feat: add bullet list element parsing and rendering #111
Conversation
peek now also considers matching functions.
mhatzl
added
the
waiting-on-reviewer
There was a problem hiding this comment.
Awesome!
I think we now have a fairly decent parsing framework. Looks like it's pretty flexible and not that hard to work with, and that makes me really happy.
As for the PR itself, I left some comments. Many of them are not directly related to this PR. You can decide for yourself, whether we should tackle this in this PR.
I am almost leaning towards merging this and then tackling the comments in other PRs so that we prevent explosion in number of changes (is already 1+k LOC large, would not like to inflate it even more). So, I will just leave comments for now (neither approved nor changes requested) so we can discuss first how should we go forward.
| fn next(&mut self) -> Option<Self::Item> { | ||
| if self.end_reached() { | ||
| if self.prefix_mismatch || self.end_reached() { | ||
| return None; | ||
| } | ||
|
|
||
| if let Some(end_fn) = self.end_match.clone() { | ||
| self.next_matching = true; | ||
|
|
||
| if (end_fn)(self) { | ||
| self.iter_end = true; | ||
| self.next_matching = false; |
There was a problem hiding this comment.
It's somewhat hard for me to follow the code in this whole function in general. Could you add some comments in this function explaining the logic, and/or updating documentation on next_matching and similar fields to better explain how are they used?
There was a problem hiding this comment.
I updated the doc on next_matching. Hopefully this makes this code clearer.
In short: Setting next_matching flag in next() prevents matching in peeking_next() during execution of the matching functions in next(). Matching functions must get the symbols as is, because they define what symbols must be skipped.
nfejzic
removed
the
waiting-on-reviewer
List of issues that this PR closes
closes #92
Relevant decisions you made in this PR
Removed
BlanklinesymbolThe
Blanklinesymbol was used to indicate two consecutive newlines.This created problems with prefix matching, because prefix is only matched after newlines.
As a better alternative, the
SymbolIteratoroffersis_empty_line()andconsumed_is_empty_line().Those functions also handle lines only consisting of whitespace symbols.
Adapt
peeking_next()to consider matching functionsWith this change,
peeking_next()no returns the same symbols asnext().This is important for nested elements, because matching functions internally use
peeking_next().Meaning that matching functions of nested elements would not be correct if
peeking_next()would not consider the matching functions of outer iterators.