css golang mit license scrawl web

Scrawl is a simple command line tool for downloading files referenced on websites using CSS selectors. This application is not meant to be a replacement for curl or Wget, but rather a precision tool for grabbing files when the context in which they are presented is known to. This capability is particularly useful when the path of the desired file is not known but the URL of the website that links to it is (common for download pages).


If you already have the Go environment and toolchain set up, you can get the latest version by running:

$ go install

Otherwise, you can use the pre-built binaries from the project page.


Executing Scrawl with the -help command line argument will trigger online help to be displayed. Below is a more detailed description of what the parameters do.


Let’s say we want to create a script to download the latest Debian package of Anki:

  1. We load up the homepage and are presented with a big download button as shown in the screenshot below:
  2. Let’s copy that link so we can download the latest version with wGet or curl from our script at any time! Hmm, it looks like the path has the version number embedded in the filename. This means that even after a new version of Anki is released, our script will keep getting version 2.0.33 (unless of course it gets deleted).
  3. Let’s inspect the download link in your favorite browser to see what additional information we can get:
  4. It appears that we can easily create a selector for this element: #linux > a:nth-child(2). Note that Chrome provides the option to copy the CSS selector for any element, making knowledge of web technology optional for this step.
  5. Now let’s create a simple download and install script:
    rm -rf /tmp/anki
    mkdir /tmp/anki
    scrawl -attr=href -dir=/tmp/anki -verbose "#linux > a:nth-child(2)"
    sudo dpkg -i /tmp/anki/*.deb
    sudo apt-get install -y -f
    In this script, we prepare an empty download directory and tell Scrawl to scrape, extracting the href property of the download link identified by the CSS selector #linux > a:nth-child(2). We then install the package and bring in any unsatisfied dependencies.