Layout of the page will hide part of the content when linking to an anchor

[ljacomet]

See for example http://www.ehcache.org/documentation/3.0/xml.html#code-cache-code-elements

[mathieucarbou]

See PR #39

I found that my fix works fine for http://localhost:4000/documentation/#resources_talks because the anchors have been correctly defined in the documentation page:

<a name="current_version" class="anchor" id="current_version"></a>

BUT here: http://localhost:4000/documentation/3.0/xml.html/#code-cache-code-elements

It does not work because there is simply no anchor at all.

<h3 id="code-cache-code-elements”>
    <a class="anchor" href="#code-cache-code-elements"></a>
    <code>&lt;cache&gt;</code> elements
</h3>

Here you only have an empty link.

I verified and the CSS works fine because the <a> element is moved 70px on top of the title, which is what is needed to make the fix work. The issue is that there is no name=“” attribute so that the browser can consider it as an anchor.

Just to confirm, I’ve replaced the above code by:

<h3>
    <a class="anchor" name="code-cache-code-elements"></a>
    <code>&lt;cache&gt;</code> elements
</h3>

And it worked fine :-)

So I don’t know if this empty link is the expected behavior, but if we really want it to be an anchor, then there is something to fix in the ehcache project perhaps ? I didn’t find where actually. I guess it is the way the asciidoc gets generated ?

For the moment, if we go to http://localhost:4000/documentation/3.0/xml.html/#code-cache-code-elements, Chrome simply tries to find an element with the given id, and scroll to it. That is why you get a bug, because it does not consider the anchor, there is none.

[mathieucarbou]

I found a workaround, by using some JS to redefine the anchor names correctly.