Compare your page with a detectable DIV table
The sample uses DIV elements with explicit ARIA table, row, and cell roles. It contains three product rows and three columns. If this example is detected but your page is not, compare the source structure and loading behaviour. A visual arrangement of boxes does not necessarily expose a usable table.
Start with a known working table
Open the sample table linked below and run the extension there. If it appears, capture is working and the original page needs a closer look. If it does not, reload the page after installing the extension and check that the extension is enabled.
Chrome restricts extension access on some pages, including browser settings and the Chrome Web Store. Test on a normal website instead. Do not use a protected browser page as your compatibility test.
Match the symptom to the next check
Keep the source page visible while diagnosing it. The important distinction is whether cells are missing from detection or rows are missing from an otherwise valid capture.
| What you see | Likely explanation | What to try |
|---|---|---|
| No table listed | Not loaded yet, restricted page, or unsupported layout | Wait for loading, reload, scan again |
| Report is inside an embedded panel | Separate iframe document | Wait for the frame; check site access |
| Only the visible rows export | Lazy loading or row virtualization | Try Advanced capture; verify the final count |
| Correct first page, missing later pages | Pagination | Capture pages separately or use the site’s export |
| Table behaves like one picture | Image or canvas rendering | Use the original data download or OCR if appropriate |
Wait for the iframe, then scan again
An iframe loads another document inside the page. The outer page can finish loading while the embedded report is still fetching data. Wait until its rows are visible, then reopen the extension and rescan.
The extension supports scanning accessible frames, but browser permissions and the embedding site can limit access. Check the extension’s site-access setting for the page you intend to capture. If the report provides an Open in new tab control, try its own page. Do not assume that every frame can be read.
Use Advanced capture for loaded-on-scroll tables
Advanced capture can scroll and look for more data as the page loads it. Let the report settle before exporting, then compare the captured count with the count shown by the website.
Some grids recycle their row elements: scrolling removes earlier rows while adding later ones. Others paginate through separate requests. Scroll assistance cannot guarantee that all of these designs become one complete table. If the site offers a CSV or Excel download, that may be the more dependable route for the full report.
Avoid combining pages until you have checked headers and repeated records. A missing page is easy to overlook when the first few rows look correct.
Recognize an image or canvas table
If selecting text never reaches individual cell contents, you may be looking at an image or canvas. This is a useful clue, though not a definitive test: a website can also disable text selection.
Look for an accessible data view or a built-in download. For a scanned PDF, OCR may be needed before table reconstruction. OCR can misread zeros, decimal points, and column boundaries, so verify those cells against the source.
What to record when a table still fails
Note the page URL, the steps that make the table appear, the expected row count, and whether the data is inside a frame. A small public example is easier to reproduce than a screenshot of a private dashboard.
Do not include confidential table contents in a bug report. If you can reproduce the issue with sample data, use that. Keep the source export until you have confirmed the capture is complete.
Check loaded rows and report access
For a changing table, compare the exported record count with the source after loading finishes. Pagination and virtualization can leave out records even when the visible columns look correct.