Skip to content

info2000/web.scraper.workers.dev

 
 

Repository files navigation

Web Scraper

Web Scraper makes it effortless to scrape websites. You provide a URL and CSS selector and it will return you JSON containing the text contents of the matching elements. You can also scrape HTML attribute values by optionally specifying an attribute name.

Website →

Deploy to Cloudflare Workers

Examples

Heading from example.com

web.scraper.workers.dev/?url=example.com&selector=h1

{"result":["Example Domain"]}

Profile details from github.com profile page

web.scraper.workers.dev/?url=github.com/adamschwartz&selector=.vcard-fullname,.d-md-block+[itemprop=worksFor],.d-md-block+[itemprop=homeLocation]&pretty=true

{
  "result": {
    ".vcard-fullname": [
      "Adam Schwartz"
    ],
    ".d-md-block [itemprop=worksFor]": [
      "@cloudflare"
    ],
    ".d-md-block [itemprop=homeLocation]": [
      "Boston, MA"
    ]
  }
}

Random quote/author from quotes.net

web.scraper.workers.dev/?url=quotes.net/random.php&selector=%23disp-quote-body,.author&pretty=true

{
  "result": {
    "#disp-quote-body": [
      "We are advertis'd by our loving friends."
    ],
    ".author": [
      "William Shakespeare"
    ]
  }
}

API

  • Requests are made as GET against https://web.scraper.workers.dev.
  • There are two required query params, url and selector.
  • There are three optional query params, attr, pretty and spaced.
https://web.scraper.workers.dev
  ?url=https://example.com
  &selector=p
  &attr=title
  &pretty=true
  &spaced=true

How it works

If at least url and selector are set, the response value will always be JSON.

If only one node is found on the page matching the selector, the result will be a string. If more than one node is found, the result will be an array of strings.

If an attr is provided, the result will be a string matching only the first node found which has a non-empty value for that HTML attribute.

Query params

url (required)

  • Supports https:// and http:// protocols.
  • If a protocol isn’t found, http:// is prepended.
    • e.g. https://web.scraper.workers.dev/?url=example.com&selector=p

selector (required)

  • Supports the same set of CSS selectors as Cloudflare Workers' HTMLRewriter class
    • As of Oct 10, 2019, this includes:
      • * – any element
      • E – any element of type E
      • E:not(s) – an E element that does not match either compound selector s
      • E.warning – an E element belonging to the class warning
      • E#myid – an E element with ID equal to myid.
      • E[foo] – an E element with a foo attribute
      • E[foo="bar"] – an E element whose foo attribute value is exactly equal to bar
      • E[foo="bar" i] – an E element whose foo attribute value is exactly equal to any (ASCII-range) case-permutation of bar
      • E[foo="bar" s] – an E element whose foo attribute value is exactly and case-sensitively equal to bar
      • E[foo~="bar"] – an E element whose foo attribute value is a list of whitespace-separated values, one of which is exactly equal to bar
      • E[foo^="bar"] – an E element whose foo attribute value begins exactly with the string bar
      • E[foo$="bar"] – an E element whose foo attribute value ends exactly with the string bar
      • E[foo*="bar"] – an E element whose foo attribute value contains the substring bar
      • E[foo|="en"] – an E element whose foo attribute value is a hyphen-separated list of values beginning with en
      • E F – an F element descendant of an E element
      • E > F – an F element child of an E element
  • Supports multiple selectors delimited with a comma.

attr (optional)

  • When attr is not set, the text contents of all matched nodes are returned.
  • When attr is set, that HTML attribute is scraped from the first matching node with a non-empty value for that attribute.

pretty (optional)

  • When false or not included, JSON is minified.
  • When true, JSON is formatted using JSON.stringify(json, null, 2).

spaced (optional)

  • When false or not included, the text nodes of children of the nodes matching selector will be concatenated raw.
  • When true, a single space character is added after the end tag of each child node found.
Examples

Consider the following DOM structure:

<div><p>This is the first paragraph.</p><p>This is another paragraph.</p></div>

If the selector is set to match div, by default the resulting text will be:

This is the first paragraph.This is another paragraph.

This is because there is no space character between </p> and <p>.

With spaced set to true, the result is:

This is the first paragraph. This is another paragraph.

Development

Web Scraper is powered by Cloudflare Workers, heavily utilizing HTMLRewriter for parsing the HTML and scraping the text.

To develop Web Scraper locally, pull down the repo, and follow these steps:

  1. Installing the Workers CLI globally.
npm i @cloudflare/[email protected] -g
  1. Run the preview/watcher inside the repo:
wrangler preview --watch

This will open up the Workers preview experience, so you can test and debug the site. The main source can be found in index.js. As you make changes you’ll see them live in the previewer.

Deploying

Web Scraper is deployed automatically when changes are pushed to master using a GitHub Action and the Workers CLI.

Author

Web Scraper was created by Adam Schwartz.

About

Scrape websites for text by CSS selector.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • JavaScript 100.0%