tables: add detectVectorGridInRegion napi export for region-scoped vector grid detection (#70)

* feat: add vector grid region detector napi export

Expose region-scoped vector PDF grid detection so TSR callers can reuse native geometry before model fallback.

Made-with: Cursor

* fix: address vector grid review feedback

Return null for rotated vector grids until the coordinate transform has coverage and reject out-of-crop cell boxes surfaced by real-PDF smoke testing.

Made-with: Cursor

* test: add crop bbox plausibility coverage

Cover in-crop, out-of-crop, slack-boundary, and non-positive DPI behavior for vector grid cell bbox validation.

Made-with: Cursor
This commit is contained in:
Abimael Martell
2026-04-28 17:20:15 -07:00
committed by GitHub
parent d196d435d1
commit c186a036fc
4 changed files with 894 additions and 4 deletions
+12
View File
@@ -7,6 +7,7 @@ import {
extractText,
extractTextWithPositions,
extractTextInRegions,
detectVectorGridInRegion,
extractPagesMarkdown,
} from './index.js';
@@ -90,6 +91,17 @@ assert.equal(typeof regionResults[0].regions[0].text, 'string');
assert.equal(typeof regionResults[0].regions[0].needsOcr, 'boolean');
console.log(' extractTextInRegions: OK');
// --- detectVectorGridInRegion ---
console.log('Testing detectVectorGridInRegion...');
const vectorGrid = detectVectorGridInRegion(fixture, 0, [0, 0, 600, 800], 72);
assert.ok(vectorGrid === null || typeof vectorGrid === 'object');
if (vectorGrid) {
assert.ok(Array.isArray(vectorGrid.structureTokens));
assert.ok(Array.isArray(vectorGrid.cellBboxes));
assert.ok(vectorGrid.cellBboxes.every(bbox => Array.isArray(bbox) && bbox.length === 4));
}
console.log(' detectVectorGridInRegion: OK');
// --- extractPagesMarkdown ---
console.log('Testing extractPagesMarkdown...');