Get the path to the currently processed Markdown file

[SnDream]

Hello!
I'm trying to make my own customized version of the auth_attach plugin made by GiovanH. I want to implement rules like this:

1. when Markdown's relative link points to a file in the STATIC path, it automatically adds {static} before the path because the file will be in a defined path.
2. when Markdown's relative link points to a file outside the CONTENT path, it automatically adds {attach} before the path because the file will not be in the compiled output.
3. otherwise, it automatically adds {filename} before the path.

Since relative linking is based on the path to the current Markdown file, this requires that you get the path to the Markdown file itself, but I found this to be more difficult than I thought.

Markdown itself doesn't provide the path, and from the code it looks like the path is already abstracted during Markdown initialization.

My current implementation uses the path to be recorded in the content_object_init signal, but after testing, I found that content_object_init comes after Markdown processes the file.

I'm currently trying to replace it with article_generator_preread , which is supposed to be a signal before the Markdown file is read, but here's another problem:

        """Return a content object parsed with the given format."""

        path = os.path.abspath(os.path.join(base_path, path))
        source_path = posixize_path(os.path.relpath(path, base_path))
        logger.debug("Read file %s -> %s", source_path, content_class.__name__)

        if not fmt:
            _, ext = os.path.splitext(os.path.basename(path))
            fmt = ext[1:]

        if fmt not in self.readers:
            raise TypeError("Pelican does not know how to parse %s", path)

        if preread_signal:
            logger.debug("Signal %s.send(%s)", preread_signal.name, preread_sender)
            preread_signal.send(preread_sender)

source_path is not passed as a parameter to article_generator_preread and I can't get that.

Is there a suitable way to solve this problem that would allow me to get the path of the currently processed Markdown during Markdown processing?

Also, I'd like to know why pelican can't just automatically {static}/{attach}/{filename} markup based on the rules above? Is there a scenario I'm not expecting? Thanks.

[SnDream]

UPDATE: I'm currently getting the filename in article_generator_preread, which is basically doing the iteration of the file all over again, and I'm not sure if there's a better way to do it.
If there is any better way of doing this, I would appreciate it, thanks!

SnDream/sndream.github.io@80005d3