@robbiecarlton the biggest thing I'm worried about with this PR is that I kinda just cargo-culted the ownership/borrowing/consuming thing until rustc shut up. Not sure if what I did was good practice. And I also don't know about the iterator chain I used for filtering the items -- feels like there ought to be some more elegant way of filtering and failing on the first error, but I tried a few ideas and then gave up. Currently it throws out runtime handlebars errors, but it does test the template once with the first item in the list to make sure the passed-in filter predicate actually works. (The way it works is that it just creates a tiny template, {{#if <condition>}}true{{else}}false{{/if}} and then passes each item to that template in turn.)

@pdaoust A few things I'd like to do to get this merged

• Please can the PR target the develop branch and I'll take care of merging it to release series branches
• Could you please give me some steps to reproduce the problem you're fixing so I can do my own manual testing?
• Are you up for adding some unit tests where appropriate for this? I'm not sure what's already here so it might not be super straight forward. I'm happy to either work on this with you or take it on myself based on some manual testing

I'm going to go ahead and re-target this to develop because the current target branch is going to be renamed

Okay, this is now ready for review and can be merged when ready

@ThetaSinner really sorry I missed all your messages from November; I don't have a good way of getting notifications from GitHub. @c12i thanks for the review; I'm really eager to get feedback on the aesthetics/idiomatic-ness of my code. Rust is very new to me, and I'd prefer iterators (coming from C# and JS) but don't always know how to do them in Rust.

I'll merge from develop and take a look at all the review stuff; thanks all!

Okay, I've tidied up all the code per your feedback @c12i and I've also merged in develop -- I noticed that the helpers code was rearranged, so I moved the new filter helper into a similar structure (it gets its own file, figure it was sufficiently complex).

I'm pretty new to writing Rust, but I think I could come up with some tests that at the very least test the initial problem that this thing was trying to fix. How far do you think I should go -- wondering if it makes sense to test all possible falsey HBS values and a few representative truthy values, or if that's just overkill.

I noticed that the helpers code was rearranged, so I moved the new filter helper into a similar structure (it gets its own file, figure it was sufficiently complex).

Ah yes, should have mentioned this earlier, glad you made the change

How far do you think I should go -- wondering if it makes sense to test all possible falsey HBS values and a few representative truthy values, or if that's just overkill.

No worries @pdaoust, we can test the happy path cases for now

@c12i could you review again? I simplified some code -- realised that a couple suggestions above tackled one problem in two different ways and cancelled each other out :) I also wrote a few basic tests and discovered/fixed a bug.

realised that a couple suggestions above tackled one problem in two different ways and cancelled each other out

thanks for the changes @pdaoust, the implementation is much simpler now