Skip to content

Panflute: Part 2

One of the pain points of translating a document is figuring out what to do with hyperlinks. Here's how we deal with it:

We have a prepare() method in our panflute filter that is run before the document filter is applied, and a finalize() method that is run after the document filter is applied.

The prepare method initializes an empty list to hold links. As we parse each section of the document, we look for links, and when we find a link, we add the link to a list.

When the text is translated, links are stripped out and the original link text becomes plain text again. Rather than try and determine which words in the translation map to the original link text, we simply aggregate the links at the bottom of the document. Each link includes a translation of the original link text that is a link to the original link.

Panflute Filter

Let's cover a bit more of the panflute magic that makes all of that happen.

This relies on two additional methods for the filter: prepare() and finalize().

The prepare method just creates an empty list that will be used to store all of the link elements in the document:

def prepare(doc):
    doc.linklist = []

We call the elem.walk() method on each element of the document, so we have a chance to see each element and determine if it is a link. If it is, we keep it simple and just save the entire element:

def strip_links(elem,doc):
    Each link will be stripped in the translation process.
    Save them for the end.
    if isinstance(elem,Link):