Grably — 网页抓取：表格和列表转 CSV

把任意表格或列表抓成 CSV、JSON，或直接粘进表格。可跟随翻页。无需账号，不走云端，全部在浏览器内完成。

Store minimum is 25 characters; this is the English one. Translate per locale only once there is install data worth optimising for.) There is a list on the screen in front of you. You want it in a spreadsheet. That should take thirty seconds and no account. Grably turns any repeating structure on a web page — a table, a product grid, a search-results list, a directory of names — into a CSV file, without sending anything anywhere. HOW IT WORKS 1. Click the Grably icon (or press Alt+Shift+G) on the page you want. 2. Grably scans the page and shows what it found: "Table · 42 rows × 5 cols". Or point at a row yourself — hovering outlines every matching row, and one click takes them all. 3. Fix up the columns: rename them, drop the ones you don't want, or add one by clicking a value on the page. One more click switches a column between the visible text, the link it points to, or an image's source. 4. Download as CSV, JSON, or copy as TSV and paste straight into Sheets. IT FOLLOWS PAGINATION — WITHOUT MOVING YOUR TAB Turn on "More pages" and Grably fetches the following pages in the background and runs the same recipe against each one. Your tab never navigates, so the panel, your column edits, and every row collected so far survive. It recognises rel="next", next-page links in all 18 shipped languages, numbered pagers, and "load more" buttons. It also tells you honestly why a run stopped: page limit reached, no further pages, the next page returned nothing new, a failed request, or you cancelled. Rows are de-duplicated across pages, so a site that quietly loops back to page 1 can't triple your export. SAVED RECIPES Once a page is set up the way you want, save it. Next time you're on that site, the recipe is waiting in the popup — one click and you have the CSV. A recipe stores selectors and your column names. It never stores the data you scraped. NOTHING LEAVES YOUR BROWSER No account, no cloud, no server, no subscription to read your own screen. Everything — detection, extraction, pagination, CSV generation — happens on your device. There is no backend for your data to sit on. PERMISSIONS, AND WHY THERE ARE SO FEW Grably does not ask to read your data on every website. It uses activeTab: nothing runs anywhere until you click the icon on a page you chose, and that access expires with the tab. Beyond that it needs only storage, for your saved recipes and settings. WHAT GRABLY DOES NOT DO It is not a cloud crawler. No scheduled overnight runs, no proxy pools, no CAPTCHA solving, no rate-limit evasion, and it refuses to follow a "next" link onto a different domain. It makes the same requests your own browser would, one at a time, with a pause between them. If you need a crawling farm, buy one — this is for the list already on your screen.