Microdata Schema Extractor

There are two ways to mark up structured data on a page: microdata, with attributes scattered through the HTML, and JSON-LD, a script block separate from the content. Google understands both, but has recommended the second for years — it is easier to generate, review and maintain, because it does not tangle with the visual markup.

Paste the HTML and the page walks the microdata scopes, builds the property tree and returns the JSON-LD equivalent. The reading respects the format's rules: each property's value comes from the attribute appropriate to its element — content on meta, address on link and anchor, source on image, date on time elements — and the element text in the remaining cases.

It is the right tool for a migration. Older sites tend to have microdata spread across page templates, and rewriting it all by hand is expensive and error-prone. By converting first and reviewing the JSON afterwards, you can add the new block, check it in Google's testing tool and only then strip the old attributes — with no window where the page has no structured data at all.