summaryrefslogtreecommitdiff
path: root/doc/tips/JavaScript_to_add_index.html_to_file:_links.mdwn
blob: a11c807e8c7ace028f3a3fba4e33131eb3f36d51 (plain)

The source file foo/bar.mdwn or foo/bar.html generates the page foo/bar/index.html, but the links to the page appear as "foo/bar/". This is fine (and recommended) for pages served by an http server, but it doesn't work when browsing the pages directly using file: URL. The latter might be desirable when testing pages before upload, or if you want to read pages when off-line without access to a web server.

Here is a JavaScript "onload" script which fixes the URLs if the local.protocol isn't http or https:

function fixLinks() {
  var scheme = location.protocol;
 if (scheme=="http:" || scheme=="https:") return;
 var links = document.getElementsByTagName("a");
  for (var i = links.length; --i >= 0; ) {
   var link = links[i];
    var href = link.href;
    var hlen = href.length;
   if (hlen > 0 && link.protocol==scheme && href.charAt(hlen-1) == "/")
     links[i].href = href + "index.html";
  }
}

This can be placed in page.tmpl:

<html>
<head>
<script language="JavaScript">
function fixLinks() {
...
}
</script>
</head>
<body onload="javascript:fixLinks();">
...
</html>

This script has not been extensively tested.