✨ NEW: Add SyntaxTreeNode

[hukkin]

Context: read the thread here hukkin/mdformat#132 (comment)

[codecov]

Codecov Report

Merging #129 (a6d5c60) into master (11eb374) will decrease coverage by 0.22%.
The diff coverage is 91.30%.

@@ Coverage Diff @@ ## master #129 +/- ## ========================================== - Coverage 96.42% 96.19% -0.23%  ========================================== Files 71 72 +1 Lines 2966 3104 +138 ========================================== + Hits 2860 2986 +126  - Misses 106 118 +12 

Flag	Coverage Δ
pytests	96.19% <91.30%> (-0.23%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
markdown_it/utils.py	94.11% <75.00%> (-2.55%)	⬇️
markdown_it/tree.py	91.66% <91.66%> (ø)
markdown_it/token.py	91.95% <100.00%> (+0.18%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 11eb374...c96b2cb. Read the comment docs.

[chrisjsewell]

Thanks firstly can you move this to a separate file: tree.py and also note in the module that this is not part of upstream markdown-it
Then also add a deprecation warning inside nest_tokens

[hukkin]

@chrisjsewell 👍 done

[chrisjsewell]

thanks, next can you add docstrings for all the methods and were we going to have is_nested and is_root properties?

[hukkin]

thanks, next can you add docstrings for all the methods

Sure thing.

and were we going to have is_nested and is_root properties?

What do you think? I realized I don't have much need for these.

is_root basically equals not self.parent and that can be useful when traversing towards root and checking if you're already there. But in that context not self.parent feels more intuitive, because if you're not there yet, what you do next is self.parent.parent...

is_nested I never need in mdformat. That would be equal to bool(self.nester_tokens) which isn't overly onerous to do. And if the use case for is_nested is checking before accesing self.nester_tokens then bool(self.nester_tokens) could be more intuitive?

So basically just didn't add these because they weren't useful to me, but have no problem with adding them.

[chrisjsewell]

yeh I think at least add the is_nested

[hukkin]

We could enforce __init__ privateness (in favor of from_tokens) by something like adding a private mandatory kwarg:

def __init__(self, __internal_call=False): if not __internal_call: raise Exception("this is private API")

just an idea, not sure if a good one 😄

[chrisjsewell]

just an idea, not sure if a good one

yeh not necessarily a bad idea lol, but probably overkill

[hukkin]

Added docstrings and is_nested. Didn't add docstrings for the @propertyies that are just a passthrough from a Token. Do you think we need? There is a comment that kind of explains them as a whole.

[chrisjsewell]

dn't add docstrings for the @propertyies that are just a passthrough from a Token. Do you think we need?

I would say yes (basically copied from the comments in Token), because they are helpful when e.g. working in editors like VS Code where you get the description when you hover over them

[hukkin]

Yep, makes sense. Copied those from Token class.

[chrisjsewell]

lastly, can you add a few extra bits to the test module, to increase the test coverage

[hukkin]

Added. Tell me if you think there's something else in particular that you want tested.

[chrisjsewell]

nah thats great cheers