[#255] Split JSON-LD tests by API

[anatoly-scherbakov]

Part of #255; stacked on #339.

Replaces the combined JSON-LD test module with function-based modules grouped by public API.

Validation:

• pytest tests/jsonld
• pytest --ignore tests/test_manifests.py
• make lint

[github-actions]

Coverage Report

File	Stmts	Miss	Cover	Missing
lib/pyld
canon.py	290	12	96%	36, 47, 51, 55, 85–86, 530, 552, 616, 626, 680–681
context_resolver.py	103	8	92%	78, 137, 151, 197, 203–205, 229
iri_resolver.py	141	3	98%	311–312, 322
jsonld.py	2501	164	93%	252–256, 305, 342, 347, 357, 374–389, 425–426, 468, 476, 496, 504–505, 613–618, 701–702, 717–718, 781–786, 811–812, 827–828, 838–839, 865, 913, 927–928, 951–952, 994, 1001, 1008, 1076, 1101, 1123–1125, 1135–1138, 1153, 1176, 1188–1191, 1287, 1296–1299, 1319, 1348, 1381, 1385, 1419, 1436–1438, 1448–1450, 1493, 1512–1515, 1528, 1577, 1680–1687, 1714–1716, 1742–1745, 1921, 1959, 2265, 2396, 2405–2412, 2506, 2533, 2557, 2565, 3159, 3209, 3228, 3293, 3346, 3349, 3412, 3497, 3531, 3539, 3817, 3822, 3824, 4019–4021, 4170, 4233, 4472, 4760, 4892, 4926, 4955–4958, 4980–4981, 5164, 5228, 5255, 5329, 5331, 5349, 5447, 5496, 5728, 5730, 5769, 5771, 5780, 5922, 6050, 6137, 6257–6271, 6403, 6630, 6636–6640, 6679, 6682, 6772–6773, 6782
util.py	72	3	96%	43, 116, 136
lib/pyld/cli
app.py	44	2	95%	77, 80
entry.py	12	3	75%	9–11, 24
input.py	52	1	98%	57
lib/pyld/documentloader
aiohttp.py	71	22	69%	29–39, 68, 74, 86, 98–106, 118–121, 138, 153–156
file.py	36	1	97%	49
requests.py	52	5	90%	57, 71, 83–91
requests_sqlite_cache.py	25	1	96%	19
scheme_directed.py	15	1	93%	38
type_directed.py	15	1	93%	41
TOTAL	3785	227	94%

Tests	Skipped	Failures	Errors	Time
1931	28 💤	0 ❌	0 🔥	3m 5s ⏱️

[mielvds]

Thanks for the action @anatoly-scherbakov ! May I suggest a slight scope change to this PR. In to_rdf, We have two nquads formatting styles mixed:

assert nquads == (
        '<http://example.com/s> <http://example.com/p> '
        '"1.0E21"^^<http://www.w3.org/2001/XMLSchema#double>  .\n\n'
    )

and

expected = """_:custom0 <http://example.org/p> _:custom1  .
_:custom1 <http://www.w3.org/1999/02/22-rdf-syntax-ns#first> "a"^^<http://www.w3.org/2001/XMLSchema#string>  .
_:custom1 <http://www.w3.org/1999/02/22-rdf-syntax-ns#rest> _:custom2  .
_:custom2 <http://www.w3.org/1999/02/22-rdf-syntax-ns#first> "b"^^<http://www.w3.org/2001/XMLSchema#string>  .
_:custom2 <http://www.w3.org/1999/02/22-rdf-syntax-ns#rest> <http://www.w3.org/1999/02/22-rdf-syntax-ns#nil>  .

"""

Could we unify those to either one? This PR seems like the place to do this

[anatoly-scherbakov]

Updated the complete exact N-Quads expectations in test_to_rdf.py to the named triple-quoted expected style, while retaining the sorted multiline comparisons, substring assertions, and existing TODOs. Focused tests and lint pass.