We've been using clean data-vocabulary.org markup for our Breadcrumbs for a few years, but the support for that is going away. So we've been trying to switch over to RDFa, but have run into an issue with Google Rich Testing Tool validation.
If we simply copy-paste the code snippet, it validates without any problems.
<ol vocab="https://schema.org/" typeof="BreadcrumbList" style="list-style:none; padding:0; margin:0;">
<li property="itemListElement" typeof="ListItem" style="display:inline-block;">
<a property="item" href="/">
<span property="name">Index</span></a>
<meta property="position" content="1"> </li>
» <li property="itemListElement" typeof="ListItem" style="display:inline-block;">
<a property="item" typeof="WebPage"
href="archive/">
<span property="name">News</span></a>
<meta property="position" content="2"> </li> » <li property="itemListElement" typeof="ListItem" style="display:inline-block;">
<a property="item" typeof="WebPage"
href="playstation-3/">
<span property="name">PlayStation 3</span></a>
<meta property="position" content="3">
</li>
</ol>
However, if we try to validate the whole URL with the code on-page, the breadcrumbs throw errors by grabbing our other meta-tags, completely outside of the breadcrumb list code.
Does anyone have any clue why this might be happening? All of the breadcrumbs are "WebPage" type, but the errors all occur on the 1st ListItem. Please see image below for example:
Is this a code problem? Or google test tool problem? The issue is fixed if I simply make the 1st list item breadcrumb a non-URL/non web page