The idea behind the creation of Python in Hypertext Markup Language (phml), is to allow for web page generation with direct access to python. This language takes inspiration directly from frameworks like Vue.js, Astro.js, Solid.js, and SvelteKit. There is conditional rendering, components, python elements, inline/embedded python blocks, and slot, named slots, and much more. Now let's dive into more the language.
Let's start with the new python
element. Python is a whitespace language. As such, phml
has the challenge of maintaining the indentation in an appropriate way as to preserve the intended whitespace. The key focus is the indended whitespace. While this can be tricky the first line with content serves as a reference. The amount of indentation for the first line is removed from each line and the remaining whitespace is left alone. For example if there is a python block that looks like this.
<python> message = "hello world" if "hello" in message: print(message) </python>
The resulting python code would look like this.
message = "hello world"
if "hello" in message:
print(message)
So now we can write python code, now what? You can define functions and variables
how you normally would and they are now available to the scope of the entire file. Consider the following example; You can define function called URL
in the python
element and it can be accessed in any other part of the file. So the code would look like this:
<python>
def URL(link: str) -> str: links = { "youtube": "https://youtube.com" } if
link in links: return links[link] else: return ""
</python>
...
<a href="{URL('youtube')}">Youtube</a>
phml combines all python
elements and treats them as one python file. This is of the likes of the script
or style
tags. With the fact that you can write any code in the python element and used it anywhere else in the file you of the full power of the python programming language at your desposal.
Next up is inline python blocks. These are represented with {{}}
in text elements. Any text in-between the brackets will be processed as python. This is mostly useful when you want to inject a value from python. Assume that there is a variable defined in the python
element called message
and it contains Hello World!
. Now this variable can be used like this, <p>{{ message }}</p>
,
which renders to, <p>Hello World!</p>
.
Note: Inline python blocks are only rendered in a Text element or inside an html attribute.
Conditional rendering with @if
, @elif
, and @else
is an extremely helpful tool in phml.
@if
can be used alone and the python inside it's value must be truthy for the element to be rendered. @elif
requires an element with a @if
or @elif
attribute immediately before it, and it's condition is rendered the same as @if
but only rendered if a @if
or @elif
first fails. @else
requires there to be either a @if
or a @else
immediately before it. It only renders if the previous element's condition fails. If @elif
or @else
is on an element, but the previous element isn't a @if
or @elif
then an exception will occur. Most importantly, the first element in a chain of conditions must be a @if
.
Other than conditions, there is also a built in for loop element. The format looks something like <For :each="item in collection>"
and it duplicates it's children at the node position of the For
element. The For
element requires there to be an each
attribute for it to be rendered. You can consider the value of this element as pythons equivelent to for item in collection:
as this is what the each
attribute expands out to. The attributes defined in the each
element, item
from the previous example, is exposed to the children of the for loop. The attributes from the iteration are scoped recursively through the children. All conditionals work for the the For
element. An added feature is when a For
iteration has an error or iterates zero times, the @elif
or @else
following the For
is used instead. This means that a For
failing or generating zero is like a failed @if
and can be treated as such. Below is an example of how a For
element could be used.
<ul
<For :each="i in range(3)">
<li>{i}</li>
</For>
<li @else>No items in range</li>
</ul>
The compiled html will be:
<ul>
<li>1</li>
<li>2</li>
<li>3</li>
</ul>
Python attributes are shortcuts for using inline python blocks in html attributes. Normally, in phml, you would inject python logic into an attribute similar to this src="{url('youtube')}"
. If you would like to make the whole attribute value a python expression you may prefix any attribute with a :
. This keeps the attribute name the same after the prefix, but tells the parser that the entire value should be processed as python. So the previous example with URL
can also be expressed as <a :href="URL('youtube')>Youtube</a>"
.
PHML includes a powerful component system. The components are partial phml files and are added to the core compiler. After adding the component whenever an element with the same name as the component is found, it is replaced. Components have scoped python
elements, while all style
and script
elements are global to the file they are injected into. Components require that there is only one element, that isn't a python
, script
, or style
tag, to be present. A sample component can look something like the example below.
<!-- Component.phml -->
<div># content goes here</div>
<python> # python code goes here </python>
<style>
/* styles go here */
</style>
<script>
// js goes here
</script>
Components can be added to the compiler by using HypertextManager.add('path/to/component.phml')
. You can define a components name when adding it to the compiler like this HypertextManager.add(('Component', 'path/to/component.phml'))
, or you can just let the compiler figure it out for you. Each directory in the path given along with the file name are combine to create the components name. So if you pass a component path that is path/to/component.phml
it will create a components name of Path.To.Component
which is then used as <Path.To.Component />
. The compiler will try to parse and understand the component name and make it Pascal case. So if you have a file name of CoMP_onEnt.phml
it will result in CoMPOnEnt
. It uses _
as a seperator between words along with capital letters. It will also recognize an all caps word bordering a new word with a capital letter.
Great now you have components. But what if you have a few components that are siblings and you don't want them to be nested in a parent element. PHML provides a <>
element which is a placeholder element. All children are treated as they are at the root of the component.
<!-- file.phml -->
...
<body>
<Component />
</body>
...
<!-- Component.phml -->
<>
<p>Hello</p>
<p>World</p>
<>
will result in the following rendered html
<!-- file.html -->
...
<body>
<p>Hello</p>
<p>World</p>
</body>
...
Now how do you pass information to component to use in rendering? That is where the Props
variable comes in. The Props
variable is a dictionary defined in the components python
element. This defines what attributes on the component are props along with their default values.
<!-- component.phml -->
<python> Props = { message: "" } </python>
<p>{{ message }}</p>
<!-- file.phml -->
...
<Component message="Hello World!" />
...
Both normal attribute values and python attributes can be used for props. The above example really only works for self closing components. What if you want to pass children to the component? That is where slots come in.
<python> Props = { message: "" } </python>
<div class="callout">
<p @if="message is not None">{{ message }}</p>
<slot />
</div>
The Slot
element must be capitalized. When a Slot
element is present any children inside of a component are inserted in place of it. If no children exist then the slot is just removed. What about having multiple slots and having certain components go to certain slot. PHML covers this with the slot
and name
attribute. The slot attribute holds the name of the slot that the child element should be placed in. The name attribute goes on the Slot
element itself giving it it's name. There may only be one Slot
of every name including the default Slot
with no name attribute. An example of this will look something like this.
<!-- component.phml -->
<div>
<slot name="top" />
<slot />
<slot name="bottom" />
</div>
<!-- file.phml -->
...
<Component>
<p slot="bottom">Bottom</p>
<p slot="top">Top</p>
Middle
</Component>
...
<!-- file.html -->
...
<p slot="top">Top</p>
Middle
<p slot="bottom">Bottom</p>
...
PHML also has very basic markdown support. The Markdown
element can be used to render markdown in place of the element itself. The markdown component is an optional feature of phml. To enable the feature you can run pip3 install phml[markdown]
or pip3 install markdown
, then use the component in a phml file. The markdown component can only reference/render a markdown file. To do so, use the :src
/src
attribute to specify the path to the markdown file, relative to the current file. If phml is rendering from a parsed dict or str, then the current working directory is used.
The markdown component uses a few default extensions while parsing. First it uses codehilite
and fenced_code
for highlighting code blocks, this also requires a css file generated with pygmentize
. Lastly, it will use tables
to add the ability to parse markdown tables. This makes the markdown close to a github flavor.
Users may need to add additional markdown extension or to configure them. That is where the :extras
/extras
and :configs
attributes come in. The :extras
attribute is a list of string names of the markdown extensions to add. The extras
attribute, also :extras
, is a space seperated string of the extension names. To configure the extensions the python attribute :configs
is used. The attribute must be a dict of the format { '<extension>': {
: <value> } }
. The options and available values are found in the markdown
modules documentation.
When the markdown is compiled, the resulting html elements are nested in an article
element. All attributes that are not src
, extras
, or :config
on the markdown component are added to the parent article
element and are left unprocessed. This means that python attributes other than :config
or :extras
on the markdown component are left "as is".
<!-- file.phml -->
<Markdown
src="../markdown/file.md"
extras="smarty footnotes"
:configs="{
"footnotes": {
"BACKLINK_TEXT": "$"
}
}"
>
⚠️ This language is in early development stages. Everything is currently subject to change. All forms of feedback are encouraged.
For more information, check out the [API Docs](https://tired-fox.github.io/phml/phml.html)
The current version is able to parse phml using an html parser. This creates a phml ast which then can be converted back to phml or to json.
Use
First import the core parser and compiler, from phml import HypertextManager
. Then you can do the following:
phml = HypertextManager().load("path/to/file.phml")
print(phml.render())
There is method chaining so most if not all methods can be chained. The obvious exception being any method that returns a value.
By default HypertextManager.render()
will return the html
string. If you want to get a json
string you may pass Formats.JSON
. HypertextManager.render(file_type=Formats.JSON)
.
If you want to write to a file you can call phml.write("path/to/output/file.phml")
. Same with render
it defaults to html. You can change this the same way as render
. core.write("path/to/otuput/file.json", file_type=Formats.JSON)
.
For both render
and write
you will first need to call phml.load("path/to/source/file.phml")
. This parses the source file and stores the ast in the parser. render
and write
then use that ast to create the desired output. Optionally if you already have a phml or html string or a properly formatted dict you can call core.parse(data)
which will parse that information similar to load
.
Every time phml.parse
or phml.load
is called it will overwrite the stored ast variable.
There are many more features such as globally exposed variables, components, slots, exposing python files to be used in phml files, etc...
For more information check out the API Docs