tool · one of 34 Pro CSV, TSV, JSON and XLSX files

Scraper

Q: Where do the column names come from?

The markup - a test id, a class token, an itemprop, a tag. Never from a value, because a column named after the first row lies about every other one.

How do I get these cards into a spreadsheet?

Pointing at one card is the whole input. From there it climbs - a click lands on the title inside a card, never on the card - until it reaches the level whose siblings share its shape, then decides the columns from every item rather than from the first one. It reads the page in front of you and no other, which is why it needs no permission over any site.

Reach it Press ⌘ + K on any tab and type scraper. Ctrl + K on Windows and Linux

Picks one item and takes the rest of the list with it, as CSV

reads the page in front of you, and nothing else

writes nothing

sends nothing

items found by climbing from the one you pick

columns decided by all items, not the first

sensitive fields a set holding one is refused whole

plan free to read and copy · Pro saves the file

Add LoupeKit

TWO BUILDS 9 browsers

Chrome from the Chrome Web Store Firefox from addons.mozilla.org Edge from Microsoft Edge Add-ons Brave takes the Chrome build Opera takes the Chrome build 4 more the Chrome build

in the panel · this one is Pro

section#about

What this actually is

any repeating block on the page - cards, results, rows - as csv, tsv or json

Any list, table of results, set of cards or search page whose contents have to become a file. Not a general crawler: it has no way to visit a second page, and that is deliberate.

section#asked

Asked about this tool

4 of them

Why does it not take a URL to scrape?

Because that needs a grant over every site, and that is a different product. This one has promised to read only the tab you invoked it on.

How does it decide two blocks are the same kind of thing?

Tag plus a majority of the class tokens, measured against the smaller set. Requiring equality drops the card carrying is-featured; requiring only an overlap groups a heading with a paragraph that shares one utility class.

Where do the column names come from?

The markup - a test id, a class token, an itemprop, a tag. Never from a value, because a column named after the first row lies about every other one.

Why is a link two columns?

Because a name and a URL joined into one cell is a spreadsheet column nobody can sort or click. Name and Name URL are separate, and the URL is read off the property so it arrives resolved.

how to use scraper

Point at one card. A click on the title inside it is enough.
It climbs from there until it reaches the level whose siblings share that shape.
Read the columns. They are decided by every item, not by the first one.
Export the file. It reads the page in front of you and has no way to visit a second one.

Add LoupeKit

TWO BUILDS 9 browsers

Chrome from the Chrome Web Store Firefox from addons.mozilla.org Edge from Microsoft Edge Add-ons Brave takes the Chrome build Opera takes the Chrome build 4 more the Chrome build

70 of these · no host permissions · three free audits