Have you ever found a website/page that has several or perhaps dozens, hundreds or thousands of files that you need downloaded but don't have the time to manually do it?

wget's recursive function called with -r does that, but also with some quirks to be warned about.
If you're doing it from a standard web based directory structure, you will notice there is still a link to .. and wget will follow that.

Eg. let's say you have files in http://serverip/documents/ and you call wget like this:
wget -r http://serverip/documents, it will get everything inside documents but also browse up to .. and basically download every traversable file that can be followed (obviously this is usually not your intention).

Another thing to watch out for is trying to use multiple sessions to traverse the same directory.
By default wget will overwrite all files in place that it finds are duplicates.  The -nc option stops it from doing it, but I prefer the -N option which compares the time and size of the local and remote files and resumes if necessary and ignores them if they are the same (it doesn't compare by checksum though).  I think -N is what most will find makes sense for them.

Avoid traversing outside of the intended path, by using -L for relative only.

Best Way To Use wget recursively

wget -nH -N -L -r http://serverip/path

-nH means no host directory, otherwise you'll get a structure downloaded that mirrors the remove path which can be annoying.

Eg. it would create serverip/path/file

-N tells us to resume files if they are incomplete but if the remote file is newer or bigger, then resume/overwrite.  Otherwise nothing is done, the file is skipped since there's no sense in downloading the same thing again and overwriting.

-L says stay in the relative path and is the behavior that you probably wanted and expected without using -L

-r is obvious, it means recursive and to download from all links in the specified path

But even the above still does some annoying things, it will traverse as many levels as it can find and see.

wget, download, directory, automatically, recursivelyhave, website, dozens, downloaded, manually, recursive, quirks, eg, http, serverip, documents, browse, traversable, multiple, sessions, traverse, default, overwrite, duplicates, nc, compares, resumes, ignores, doesn, checksum, traversing, relative, recursively, nh, ll, mirrors, resume, incomplete, newer, skipped, downloading, overwriting, links, specified, levels,

Latest Articles