]> dreyeck.freedombox.rocks Git - topic-maps.git/blob - scripts/check-cxtm.py
Implement Topic-set Scope across TMDM, CTM and CXTM
[topic-maps.git] / scripts / check-cxtm.py
1 #!/usr/bin/env python3
2 """Independent raw UTF-8 comparison, XML reference integrity and C14N checks."""
3 import json
4 import pathlib
5 import sys
6 import xml.etree.ElementTree as ET
7 root = pathlib.Path(__file__).resolve().parent.parent
8 report = json.loads(pathlib.Path(sys.argv[1]).read_text())
9 assert report['kind'] == 'cxtm-byte-equality' and report['failed'] == 0
10 assert report['passed'] == len(report['checks']) == len(set(report['checks']))
11 expected_count = {'--slice-13': 61, '--slice-12a': 58, '--slice-5': 28, '--slice-6': 33, '--slice-7': 39, '--slice-8': 46, '--slice-9': 51, '--slice-10': 55, '--slice-11': 56}.get(sys.argv[2] if len(sys.argv) > 2 else '', 24)
12 assert len(report['cases']) == expected_count
13 for case in report['cases']:
14 actual = case['actualCXTM'].encode('utf-8')
15 expected = (root / 'vendor/cxtm/ctm/baseline' / (pathlib.Path(case['id']).name + '.cxtm')).read_bytes()
16 assert expected == case['expectedCXTM'].encode('utf-8')
17 discrepancy = expected_count >= 55 and (case['id'] == 'ctm/in/occurrence-datetime2.ctm' or expected_count >= 56 and case['id'] == 'ctm/in/occurrence-datetime-timezone2.ctm')
18 if discrepancy:
19 assert actual != expected and case['status'] == 'fail'
20 if case['id'].endswith('timezone2.ctm'):
21 assert b'<value>2010-12-31T22:00:00Z</value>' in actual
22 assert b'<value>2011-01-01T00:00:00+02:00</value>' in expected
23 else:
24 assert b'<value>1976-09-19T12:02:00</value>' in actual
25 assert b'<value>1976-09-19T12:01:60</value>' in expected
26 else:
27 assert actual == expected and case['status'] == 'pass', case['id']
28 assert case['byteCount'] == len(actual)
29 assert case['baseLocator'].endswith('/' + case['id'])
30 # CXTM 3.1 adds an LF after the root. XML reparsing drops that Infoset
31 # sibling; this check is separate from the unmodified byte comparison.
32 assert actual.endswith(b'</topicMap>\n') and not actual.startswith(b'\xef\xbb\xbf')
33 assert ET.canonicalize(case['actualCXTM']) + '\n' == case['actualCXTM'], case['id']
34 xml = ET.fromstring(actual)
35 topics = xml.findall('topic'); associations = xml.findall('association')
36 assert [t.attrib['number'] for t in topics] == list(map(str,range(1,len(topics)+1)))
37 role_refs = {}
38 for i, a in enumerate(associations,1):
39 assert a.attrib['number'] == str(i)
40 for j, r in enumerate(a.findall('role'),1):
41 assert r.attrib['number'] == str(j)
42 role_refs[f'association.{i}.role.{j}'] = r.find('player').attrib['topicref']
43 for t in topics:
44 assert [o.attrib['number'] for o in t.findall('occurrence')] == list(map(str, range(1,len(t.findall('occurrence'))+1)))
45 for r in t.findall('rolePlayed'):
46 assert role_refs[r.attrib['ref']] == t.attrib['number']
47 assert sum(len(t.findall('rolePlayed')) for t in topics) == len(role_refs)
48 for ref in xml.iter():
49 if 'topicref' in ref.attrib:
50 assert 1 <= int(ref.attrib['topicref']) <= len(topics)
51 passes = sum(c['status'] == 'pass' for c in report['cases'])
52 fails = sum(c['status'] == 'fail' for c in report['cases'])
53 print(f'CXTM-INDEPENDENT-PASS: {passes} byte-exact unchanged baselines; {fails} explicit normative discrepancy; canonical XML and reference integrity')