]> dreyeck.freedombox.rocks Git - topic-maps.git/blob - scripts/check-merging.py
test: register seven negative XTM originals (slice 21C)
[topic-maps.git] / scripts / check-merging.py
1 #!/usr/bin/env python3
2 """Compare selected model properties with frozen baselines, never canonical XML.
3
4 Match topics by identifier sets, not observation order or CXTM numbering.
5 This bounded semantic oracle does not implement ISO 13250-4 canonicalization.
6 """
7 import json
8 import pathlib
9 import sys
10 import xml.etree.ElementTree as ET
11
12 root = pathlib.Path(__file__).resolve().parent.parent
13 report = json.loads(pathlib.Path(sys.argv[1]).read_text())
14 assert report['kind'] == 'semantic-observations-not-cxtm'
15 assert report['failed'] == 0 and report['passed'] == len(report['checks'])
16 assert len(set(report['checks'])) == len(report['checks'])
17 expected_ids = {'ctm/in/'+name+'.ctm' for name in [
18 'topic-subject-identifier', 'topic-subject-identifier-wrapped',
19 'topic-subject-identifier-duplicate', 'topic-subject-locator',
20 'topic-subject-locator-wrapped', 'topic-subject-locator-semi',
21 'topic-subject-locator-duplicate', 'topic-item-identifier',
22 'topic-item-identifier-wrapped', 'topic-item-identifier-duplicate', 'name-duplicate']}
23 assert len(report['cases']) == len(expected_ids)
24 assert {case['id'] for case in report['cases']} == expected_ids
25 for case in report['cases']:
26 baseline = root/'vendor/cxtm/ctm/baseline'/(pathlib.Path(case['id']).name+'.cxtm')
27 xml = ET.parse(baseline).getroot()
28 actual = case['actualObjects']
29 expected_topics = xml.findall('topic')
30 assert xml.tag == 'topicMap'
31 assert actual['topicCount'] == len(expected_topics)
32 assert actual['nameCount'] == sum(len(t.findall('name')) for t in expected_topics)
33
34 def key(topic, xml_item=False):
35 values = []
36 for prop in ['itemIdentifiers', 'subjectIdentifiers', 'subjectLocators']:
37 locators = ([e.text for e in topic.findall(prop+'/locator')]
38 if xml_item else topic[prop])
39 locators = [case['documentIRI']+v if v.startswith('#') else v for v in locators]
40 assert len(locators) == len(set(locators))
41 values.append(tuple(sorted(locators)))
42 return tuple(values)
43
44 by_identity = {key(t): t for t in actual['topics']}
45 assert len(by_identity) == actual['topicCount']
46 mapping = {}
47 for expected in expected_topics:
48 observed = by_identity[key(expected, True)]
49 assert observed['parentIsMap'] is True
50 assert all(c.tag in ['itemIdentifiers', 'subjectIdentifiers', 'subjectLocators', 'name'] for c in expected)
51 mapping[int(expected.attrib['number'])] = observed['ordinal']
52 for expected in expected_topics:
53 observed = by_identity[key(expected, True)]
54 names = [(n['value'], n['typeTopicOrdinal'], n['parentTopicOrdinal']) for n in observed['names']]
55 expected_names = [(n.find('value').text, mapping[int(n.find('type').attrib['topicref'])],
56 observed['ordinal']) for n in expected.findall('name')]
57 assert sorted(names) == sorted(expected_names)
58 print('MERGING-BASELINE-SEMANTIC-EVIDENCE-PASS: 11 frozen baselines; no canonicalization or canonical numbering')