Conformance: JSON-LD in HTML¶

The W3C’s JSON-LD 1.1 test suite has 50 tests of JSON-LD inside HTML: which of a page’s scripts a processor extracts (the first, the one a fragment names, or all of them), which script text is an error, which base IRI the page sets, and what expanding, compacting, flattening or turning the result into RDF gives. Sluicer’s JSON-LD reader, its extruct interface and extruct itself read every page; each answer, as the reader gives it, is processed by PyLD, a JSON-LD processor, with the test’s options, and compared with the suite’s expected result by the suite’s own rules. Regenerated on 2026-09-25 from commit a6e42e6 by uv run bench/w3c_jsonld.py, against the suite at ffdb326121ea; the rules are in bench/PREREG.md.

Read this before the numbers

A reader of JSON-LD is not a JSON-LD processor, and these are a processor’s tests. A reader is given a page and its address and answers the JSON its scripts hold; it takes no option, so it cannot be told to read only the first script or the one a fragment names, and its answer carries no base IRI. Every test that asks for one of those fails for every reader alike, and is listed below as such. What the page measures is what a reader’s answer keeps of what a processor needs: the scripts’ JSON as written, and a refusal where the text is not JSON-LD.

Results¶

Tests passed, of the 50, with the Wilson interval of the share. PyLD reading every page itself is the check that the harness scores a processor as the suite does: it passes 49 of 50.

reads the page	all	expand (21)	compact (4)	flatten (5)	toRdf (20)	negative tests
PyLD 3.3.0	0.980 (0.89–1.00)	21/21	4/4	4/5	20/20	15/15
Sluicer’s JSON-LD reader 0.8.0	0.480 (0.34–0.62)	11/21	1/4	2/5	10/20	7/15
sluicer.compat.extruct 0.8.0	0.660 (0.52–0.78)	15/21	2/4	2/5	14/20	13/15
extruct 0.18.0	0.660 (0.52–0.78)	15/21	2/4	2/5	14/20	13/15

21 of the tests ask for what no reader’s answer can give: a script named by its fragment, the first of several scripts alone, or the base a page’s <base href> sets. On the other 29, which a reader passes or fails by what it answers, Sluicer’s JSON-LD reader 0.690 (0.50–0.83), sluicer.compat.extruct 1.000 (0.88–1.00) and extruct 1.000 (0.88–1.00).

first	against	difference in tests passed (95% interval)	verdict
Sluicer’s JSON-LD reader 0.8.0	extruct 0.18.0	-0.180 (-0.300 to -0.079)	worse
Sluicer’s JSON-LD reader 0.8.0	sluicer.compat.extruct 0.8.0	-0.180 (-0.300 to -0.079)	worse

The difference is the first side’s rate minus the second’s, over the same pages. Its interval is the 95% percentile interval of 10,000 resamples of the pages, drawn together for both sides (bench/stats.py, seed 20260924): better when the interval is above zero, worse when it is below, inconclusive when it holds zero. These are 2 comparisons, made with no correction for making many: where two sides did not differ at all, about one in twenty would still be called better or worse, so read the verdicts as a table, not one at a time.

Where Sluicer’s JSON-LD reader fails¶

26 tests fail. By the first reason that holds, 8 name one script by its fragment, 6 want an error the reader did not give, 5 want the first of several scripts, 4 set their base with the page’s <base href> and 3 hold a @graph the reader answers node by node.

test	what it asks	the answer	why it fails
`te002`	Expands first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (3 values).
`te003`	Expands targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (3 values).
`te004`	Expands all embedded JSON-LD script elements with extractAllScripts option	differs	The reader answers the nodes of a script’s `@graph` one by one, each with the script’s `@context`: the same statements, in the page’s default graph, where a processor given every script puts them in a graph of their own.
`te011`	Errors if no element found at target	answered	The test names the script `#third`; a reader is given no fragment and answers every script (3 values).
`te014`	Errors if uncommented script text contains comment	answered	The test wants the error `invalid script element`: the script is not JSON-LD as written. The reader mended it and answered 1 value.
`te015`	Errors if end comment missing	answered	The test wants the error `invalid script element`: the script is not JSON-LD as written. The reader mended it and answered 1 value.
`te016`	Errors if start comment missing	answered	The test wants the error `invalid script element`: the script is not JSON-LD as written. The reader mended it and answered 1 value.
`te020`	Expands embedded JSON-LD script element relative to HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`te021`	Expands embedded JSON-LD script element relative to relative HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`te022`	Expands targeted JSON-LD script element with fragment and HTML base	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).
`tc002`	Compacts first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (3 values).
`tc003`	Compacts targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (3 values).
`tc004`	Compacts all embedded JSON-LD script elements with extractAllScripts option	differs	The reader answers the nodes of a script’s `@graph` one by one, each with the script’s `@context`: the same statements, in the page’s default graph, where a processor given every script puts them in a graph of their own.
`tf002`	Flattens first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (3 values).
`tf003`	Flattens targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (3 values).
`tf004`	Flattens first script element by default	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (3 values).
`tr002`	Transforms first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (3 values).
`tr003`	Transforms targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (3 values).
`tr004`	Expands all embedded JSON-LD script elements with extractAllScripts option	differs	The reader answers the nodes of a script’s `@graph` one by one, each with the script’s `@context`: the same statements, in the page’s default graph, where a processor given every script puts them in a graph of their own.
`tr011`	Errors if no element found at target	answered	The test names the script `#third`; a reader is given no fragment and answers every script (3 values).
`tr014`	Errors if uncommented script text contains comment	answered	The test wants the error `invalid script element`: the script is not JSON-LD as written. The reader mended it and answered 1 value.
`tr015`	Errors if end comment missing	answered	The test wants the error `invalid script element`: the script is not JSON-LD as written. The reader mended it and answered 1 value.
`tr016`	Errors if start comment missing	answered	The test wants the error `invalid script element`: the script is not JSON-LD as written. The reader mended it and answered 1 value.
`tr020`	Expands embedded JSON-LD script element relative to HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`tr021`	Expands embedded JSON-LD script element relative to relative HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`tr022`	Expands targeted JSON-LD script element with fragment and HTML base	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).

Where sluicer.compat.extruct fails¶

17 tests fail. By the first reason that holds, 8 name one script by its fragment, 5 want the first of several scripts and 4 set their base with the page’s <base href>.

test	what it asks	the answer	why it fails
`te002`	Expands first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (2 values).
`te003`	Expands targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).
`te011`	Errors if no element found at target	answered	The test names the script `#third`; a reader is given no fragment and answers every script (2 values).
`te020`	Expands embedded JSON-LD script element relative to HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`te021`	Expands embedded JSON-LD script element relative to relative HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`te022`	Expands targeted JSON-LD script element with fragment and HTML base	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).
`tc002`	Compacts first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (2 values).
`tc003`	Compacts targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).
`tf002`	Flattens first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (2 values).
`tf003`	Flattens targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).
`tf004`	Flattens first script element by default	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (3 values).
`tr002`	Transforms first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (2 values).
`tr003`	Transforms targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).
`tr011`	Errors if no element found at target	answered	The test names the script `#third`; a reader is given no fragment and answers every script (2 values).
`tr020`	Expands embedded JSON-LD script element relative to HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`tr021`	Expands embedded JSON-LD script element relative to relative HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`tr022`	Expands targeted JSON-LD script element with fragment and HTML base	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).

Where extruct fails¶

17 tests fail. By the first reason that holds, 8 name one script by its fragment, 5 want the first of several scripts and 4 set their base with the page’s <base href>.

test	what it asks	the answer	why it fails
`te002`	Expands first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (2 values).
`te003`	Expands targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).
`te011`	Errors if no element found at target	answered	The test names the script `#third`; a reader is given no fragment and answers every script (2 values).
`te020`	Expands embedded JSON-LD script element relative to HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`te021`	Expands embedded JSON-LD script element relative to relative HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`te022`	Expands targeted JSON-LD script element with fragment and HTML base	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).
`tc002`	Compacts first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (2 values).
`tc003`	Compacts targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).
`tf002`	Flattens first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (2 values).
`tf003`	Flattens targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).
`tf004`	Flattens first script element by default	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (3 values).
`tr002`	Transforms first embedded JSON-LD script element	differs	The test wants the first of the page’s 2 scripts; the reader answers all of them (2 values).
`tr003`	Transforms targeted JSON-LD script element	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).
`tr011`	Errors if no element found at target	answered	The test names the script `#third`; a reader is given no fragment and answers every script (2 values).
`tr020`	Expands embedded JSON-LD script element relative to HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`tr021`	Expands embedded JSON-LD script element relative to relative HTML base	differs	The page’s `<base href>` sets the base IRI; a reader’s answer does not carry it, so it is read against the page’s address.
`tr022`	Expands targeted JSON-LD script element with fragment and HTML base	differs	The test names the script `#second`; a reader is given no fragment and answers every script (2 values).

Where PyLD fails¶

test	what it asks	the answer	why it fails
`tf004`	Flattens first script element by default	differs

Method¶

Suite. w3c/json-ld-api at commit ffdb326121ea, tests/html-manifest.jsonld and the pages and expected results it names, under the W3C Software and Document License. Downloaded into bench/cache/, never committed here.
Readers. Each is given the page’s bytes and its address, https://w3c.github.io/json-ld-api/tests/ and the test’s input, its fragment included. Sluicer’s reader is read_jsonld on sluicer.document.load(html, url=...), what Sluicer’s records are made from; the other two are extract(html, base_url=url, syntaxes=["json-ld"]). Sluicer’s reader answers a number as the text the page wrote, 41.90 and not 41.9, since a record holds every value as text; a processor reads that as a string. None of these pages writes a number.
Processing. The answer, a list of values, is the document PyLD expands, compacts (with the test’s context), flattens or turns into N-Quads, with the test’s base option or else the page’s address as its base, and JSON-LD 1.1’s processing mode.
Comparison. The suite’s JSON-LD object comparison: objects member by member, arrays in any order but a @list’s, language tags in any case; flattened results also with blank nodes relabelled, when they are the same graph; RDF by canonical N-Quads (URDNA2015).
Negative tests. A processor must raise the test’s error. A reader passes one by answering nothing or by raising.
PyLD reads each page itself through a document loader that serves the suite from disk, with the test’s options, extractAllScripts included; it passes a negative test only with the test’s error code.
Environments. Python 3.12; extruct from bench/requirements/extruct.txt, PyLD from bench/requirements/pyld.txt, Sluicer from this checkout.