Datasette plugin for rendering HTML based on JSON values, using the render_cell plugin hook.
This plugin looks for cell values that match a very specific JSON format and converts them into HTML when they are rendered by the Datasette interface.
{
"href": "https://simonwillison.net/",
"label": "Simon Willison"
}
Will be rendered as an <a href="">
link:
<a href="https://simonwillison.net/">Simon Willison</a>
You can set a tooltip on the link using a "title"
key:
{
"href": "https://simonwillison.net/",
"label": "Simon Willison",
"title": "My blog"
}
Produces:
<a href="https://simonwillison.net/" title="My blog">Simon Willison</a>
You can also include a description, which will be displayed below the link. If descriptions include newlines they will be converted to <br>
elements:
select json_object(
"href", "https://simonwillison.net/",
"label", "Simon Willison",
"description", "This can contain" || x'0a' || "newlines"
)
Produces:
<strong><a href="https://simonwillison.net/">Simon Willison</a></strong><br>This can contain<br>newlines
[
{
"href": "https://simonwillison.net/",
"label": "Simon Willison"
},
{
"href": "https://github.com/simonw/datasette",
"label": "Datasette"
}
]
Will be rendered as a comma-separated list of <a href="">
links:
<a href="https://simonwillison.net/">Simon Willison</a>,
<a href="https://github.com/simonw/datasette">Datasette</a>
The href
property must begin with https://
or http://
or /
, to avoid potential XSS injection attacks (for example URLs that begin with javascript:
).
Lists of links cannot include "description"
keys.
The image tag is more complex. The most basic version looks like this:
{
"img_src": "https://placekitten.com/200/300"
}
This will render as:
<img src="https://placekitten.com/200/300">
But you can also include one or more of alt
, caption
, width
and href
.
If you include width or alt, they will be added as attributes:
{
"img_src": "https://placekitten.com/200/300",
"alt": "Kitten",
"width": 200
}
Produces:
<img src="https://placekitten.com/200/300"
alt="Kitten" width="200">
The href
key will cause the image to be wrapped in a link:
{
"img_src": "https://placekitten.com/200/300",
"href": "http://www.example.com"
}
Produces:
<a href="http://www.example.com">
<img src="https://placekitten.com/200/300">
</a>
The caption
key wraps everything in a fancy figure/figcaption block:
{
"img_src": "https://placekitten.com/200/300",
"caption": "Kitten caption"
}
Produces:
<figure>
<img src="https://placekitten.com/200/300"></a>
<figcaption>Kitten caption</figcaption>
</figure>
You can use {"pre": "text"}
to render text in a <pre>
HTML tag:
{
"pre": "This\nhas\nnewlines"
}
Produces:
<pre>This
has
newlines</pre>
If the value attached to the "pre"
key is itself a JSON object, that JSON will be pretty-printed:
{
"pre": {
"this": {
"object": ["is", "nested"]
}
}
}
Produces:
<pre>{
"this": {
"object": [
"is",
"nested"
]
}
}</pre>
- Preformatted text with JSON demo
- Preformatted text demo showing the Mandelbrot Set using this example from the SQLite documentation
The most powerful way to make use of this plugin is in conjunction with SQLite's JSON functions. For example:
select json_object(
"href", "https://simonwillison.net/",
"label", "Simon Willison"
);
You can use these functions to construct JSON objects that work with the plugin from data in a table:
select id, json_object(
"href", url, "label", text
) from mytable;
The json_group_array()
function is an aggregate function similar to group_concat()
- it allows you to construct lists of JSON objects in conjunction with a GROUP BY
clause.
This means you can use it to construct dynamic lists of links, for example:
select
substr(package, 0, 12) as prefix,
json_group_array(
json_object(
"href", url,
"label", package
)
) as package_links
from packages
group by prefix
Since this plugin is designed to be used with SQL that constructs the underlying JSON structure, it is likely you will need to construct dynamic URLs from results returned by a SQL query.
This plugin registers a custom SQLite function called urllib_quote_plus()
to help you do that. It lets you use Python's urllib.parse.quote_plus() function from within a SQL query.
Here's an example of how you might use it:
select id, json_object(
"href",
"/mydatabase/other_table?_search=" || urllib_quote_plus(text),
"label", text
) from mytable;