tegwick
2b4500bc46
Verifies evidence-source ingest pairs with evidence-anchor DOM selector create/resolve for non-paginated formats (EANCH-WP-0003).
68 lines
No EOL
2.1 KiB
TypeScript
68 lines
No EOL
2.1 KiB
TypeScript
/**
|
|
* Integration round-trip: HTML/Markdown ingest → createSelectors → resolveSelectors.
|
|
*
|
|
* Crosses the source ↔ anchor boundary for non-paginated formats (EANCH-WP-0003).
|
|
*/
|
|
|
|
import { describe, expect, it } from "vitest";
|
|
|
|
import { ingestHtml, ingestMarkdown } from "@citation-evidence/evidence-source";
|
|
import { createSelectors, resolveSelectors } from "@citation-evidence/evidence-anchor/selectors";
|
|
import type { DomSelectionCapture } from "@citation-evidence/evidence-anchor/types";
|
|
|
|
const HTML_FIXTURE = `<!DOCTYPE html>
|
|
<html><body>
|
|
<h1>Hello World</h1>
|
|
<p>Known good quote for HTML ingest testing.</p>
|
|
</body></html>`;
|
|
|
|
const MARKDOWN_FIXTURE = `# Hello World
|
|
|
|
Known good quote for Markdown ingest testing.
|
|
|
|
**bold** text.
|
|
`;
|
|
|
|
const CASES = [
|
|
{
|
|
id: "html-basic",
|
|
ingest: () => ingestHtml(HTML_FIXTURE, { filename: "sample.html" }),
|
|
quote: "Known good quote for HTML ingest testing.",
|
|
},
|
|
{
|
|
id: "markdown-basic",
|
|
ingest: () => ingestMarkdown(MARKDOWN_FIXTURE, { filename: "sample.md" }),
|
|
quote: "Known good quote for Markdown ingest testing.",
|
|
},
|
|
] as const;
|
|
|
|
function domCapture(text: string): DomSelectionCapture {
|
|
return {
|
|
kind: "dom",
|
|
text,
|
|
startPath: [0, 0],
|
|
startOffset: 0,
|
|
endPath: [0, 0],
|
|
endOffset: text.length,
|
|
structuralPath: [{ kind: "section", index: 0 }],
|
|
};
|
|
}
|
|
|
|
describe("create + resolve round-trip — HTML/Markdown", () => {
|
|
for (const c of CASES) {
|
|
it(`${c.id}: known-good quote round-trips with confidence ≥ 0.9`, async () => {
|
|
const { representation } = await c.ingest();
|
|
const selectors = createSelectors(domCapture(c.quote), representation);
|
|
const resolution = resolveSelectors(selectors, representation);
|
|
|
|
expect(resolution.status).toBe("resolved");
|
|
expect(resolution.confidence).toBeGreaterThanOrEqual(0.9);
|
|
|
|
const span = resolution.candidates[0]?.textPosition;
|
|
expect(span).toBeDefined();
|
|
const canonical = representation.canonicalText ?? "";
|
|
expect(canonical.slice(span!.start, span!.end)).toBe(c.quote);
|
|
expect(selectors.some((s) => s.type === "DomRangeSelector")).toBe(true);
|
|
});
|
|
}
|
|
}); |