Datasets:
File size: 8,392 Bytes
0e239c6 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 | """Read real execution evidence from the five frozen Node/npm protocols."""
from collections import Counter
import json
from pathlib import Path
import re
def packets(text, prefix):
return [json.loads(line[len(prefix):]) for line in text.splitlines() if line.startswith(prefix)]
def parse_report(iid, repository, kind, stdout, stderr, returncode, cell):
phase = 'p2p' if kind == 'pass_to_pass' else 'f2p'
rows = []; issues = []; metadata = {}
try:
if iid == 'buildpkg-npm-cli-8703':
raw = packets(stdout, 'CROSSREPO_GIT_CHAIN_CASE ')
summaries = packets(stdout, 'CROSSREPO_GIT_CHAIN_SUMMARY ')
expected = 19 if phase == 'p2p' else 31
if len(raw) != expected or len(summaries) != 1:
raise ValueError('Incomplete public Git-chain cases')
s = summaries[0]
rows = [{'id': r['name'], 'passed': r['pass']} for r in raw]
if any(type(r['passed']) is not bool for r in rows):
issues.append('Nonboolean Git-chain case outcome')
if s['mode'] != phase or s['executed'] != expected or s['passed'] != sum(r['passed'] for r in rows) or s['failed'] != sum(not r['passed'] for r in rows):
issues.append('Git-chain summary differs from cases')
if not s['real_cli'] or not s['provider'].startswith('/testbed/npm/pacote/'):
issues.append('Missing paired provider/real CLI evidence')
metadata['count_unit'] = 'public provider/CLI behavioral case; shared suite executed separately per repository command'
elif iid == 'buildpkg-npm-cli-9026':
report_path = Path(cell) / (repository + '-' + kind + '-tap-report-v2.json')
if not report_path.exists(): report_path = Path(cell) / (repository + '-' + kind + '-tap-report.json')
report = json.loads(report_path.read_text())
all_assertions = report['assertions']; assertions = []
filtered = []
for r in all_assertions:
if r.get('skip'):
if 'filter' not in str(r['skip']).lower():
issues.append('Native TAP assertion skipped outside frozen grep selection')
filtered.append(r)
elif r.get('todo'):
issues.append('Native TAP TODO assertion')
else:
assertions.append(r)
if report['parser_errors'] or report['bailouts'] or not assertions:
issues.append('Native TAP plan/parser/bailout or empty execution error')
summaries = report['summaries']
if not summaries or any(s['plan']['start'] is None or s['plan']['end'] is None for s in summaries):
issues.append('Missing native TAP completion plan')
native_failed = sum(r['ok'] is False for r in assertions)
if returncode != int(bool(native_failed)):
issues.append('Harness exit differs from executed native assertions')
rows = [{'id': repository + '/' + phase + '/native-TAP-contract', 'passed': not native_failed}]
metadata = {'count_unit': 'native TAP contract command', 'native_assertions': len(assertions),
'native_passed': len(assertions) - native_failed, 'native_failed': native_failed,
'grep_filtered_assertions': len(filtered), 'native_cases': assertions,
'tap_summary_count': len(summaries)}
elif iid == 'buildpkg-npm-cli-9586':
summaries = packets(stdout, 'CROSSSWE_COUNTS ')
if len(summaries) != 1:
raise ValueError('Missing npm policy scenario outcomes')
s = summaries[0]; raw = s['cases']
expected = {('npm-cli', 'p2p'): 3, ('npm-packlist', 'p2p'): 2, ('npm-cli', 'f2p'): 7, ('npm-packlist', 'f2p'): 1}[repository, phase]
rows = [{'id': r['name'], 'passed': r['passed']} for r in raw]
if len(rows) != expected or s['tests_run'] != expected or s['mode'] != phase or s['repo'] != repository or s['skipped']:
issues.append('Incomplete npm policy scenario inventory')
if s['passed'] != sum(r['passed'] for r in rows) or s['failed'] != sum(not r['passed'] for r in rows):
issues.append('npm policy summary differs from raw cases')
metadata['count_unit'] = 'public npm install/pack scenario'
elif iid == 'mcp-specification-2508':
data = [json.loads(line) for line in stdout.splitlines() if line.startswith('{')]
compilers = [r for r in data if 'compiler' in r]
positives = [r for r in data if r.get('check') == 'valid-json-capabilities']
final = [r for r in data if r.get('status') == 'passed']
if not compilers or not positives or any(r['diagnostics'] for r in positives):
issues.append('Missing real compiler or positive regression evidence')
if returncode == 0:
if len(final) != 1 or final[0]['repo'] != repository or final[0]['mode'] != phase:
issues.append('Missing full MCP runtime completion')
else:
if phase != 'f2p' or 'JSON safety regression: compiler accepted non-JSON extension settings:' not in stderr:
issues.append('MCP failure is not the frozen behavioral assertion')
if not any(r.get('check') == 'non-json-capabilities' for r in data):
issues.append('Negative compiler behavior did not execute')
if repository == 'typescript_sdk':
commands = [r for r in data if 'command' in r]
if not commands or any(r['status'] != 0 for r in commands) or not any(r.get('generatedFileRestored') for r in data):
issues.append('Production generator did not complete and restore its output')
rows = [{'id': repository + '/' + phase + '/compiler-generator-contract', 'passed': returncode == 0}]
metadata = {'count_unit': 'compiler/generator contract command', 'compiler_evidence': compilers,
'checks': [r for r in data if 'check' in r], 'completion': final}
elif iid == 'mcp-specification-2663':
raw = re.findall(r'^CONTRACT_(PASS|FAIL) (.+)$', stdout, re.M)
summaries = packets(stdout, 'CONTRACT_SUMMARY ')
expected_names = {('spec', 'p2p'): ['core-schema-valid-and-invalid', 'mrtr-and-tool-result-regression'],
('spec', 'f2p'): ['core-request-removal', 'public-ts-client-request-removal'],
('conformance', 'p2p'): ['default-connection-core-wire'],
('conformance', 'f2p'): ['extension-public-behavior']}[repository, phase]
if [r[1] for r in raw] != expected_names or len(summaries) != 1:
raise ValueError('Incomplete MCP scenario execution')
rows = [{'id': name, 'passed': status == 'PASS'} for status, name in raw]
s = summaries[0]
if s['group'] != repository + ':' + phase or s['total'] != len(rows) or s['skipped'] or s['passed'] != sum(r['passed'] for r in rows) or s['failed'] != sum(not r['passed'] for r in rows):
issues.append('MCP scenario summary differs from raw execution')
metadata = {'count_unit': 'schema or live HTTP conformance scenario',
'observations': packets(stdout, 'OBSERVATION '), 'wire_evidence_records': len(packets(stdout, 'WIRE_EVIDENCE '))}
else:
raise ValueError('Unsupported batch03 protocol')
except (ValueError, KeyError, TypeError, IndexError, OSError) as error:
issues.append('Incomplete/malformed case evidence: ' + type(error).__name__)
ids = [r['id'] for r in rows]
if not rows or len(ids) != len(set(ids)):
issues.append('Empty or duplicate case inventory')
if any(token in stdout + stderr for token in ['MODULE_NOT_FOUND', 'ERR_MODULE_NOT_FOUND', 'Cannot find module', 'ModuleNotFoundError:', 'TimeoutExpired:']):
issues.append('Import/dependency/infrastructure failure')
if rows and returncode != int(any(r['passed'] is False for r in rows)):
issues.append('Exit status disagrees with case outcomes')
return rows, issues, metadata
|