Take this input.db file:

<chapter xml:id="_responsibility">
<simpara>See regular link: <xref linkend="SEC-Definitions"/>.</simpara>
<simpara>See link with endterm: <xref linkend="SEC-Definitions" endterm="link"/>.</simpara>
</chapter>
<chapter xml:id="SEC-Definitions">
<title>Definitions</title>
<simpara>Some definitions.</simpara>
</chapter>

Convert it to output.docx:

It would be cool if it would remember settings I gave it, so I can have the same settings for multiple files (or even support multi file processing). ATM I need to remember which settings I have set and it is hard to adjust width/height to exactly that dimensions as before.

Here's an interesting one...

<a href="#linktarget">Link</a> works.

<a href="#linktarget(thing)">Link</a> doesn't work.

Adding brackets into the link target breaks the link in the PDF (works in Chrome).

As per angularsen/UnitsNet#328 (comment)

Having a set of tests to validate that the correct abbreviation notations and symbols are used would be helpful, as we have seen from before that discrepancies tend to creep in over time.

See #328 for the (currently work in progress) guideline, go through the comments and identify what tests to write for the guid

Right now, we're doing a lot of this manually, and flattening the package-lock.json file. It'll save a lot (and likely protect against future breakage) to use this instead:

https://www.npmjs.com/package/npm-logical-tree

• Guidance on adding a linter new rule-action (if none of the existing ones can perform the action, or be extended)
• Expand on guidelines re: keeping commits to logical areas of the code / individual packages where possible
• Use of conventional-commit standard going forward
• Advice not to include any package-lock.json changes PRs, for security reasons. File will be regene