Parse &str instead of &[u8]
#541
Conversation
|
Should I turn the third commit https://github.com/djc/askama/pull/541/commits/edc3cf80d6708800322f716b90ce1ae3a939f6c0 into a new PR? |
|
No, the third commit can stay in this PR. So I guess my feeling is that a lot of the substantial changes to |
|
So this is kind of the other way around from what I intended. I'm actually more interested in the (If that is harder than I'm making it out to be or if you're just not interested in doing the work, that's fine too.) |
|
Iterating strings is just no fun in Rust. One little mistake and you are splitting a codepoint, which causes a panic. That's why I wanted to let nom do the error prone stuff instead of porting the &[u8] implementation directly. Also I'm not 100% sure the current implementation in main is 100% correct. E.g.
|
askama_shared/src/parser.rs
Outdated
| /// Skips over the any amout of `inner` until `end` was found. | ||
| /// Returns a tuple of the skipped string and the found end marker. | ||
| fn skip_till<'a, I, O>( | ||
| mut inner: impl FnMut(&'a [u8]) -> IResult<&'a [u8], I>, |
There was a problem hiding this comment.
I think take_content() is now the only caller for skip_till, which means inner is always anychar. Can we simplify? Maybe even inline skip_till() into take_content()?
There was a problem hiding this comment.
The function is used in #546, too.
I removed the "inner" parameter, because it's "anychar" for both its uses.
I cleaned up the implementation a bit. I think it's much easier on the eyes now. :)
|
@Kijewski do you have a moment to work on this soon? Otherwise I'll probably take it over, I think it's the last blocker to 0.11. |
Askama's takes valid UTF-8 files as input. So why operate on byte slices instead of strings? This makes writing some functions a lot simpler.
Askama's takes valid UTF-8 files as input. So why operate on byte slices
instead of strings? This makes writing some functions a lot simpler.