wget notes

Wget is a handy commandline utility for grabbing stuff off the web. It is built into most UNIX systems. If you are on windows, install Git for Windows which comes packaged with Git Bash, a very handy UNIX terminal emulator (note: check if Git Bash has wget installed by typing wget --version. If it is not installed, check this note). Other windows options are Cygwin, MobaXterm, or Bash on Windows Subsystem.

See all the commands by typing wget --help in your terminal or visiting the man page.

Download files from a list

Create a plain text list of urls you want to download, one per line. Use the -i option to pass that list to wget.

wget -i download-file-list.txt

Download all files of a type from a domain

Use the -r (recursive) and -A (accept) options. For example, download all PDFs:

wget -r -A.pdf http://page-with-pdfs.com/

limit to file type: --accept html

wget --spider -o test.log -e robots=off -w 5 --random-wait --mirror --page-requisites --no-parent http://www.example.com

Fully mirror a site

Download everything you need to have an exact copy of a full site locally:

wget --mirror --convert-links --adjust-extension --page-requisites --wait=2 --random-wait http://example.org

If you only want files from a specific set of domains use --domains=LIST with a comma separated list. If you just want to crawl part of a domain use --no-parent flag. If the site has redirects to send you to an authorative version of a page, you might want to use --trust-server-names, which will name the files based on the file URL rather than the original URL (before redirects).

WARC web archive options

WARC is a web archive format that stores page content, response headers, and metadata for a group of web pages. One WARC can contain all the pages gathered during a web harvest. In addition to HTML documents, it can contain binary content such as images.

To create a WARC for any wget crawl, simply add the --warc-file="filename" flag to the command. Wget will harvest the site assets as normal, but additionally create a WARC compressed as a .gz file. For larger sites it’s a good idea to add --warc-max-size=1G to limit the max size of each WARC so they don’t get too big.

Combined with a mirror, it might look like:

wget --mirror --wait=2 --random-wait --warc-file="example.org-archive" http://example.org

If necessary you can also send a different user agent if the server refuses to give content to wget’s default one, like --user-agent=Mozilla. Occasionally it might be necessary to ignore robot.txt for archival purposes, add -e robots=off to the command.

If you want to view your WARC later, use Webrecorder Player.