This is a tool I built to analyze a sample of WordPress.com posts for the most common types of content.
This will analyze each post and make a note of the blocks used. If it's a classic post (or a Gutenberg post with one classic block) it will try to convert to blocks first and use the result of that conversion.
First, you need a CSV file named posts.csv with a blog ID and post ID in each row.
I didn't include the list in the repository for privacy reasons, but you can easily generate one with curl and jq:
SITE=koke.me; curl -s https://public-api.wordpress.com/rest/v1.1/sites/$SITE/posts | jq '.posts[] | [.site_ID, .ID] | @csv'
Then you need to install dependencies with yarn install or npm install.
Then run ./fetch.sh to get the posts metadata. This will store a bunch of JSON files in cache.
Once that's done, run node . cache/* and it will show a report like this:
Sample size: 1578
core/paragraph 88% (1385)
core/image 62% (974)
core/heading 13% (205)
core/list 8% (123)
core/quote 7% (115)
core/separator 5% (81)
core/embed 4% (61)
core/table 0% (7)
core/preformatted 0% (3)
core/shortcode 0% (1)
core/freeform 0% (1)