Add ByteRanges #26
Add ByteRanges #26
Conversation
|
Some unnamed jerk just merged some code that causes conflicts here ;-) |
| //! | ||
| //! fn main() { | ||
| //! let raw_data = [0x00, 0x01, 0x10, 0xFF, 0x42]; | ||
| //! // Search for the first byte that's either 0x70...0x7F, 0xFD, 0xFE, or 0xFF |
There was a problem hiding this comment.
I'd say the nomenclature should be the same here:
0x70...0x7F, or 0xFD...0xFF
I then wonder if it should be Rust-standard:
0x70..=0x7F, or 0xFD..=0xFF
I then wonder if we can use an actual RangeInclusive instead of a tuple
| T: PartialOrd, | ||
| { | ||
| fn find_any_in_ranges(&self, ranges: &[(T, T)]) -> Option<usize> { | ||
| self.iter().position(|c| ranges.iter().position(|r| r.0 <= *c && *c <= r.1).is_some()) |
| let mut ranges: [(u8, u8); 8] = pairs; | ||
| for (ref lo, hi) in ranges.iter_mut() { | ||
| let extra = *hi; | ||
| *hi = if extra > 255_u8 - lo { 255_u8 } else { lo + extra }; |
There was a problem hiding this comment.
this might be clearer as either of these pseudocode:
-
let mut x = [a, b]; x.sort(); let [a, b] = x;
-
any::<u8>.prop_map(|a| (a, range(a, u8::max)))
There was a problem hiding this comment.
or if a > b { mem::swap(a, b) }
|
What is required for this PR to move forward? I could really use range search for some ASCII character class checks. |
|
I think it would be acceptable for you to grab this PR, make the changes, and then submit a new PR... |
|
I played with this PR as-is - it works, but found out that SIMD version it actually quite slower than regular comparisons, so now I'm not even sure if it's worth it :/ |
This adds support for finding bytes in a set of ranges of bytes.
I hereby license this contribution under the dual MIT/Apache-2.0 license, allowing licensees to choose either at their option.