Steffen Becker
1efe981872
A STYLEREF field's result runs are the chapter a reader shows before it repaginates. The DOCX import dropped them and set the field's text empty, so a header saved as .docx came back with a blank running head. ODF kept it. Found by comparing the header and footer zones on a round trip, which nothing did — they are documents of their own and hold what only they can. The corpus gains two legs besides: a second save, and Word's dialect on its way out as ODT.
66 lines
3.0 KiB
TypeScript
66 lines
3.0 KiB
TypeScript
// The one document carrying every feature at once (kitchenSink.ts) had only ever been
|
|
// written — schema-checked and rendered, never read back. Here it makes the trip.
|
|
import { describe, it, expect } from 'vitest';
|
|
import { buildOdt } from '../src/lib/export/odt';
|
|
import { importOdt } from '../src/lib/import/odt';
|
|
import { buildDocx } from '../src/lib/export/docx';
|
|
import { importDocx } from '../src/lib/import/docx';
|
|
import { normalize, firstDiff, stripFontHoist } from './normalize';
|
|
import { kitchenSinkDoc, kitchenSinkSheet, kitchenSinkOptions } from './kitchenSink';
|
|
|
|
type N = any;
|
|
|
|
// A floating frame is placed by its wrap, not by where its anchor sits among the runs,
|
|
// so either format may hand it back at the head of the paragraph.
|
|
const floats = (c: N) => (c.type === 'image' || c.type === 'textBox') && c.attrs?.wrap && c.attrs.wrap !== 'inline';
|
|
function hoistFloats(node: N): N {
|
|
if (node.content?.some(floats)) node.content = [...node.content.filter(floats), ...node.content.filter((c: N) => !floats(c))];
|
|
for (const c of node.content ?? []) hoistFloats(c);
|
|
return node;
|
|
}
|
|
|
|
// ODF leaves a default column width implicit where DOCX always writes w:tcW.
|
|
function stripColwidth(node: N): N {
|
|
if (node.attrs?.colwidth) {
|
|
delete node.attrs.colwidth;
|
|
if (!Object.keys(node.attrs).length) delete node.attrs;
|
|
}
|
|
for (const c of node.content ?? []) stripColwidth(c);
|
|
return node;
|
|
}
|
|
|
|
// ODF has no section that leaves the page: naming a master page is a break (probed),
|
|
// so a section start always reads back as one.
|
|
function stripSectionBreak(node: N): N {
|
|
if (node.attrs?.sectionBreak) delete node.attrs.breakBefore;
|
|
for (const c of node.content ?? []) stripSectionBreak(c);
|
|
return node;
|
|
}
|
|
|
|
const clean = (n: N) => stripSectionBreak(hoistFloats(stripColwidth(stripFontHoist(normalize(structuredClone(n))))));
|
|
|
|
const o = kitchenSinkOptions();
|
|
const args = [{ top: 2, bottom: 2, left: 2, right: 2 }, 'portrait', o.hf, o.language, 'A4',
|
|
kitchenSinkSheet(), 1.25, 'add', false, o.notesSettings, o.props, false, o.pageNumbering,
|
|
o.decor, o.lineNumbering, false, false, true] as const;
|
|
|
|
describe.each([['ODT', buildOdt, importOdt], ['DOCX', buildDocx, importDocx]] as const)(
|
|
'the kitchen sink survives a %s round trip', (_name, build, read) => {
|
|
it('comes back identical, with no warnings', async () => {
|
|
const doc = kitchenSinkDoc() as N;
|
|
const res = read(await (build as N)(doc, ...args)) as N;
|
|
expect(res.warnings ?? []).toEqual([]);
|
|
expect(firstDiff(clean(doc), clean(res.content))).toBeNull();
|
|
});
|
|
|
|
// The zones are documents of their own, and hold what only they can: the running
|
|
// head, the page number and count.
|
|
it('brings its header and footer back as they went out', async () => {
|
|
const res = read(await (build as N)(kitchenSinkDoc() as N, ...args)) as N;
|
|
for (const zone of ['header', 'footer'] as const) {
|
|
expect(firstDiff(clean((o.hf as N)[zone]), clean(res[zone])), zone).toBeNull();
|
|
}
|
|
});
|
|
},
|
|
);
|