Usage¶

Elements are imported directly from the htpy module as their name. HTML attributes are specified by parenthesis (() / "call"). Children are specified using square brackets ([] / "getitem").

>>> from htpy import div
>>> print(div(id="hi")["Hello!"])
<div id="hi">Hello!</div>

Elements¶

Children can be strings, markup, other elements or lists/iterators.

Elements can be arbitrarily nested:

Nested elements

>>> from htpy import article, section, p
>>> print(section[article[p["Lorem ipsum"]]])
<section><article><p>Lorem ipsum</p></article></section>

Text/Strings¶

It is possible to pass a string directly:

Using a string as children

>>> from htpy import h1
>>> print(h1["Welcome to my site!"])
<h1>Welcome to my site!</h1>

Strings are automatically escaped to avoid XSS vulnerabilities. It is convenient and safe to directly insert variable data via f-strings:

>>> from htpy import h1
>>> user_supplied_name = "bobby </h1>"
>>> print(h1[f"hello {user_supplied_name}"])
<h1>hello bobby &lt;/h1&gt;</h1>

Conditional Rendering¶

True, False and None will not render anything. Python's and and or operators will short-circuit. You can use this to conditionally render content with inline and and or.

Conditional rendering with a value that may be None

>>> from htpy import div, b
>>> error = None

>>> # No <b> tag will be rendered since error is None
>>> print(div[error and b[error]])
<div></div>

>>> error = 'Enter a valid email address.'
>>> print(div[has_error and b[error_message]])
<div><b>Enter a valid email address.</b></div>

# Inline if/else can also be used:
>>> print(div[b[error] if error else None])
<div><b>Enter a valid email address.</b></div>

Conditional rendering based on a bool variable

>>> from htpy import div
>>> is_happy = True
>>> print(div[is_happy and "😄"])
<div>😄</div>

>>> is_sad = False
>>> print(div[is_sad and "😔"])
<div></div>

>>> is_allowed = True
>>> print(div[is_allowed or "Access denied!"])
<div></div>

>>> is_allowed = False
>>> print(div[is_allowed or "Access denied!"])
<div>Access denied</div>

Loops / Iterating Over Children¶

You can pass a list, tuple or generator to generate multiple children:

Iterate over a generator

>>> from htpy import ul, li
>>> print(ul[(li[letter] for letter in "abc")])
<ul><li>a</li><li>b</li><li>c</li></ul>

Note

The generator will be lazily evaluated when rendering the element, not directly when the element is constructed. See Streaming for more information.

A list can be used similar to a JSX fragment:

Render a list of child elements

>>> from htpy import div, img
>>> my_images = [img(src="a.jpg"), img(src="b.jpg")]
>>> print(div[my_images])
<div><img src="a.jpg"><img src="b.jpg"></div>

Custom Elements / Web Components¶

Custom elements / web components are HTML elements that contains at least one dash (-). Since - cannot be used in Python identifiers, use underscore (_) instead:

Using custom elements

>>> from htpy import my_custom_element
>>> print(my_custom_element['hi!'])
<my-custom-element>hi!</my-custom-element>

Injecting Markup¶

If you have HTML markup that you want to insert without further escaping, wrap it in Markup from the markupsafe library. markupsafe is a dependency of htpy and is automatically installed:

Injecting markup

>>> from htpy import div
>>> from markupsafe import Markup
>>> print(div[Markup("<foo></foo>")])
<div><foo></foo></div>

If you are generate Markdown and want to insert it into an element, use Markup:

Injecting generated markdown

>>> from markdown import markdown
>>> from markupsafe import Markup
>>> from htpy import div
>>> print(div[Markup(markdown('# Hi'))])
<div><h1>Hi</h1></div>

HTML Doctype¶

The HTML doctype is automatically prepended to the <html> tag:

>>> from htpy import html
>>> print(html)
<!doctype html><html></html>

HTML Comments¶

Since the Python code is the source of the HTML generation, to add a comment to the code, most of the time regular Python comments (#) are used.

If you want to emit HTML comments that will be visible in the browser, use the comment function:

>>> from htpy import div, comment
>>> print(div[comment("This is a HTML comment, visible in the browser!")])
<div><!-- This is a HTML comment, visible in the browser! --></div>

It is safe to pass arbitrary text to the comment function. Double dashes (--) will be removed to avoid being able to break out of the comment.

If you need full control over the exact rendering of the comment, you can create comments or arbitrary text by injecting your own markup. See the Injecting Markup section above for details.

Attributes¶

HTML attributes are defined by calling the element. They can be specified in a couple of different ways.

Elements Without Attributes¶

Some elements do not have attributes, they can be specified by just the element itself:

>>> from htpy import hr
>>> print(hr)
<hr>

Keyword Arguments¶

Attributes can be specified via keyword arguments:

>>> from htpy import img
>>> print(img(src="picture.jpg"))
<img src="picture.jpg">

In Python, class and for cannot be used as keyword arguments. Instead, they can be specified as class_ or for_ when using keyword arguments:

>>> from htpy import label
>>> print(label(for_="myfield"))
<label for="myfield"></label>

Attributes that contains dashes - can be specified using underscores:

>>> from htpy import form
>>> print(form(hx_post="/foo"))
<form hx-post="/foo"></form>

Id/Class Shorthand¶

Defining id and class attributes is common when writing HTML. A string shorthand that looks like a CSS selector can be used to quickly define id and classes:

Define id

>>> from htpy import div
>>> print(div("#myid"))
<div id="myid"></div>

Define multiple classes

>>> from htpy import div
>>> print(div(".foo.bar"))
<div class="foo bar"></div>

Combining both id and classes

>>> from htpy import div
>>> print(div("#myid.foo.bar"))
<div id="myid" class="foo bar"></div>

Attributes as Dict¶

Attributes can also be specified as a dict. This is useful when using attributes that are reserved Python keywords (like for or class), when the attribute name contains a dash (-) or when you want to define attributes dynamically.

Using Alpine.js with @-syntax (shorthand for x-on)

>>> from htpy import button
>>> print(button({"@click.shift": "addToSelection()"}))
<button @click.shift="addToSelection()"></button>

Using an attribute with a reserved keyword

>>> from htpy import label
>>> print(label({"for": "myfield"}))
<label for="myfield"></label>

Boolean/Empty Attributes¶

In HTML, boolean attributes such as disabled are considered "true" when they exist. Specifying an attribute as True will make it appear (without a value). False will make it hidden. This is useful and brings the semantics of bool to HTML.

True bool attribute

>>> from htpy import button
>>> print(button(disabled=True))
<button disabled></button>

False bool attribute

>>> from htpy import button
>>> print(button(disabled=False))
<button></button>

Conditionally Mixing CSS Classes¶

To make it easier to mix CSS classes, the class attribute accepts a list of class names or a dict. Falsey values will be ignored.

>>> from htpy import button
>>> is_primary = True
>>> print(button(class_=["btn", {"btn-primary": is_primary}]))
<button class="btn btn-primary"></button>
>>> is_primary = False
>>> print(button(class_=["btn", {"btn-primary": is_primary}]))
<button class="btn"></button>
>>>

Combining Modes¶

Attributes via id/class shorthand, keyword arguments and dictionary can be combined:

Specifying attribute via multiple arguments

>>> from htyp import label
>>> print(label("#myid.foo.bar", {'for': "somefield"}, name="myname",))
<label id="myid" class="foo bar" for="somefield" name="myname"></label>

Escaping of Attributes¶

Attributes are always escaped. This makes it possible to pass arbitrary HTML fragments or scripts as attributes. The output may look a bit obfuscated since all unsafe characters are escaped but the browser will interpret it correctly:

>>> from htpy import button
>>> print(button(id="example", onclick="let name = 'andreas'; alert('hi' + name);")["Say hi"])
<button onclick="let name = &#39;andreas&#39;; alert(&#39;hi&#39; + name);">Say hi</button>

In the browser, the parsed attribute as returned by document.getElementById("example").getAttribute("onclick") will be the original string let name = 'andreas'; alert('hi' + name);.

Escaping will happen whether or not the value is wrapped in markupsafe.Markup or not. This may seem confusing at first but is useful when embedding HTML snippets as attributes:

Escaping of Markup

>>> from htpy import ul
>>> from markupsafe import Markup
>>> # This markup may come from another library/template engine
>>> some_markup = Markup("""<li class="bar"></li>""")
>>> print(ul(data_li_template=some_markup))
<ul data-li-template="&lt;li class=&#34;bar&#34;&gt;&lt;/li&gt;"></ul>

Render elements without a parent (orphans)¶

In some cases such as returning partial content it is useful to render elements without a parent element. This is useful in HTMX partial responses.

You may use render_node to achieve this:

Render elements without a parent

>>> from htpy import render_node, tr
>>> print(render_node([tr["a"], tr["b"]]))
<tr>a</tr><tr>b</tr>

render_node() accepts all kinds of Node objects. You may use it to render anything that would normally be a children of another element.

Best practice: Only use render_node() to render non-Elements

You can render regular elements by using str(), e.g. str(p["hi"]). While render_node() would give the same result, it is more straightforward and better practice to just use str() when rendering a regular element. Only use render_node() when you do not have a parent element.

Iterating of the Output¶

Iterating over a htpy element will yield the resulting contents in chunks as they are rendered:

>>> from htpy import ul, li
>>> for chunk in ul[li["a"], li["b"]]:
...     print(f"got a chunk: {chunk!r}")
...
got a chunk: '<ul>'
got a chunk: '<li>'
got a chunk: 'a'
got a chunk: '</li>'
got a chunk: '<li>'
got a chunk: 'b'
got a chunk: '</li>'
got a chunk: '</ul>'

Just like render_node(), there is iter_node() that can be used when you need to iterate over a list of elements without a parent:

>>> from htpy import li, iter_node
>>> for chunk in iter_node([li["a"], li["b"]]):
...     print(f"got a chunk: {chunk!r}")
...
got a chunk: '<li>'
got a chunk: 'a'
got a chunk: '</li>'
got a chunk: '<li>'
got a chunk: 'b'
got a chunk: '</li>'

Passing Data with Context¶

Usually, you pass data via regular function calls and arguments via your components. Contexts can be used to avoid having to pass the data manually between components. Contexts in htpy is conceptually similar to contexts in React.

Using contexts in htpy involves:

Creating a context object with my_context = Context(name[, *, default]) to define the type and optional default value of a context variable.
Using my_context.provider(value, lambda: children) to set the value of a context variable for a subtree.
Adding the @my_context.consumer decorator to a component that requires the context value. The decorator will add the context value as the first argument to the decorated function:

The Context class is a generic and fully supports static type checking.

The values are passed as part of the tree used to render components without using global state. It is safe to use contexts for lazy constructs such as callables and generators.

A context value can be passed arbitrarily deep between components. It is possible to nest multiple context provider and different values can be used in different subtrees.

A single component can consume as many contexts as possible by using multiple decorators:

@context_b.consumer
@context_a.consumer
def my_component(a, b):
    ...

Example¶

This example shows how context can be used to pass data between components:

theme_context: Context[Theme] = Context("theme", default="light") creates a context object that can later be used to define/retrieve the value. In this case, "light" acts as the default value if no other value is provided.
theme_context.provider(value, lambda: subtree) defines the value of the theme_context for the subtree. In this case the value is set to "dark" which overrides the default value.
The sidebar component uses the @theme_context.consumer decorator. This will make htpy pass the current context value as the first argument to the component function.
In this example, a Theme type is used to ensure that the correct types are used when providing the value as well as when it is consumed.

from typing import Literal

from htpy import Context, Node, div, h1

Theme = Literal["light", "dark"]

theme_context: Context[Theme] = Context("theme", default="light")


def my_page() -> Node:
    return theme_context.provider(
        "dark",
        lambda: div[
            h1["Hello!"],
            sidebar("The Sidebar!"),
        ],
    )


@theme_context.consumer
def sidebar(theme: Theme, title: str) -> Node:
    return div(class_=f"theme-{theme}")[title]


print(my_page())

Output:

<div><h1>Hello!</h1><div class="theme-dark">The Sidebar!</div></div>