bower
doesn't work like that. You either install the whole package or none of the package. The issue is that bower
doesn't know what parts of the package are or aren't used, so it has to download the whole thing.
Now, you could get the page you want from source (wget
, downloading it, etc) & not worry about the rest, but this might be more trouble than it's worth, as you'd have to be sure the page doesn't link to other pages. You can't do this directly from bower
, either, so you'd have to set it up yourself, and at that point why use bower
?
You can also install the whole package but only use part of it (as in, reference a page, as in <link href="/path/to/package/page_in_package" />
), if you don't care about space issues. I don't believe this will have any effect on bandwidth, etc., as the whole package is on your server, but the client would only request that one page (although the page may cause the client to request other pages with e.g. <link />
s inside it).
Finally, you could install the whole package then delete parts of it, but you have to know what parts to delete, and bower
may try to redownload those files/folders. As far as I can tell through experimentation, bower
won't attempt to re-download anything you delete, as long as the package is the same version as the one in your bower.json
, unless you tell it bower install
or bower install <any package>
. It won't even redownload them if you say bower update <whatever>
. If you don't have the package in your bower.json
, bower
will never delete anything in that folder, but it won't help you if another package needs a different version of this package, either, which is the whole reason for bower
, isn't it?