Pages detail API should allow fetching a page by path

Currently the API endpoints provided by wagtail.api.v2.endpoints.PagesAPIEndpoint do not provide the ability to fetch a Page by the page’s URL.

The docs regarding filtering pages makes it seem like a request such as

curl http://mysite.com/api/pages/?html_url=<some_url>

should work but instead the API returns:

{
    "message": "query parameter is not an operation or a recognised field: html_url"
}

Would it be possible to support filtering on this field? Perhaps also making the detail view able to return the page by it’s path? i.e. allowing all of:

# filtering
curl http://mysite.com/api/pages/?html_url=<my_page_url>

# detail
curl http://mysite.com/api/pages/1
curl http://mysite.com/api/pages/my_page_path

If this is something that wagtail would like to add then I may be able to try submitting a PR.

The motivation for this feature is to make it easy for a frontend webapp to fetch pages and render them. It would be useful for SPA apps if they could integrate with a wagtail CMS backend by using the current pages URL to do a wagtail API request and use the JSON response to render a page using whatever framework the SPA is written with (react/angular/vue etc.).

[Place1]

It’s awesome to see this issue has had some attention. I just had a look at #4195 and i’m wondering if that view returning a redirect will result in unnecessary round trips for API clients who just want to GET a page by it’s path. I know for my usecase it likely will. Would a permanent redirect help? Could the view just use Django’s reverse() to skip the redirect alltogether? @kaedroho

[Place1]

After diving into the code a bit (i’m not familiar with it) I realize that achieving this through filtering probably isn’t feasible/practical for the API, but allowing individual pages to be retrieved by their path is possible.

i’ve hacked a little bit locally and this has some of my ideas in it:

class PagesAPIEndpoint(BaseAPIEndpoint):

    # ...

    def get_object(self):
        lookup = self.kwargs[self.lookup_field]
        if not re.match('^\d+$', lookup):
            return self.get_page_at_path(lookup)
        base = super(PagesAPIEndpoint, self).get_object()
        return base.specific

    def get_page_at_path(self, path):
        path_components = [component for component in path.split('/') if component]
        try:
            return self.request.site.root_page.specific.route(self.request, path_components).page.specific
        except Http404:
            raise Http404('No %s matches the given query.' % self.model._meta.object_name)

This would allow API clients to do the following:

curl http://mysite/api/pages/my/full/page/path

but there is no way to fetch the root page

curl http://mysite/api/pages/  # <-- this is the root path but also the List view :(

So this isn’t the solution but hopefully others have some better ideas?