django-tag-parser 2.1

Arguments and keyword arguments

To parse a syntax like:

{% my_tag "arg1" keyword1="bar" keyword2="foo" %}

use:

from django.template import Library
from tag_parser import template_tag
from tag_parser.basetags import BaseNode

register = Library()


@template_tag(register, 'my_tag')
class MyTagNode(BaseNode):
    max_args = 1
    allowed_kwargs = ('keyword1', 'keyword2',)

    def render_tag(self, context, *tag_args, **tag_kwargs):
        return "Tag Output"

Inclusion tags

To create an inclusion tag with overwritable template_name:

{% my_include_tag "foo" template="custom/example.html" %}

use:

from django.template import Library
from tag_parser import template_tag
from tag_parser.basetags import BaseInclusionNode

register = Library()


@template_tag(register, "my_include_tag")
class MyIncludeTag(BaseInclusionNode):
    template_name = "mytags/default.html"
    max_args = 1

    def get_context_data(self, parent_context, *tag_args, **tag_kwargs):
        (foo,) = *tag_args
        return {
            'foo': foo
        }

The get_template_name() method can be overwritten too to support dynamic resolving of template names. By default it checks the template tag_kwarg, and template_name attribute. Note the template nodes are cached afterwards, it’s not possible to return random templates at each call.

Assignment tags

To create assignment tags that can either render itself, or return context data:

{% get_tags template="custom/example.html" %}
{% get_tags as popular_tags %}

use:

from django.template import Library
from tag_parser import template_tag
from tag_parser.basetags import BaseAssignmentOrInclusionNode

register = Library()


@template_tag(register, 'get_tags')
class GetPopularTagsNode(BaseAssignmentOrInclusionNode):
    template_name = "myblog/templatetags/popular_tags.html"
    context_value_name = 'tags'
    allowed_kwargs = (
        'order', 'orderby', 'limit',
    )

    def get_value(self, context, *tag_args, **tag_kwargs):
        return query_tags(**tag_kwargs)   # Something that reads the tags.

Block tags

To have a “begin .. end” block, define end_tag_name in the class:

{% my_tag keyword1=foo %}
    Tag contents, possibly other tags.
{% end_my_tag %}

use:

from django.template import Library
from tag_parser import template_tag
from tag_parser.basetags import BaseAssignmentOrInclusionNode

register = Library()


@template_tag(register, 'my_tag')
class MyTagNode(BaseNode):
    max_args = 1
    allowed_kwargs = ('keyword1', 'keyword2',)
    end_tag_name = 'end_my_tag'

    def render_tag(self, context, *tag_args, **tag_kwargs):
        # Render contents inside
        return self.nodelist.render(context)

Custom parsing

With the standard Node class from Django, it’s easier to implement custom syntax. For example, to parse:

{% getfirstof val1 val2 as val3 %}

use:

from django.template import Library, Node, TemplateSyntaxError
from tag_parser import template_tag, parse_token_kwargs, parse_as_var

register = Library()


@template_tag(register, 'getfirstof')
class GetFirstOfNode(Node):
    def __init__(self, options, as_var):
        self.options = options    # list of FilterExpression nodes.
        self.as_var = as_var

    @classmethod
    def parse(cls, parser, token):
        bits, as_var = parse_as_var(parser, token)
        tag_name, options, _ = parse_token_kwargs(parser, bits, allowed_kwargs=())

        if as_var is None or not choices:
            raise TemplateSyntaxError("Expected syntax: {{% {0} val1 val2 as val %}}".format(tag_name))

        return cls(options, as_var)

    def render(self, context):
        value = None
        for filterexpr in self.options:
            # The ignore_failures argument prevents that the value becomes TEMPLATE_STRING_IF_INVALID.
            value = filterexpr.resolve(context, ignore_failures=True)
            if value is not None:
                break

        context[self.as_var] = value
        return ''

Contributing

This module is designed to be generic. In case there is anything you didn’t like about it, or think it’s not flexible enough, please let us know. We’d love to improve it!

If you have any other valuable contribution, suggestion or idea, please let us know as well because we will look into it. Pull requests are welcome too. :-)