Skip to content

Latest commit

 

History

History
191 lines (137 loc) · 6.23 KB

README.md

File metadata and controls

191 lines (137 loc) · 6.23 KB
Raw

page-fetch

page-fetch is a tool for researchers that lets you:

  • Fetch web pages using headless Chrome, storing all fetched resources including JavaScript files
  • Run arbitrary JavaScript on many web pages and see the returned values

Installation

page-fetch is written with Go and can be installed with go install:

▶ go install github.com/detectify/page-fetch@latest

Or you can clone the respository and build it manually:

▶ git clone https://github.com/detectify/page-fetch.git
▶ cd page-fetch
▶ go install

Dependencies

page-fetch uses chromedp, which requires that a Chrome or Chromium browser be installed. It uses the following list of executable names in attempting to execute a browser:

  • headless_shell
  • headless-shell
  • chromium
  • chromium-browser
  • google-chrome
  • google-chrome-stable
  • google-chrome-beta
  • google-chrome-unstable
  • /usr/bin/google-chrome

Basic Usage

page-fetch takes a list of URLs as its input on stdin. You can provide the input list using IO redirection:

▶ page-fetch < urls.txt

Or using the output of another command:

▶ grep admin urls.txt | page-fetch

By default, responses are stored in a directory called 'out', which is created if it does not exist:

▶ echo https://detectify.com | page-fetch
GET https://detectify.com/ 200 text/html; charset=utf-8
GET https://detectify.com/site/themes/detectify/css/detectify.css?v=1621498751 200 text/css
GET https://detectify.com/site/themes/detectify/img/detectify_logo_black.svg 200 image/svg+xml
GET https://fonts.googleapis.com/css?family=Merriweather:300i 200 text/css; charset=utf-8
...
▶ tree out
out
├── detectify.com
│   ├── index
│   ├── index.meta
│   └── site
│       └── themes
│           └── detectify
│               ├── css
│               │   ├── detectify.css
│               │   └── detectify.css.meta
...

The directory structure used in the output directory mirrors the directory structure used on the target websites. A ".meta" file is stored for each request that contains the originally requested URL, including the query string), the request and response headers etc.

Options

You can get the page-fetch help output by running page-fetch -h:

▶ page-fetch -h
Request URLs using headless Chrome, storing the results

Usage:
  page-fetch [options] < urls.txt

Options:
  -c, --concurrency <int>   Concurrency Level (default 2)
  -d, --delay <int>         Milliseconds to wait between requests (default 0)
  -e, --exclude <string>    Do not save responses matching the provided string (can be specified multiple times)
  -i, --include <string>    Only save requests matching the provided string (can be specified multiple times)
  -j, --javascript <string> JavaScript to run on each page
  -o, --output <string>     Output directory name (default 'out')
  -p, --proxy <string>      Use proxy on given URL
  -w, --overwrite           Overwrite output files when they already exist
      --no-third-party      Do not save responses to requests on third-party domains
      --third-party         Only save responses to requests on third-party domains

Concurrency

You can change how many headless Chrome processes are used with the -c / --concurrency option. The default value is 2.

Excluding responses based on content-type

You can choose to not save responses that match particular content types with the -e / --exclude option. Any response with a content-type that partially matches the provided value will not be stored; so you can, for example, avoid storing image files by specifying:

▶ page-fetch --exclude image/

The option can be specified multiple times to exclude multiple different content-types.

Including responses based on content-type

Rather than excluding specific content-types, you can opt to only save certain content-types with the -i / --include option:

▶ page-fetch --include text/html

The option can be specified multiple times to include multiple different content-types.

Running JavaScript on each page

You can run arbitrary JavaScript on each page with the -j / --javascript option. The return value of the JavaScript is converted to a string and printed on a line prefixed with "JS":

▶ echo https://example.com | page-fetch --javascript document.domain
GET https://example.com/ 200 text/html; charset=utf-8
JS (https://example.com): example.com

This option can be used for a very wide variety of purposes. As an example, you could extract the href attribute from all links on a webpage:

▶ echo https://example.com | page-fetch --javascript '[...document.querySelectorAll("a")].map(n => n.href)' | grep ^JS
JS (https://example.com): [https://www.iana.org/domains/example]

Setting the output directory name

By default, files are stored in a directory called out. This can be changed with the -o / --output option:

▶ echo https://example.com | page-fetch --output example
GET https://example.com/ 200 text/html; charset=utf-8
▶ find example/ -type f
example/example.com/index
example/example.com/index.meta

The directory is created if it does not already exist.

Using a proxy

The -p / --proxy option can be used to specify a proxy for all requests to use. For example, to use the Burp Suite proxy with default settings, you could run:

▶ echo https://example.com | page-fetch --proxy http://localhost:8080

Overwriting files

By default, when a file already exists, a new file is created with a numeric suffix, e.g. if index already exists, index.1 will be created. This behaviour can be overridden with the -w / --overwrite option. When the option is used matching files will be overwritten instead.

Excluding third-party responses

You may sometimes wish to exclude responses from third-party domains. This can be done with the --no-third-party option. Any responses to requests for domains that do not match the input URL, or one of its subdomains, will not be saved.

Including only third-party responses

On rare occasions you may wish to only store responses to third party domains. This can be done with the --third-party option.