PDI']), ('control layout', [['LRI', 'AN', 'L', 'AL', 'ON'], 'R'], ['LRI', 'AN', 'L', 'R', 'ON']), ('control layout', [['EN', 'EN'], 'L'], ['L', 'L'])], [('regression: leftover separator neutralisation', [['AL', 'ET', 'LRI', 'ES', 'NSM', 'R', 'LRI', 'ES', 'R', 'EN'], 'L'], ['R', 'ON', 'LRI', 'ON', 'ON', 'R', 'LRI', 'ON', 'R', 'EN']), ('regression: leftover separator neutralisation', [['ET', 'ON', 'EN', 'AL', 'EN', 'R', 'ON', 'AN', 'EN'], 'R'], ['ON', 'ON', 'EN', 'R', 'AN', 'R', 'ON', 'AN', 'EN']), ('partial-repair probe', [['EN', 'NSM', 'ES', 'NSM', 'EN', 'LRI', 'NSM', 'WS', 'L'], 'R'], ['EN', 'EN', 'ON', 'ON', 'EN', 'LRI', 'ON', 'WS', 'L']), ('partial-repair probe', [['ES', 'EN'], 'L'], ['ON', 'L']), ('NSM chain after R', [['R', 'NSM', 'NSM', 'EN'], 'L'], ['R', 'R', 'R', 'EN']), ('Arabic number context', [['AL', 'EN', 'CS', 'EN'], 'L'], ['R', 'AN', 'AN', 'AN']), ('control layout', [['L', 'AL', 'PDI', 'LRI', 'AL', 'EN', 'CS', 'EN', 'WS', 'CS'], 'R'], ['L', 'R', 'PDI', 'LRI', 'R', 'AN', 'AN', 'AN', 'WS', 'ON']), ('control layout', [['AL', 'CS', 'PDI'], 'L'], ['R', 'ON', 'PDI'])]]
for label, args, expected in fixtures[N - 1]:
    check(label, solve(args), expected)
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
regression: leftover separator neutralisation	`['R', 'PDI', 'ON', 'ON', 'ON']`	`['R', 'PDI', 'ON', 'ON', 'ON']`	Passed
regression: leftover separator neutralisation	`['R', 'ON', 'ON', 'ON', 'LRI', 'ON', 'R', 'R', 'ON', 'R', 'ON']`	`['R', 'ON', 'ON', 'ON', 'LRI', 'ON', 'R', 'R', 'ON', 'R', 'ON']`	Passed
partial-repair probe	`['R', 'LRI', 'WS', 'ON', 'L', 'ON', 'LRI', 'AN', 'AN']`	`['R', 'LRI', 'WS', 'ON', 'L', 'ON', 'LRI', 'AN', 'AN']`	Passed
partial-repair probe	`['ON', 'L', 'ON', 'R', 'R', 'L', 'ON']`	`['ON', 'L', 'ON', 'R', 'R', 'L', 'ON']`	Passed
percent after number	`['EN', 'EN', 'EN', 'L']`	`['EN', 'EN', 'EN', 'L']`	Passed
NSM chain after R	`['R', 'R', 'R', 'EN']`	`['R', 'R', 'R', 'EN']`	Passed
control layout	`['AN', 'AN', 'AN']`	`['AN', 'AN', 'AN']`	Passed
control layout	`['L', 'R', 'EN', 'ON']`	`['L', 'R', 'EN', 'ON']`	Passed

SHA-256 / ed427f86a60e91015a58c586e339c520581e311aad211053b8c14bf414895716

Verification & scope

A deterministic toy bidi model over stipulated class labels and integer levels; it is inspired by, but does not claim conformance to, any published algorithm. 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 2026-09-29T14:49:53.529003+00:00.

Case digest / ceb1db1eb65c4e26b2d9e5ed294f3277b475442511adabe54193876add9d8515