Fix the codebase's use of chars & put range asserts on chars
#12086
Conversation
| } | ||
| bump(rdr); | ||
| accum_int *= 16; | ||
| accum_int += hex_digit_val(n); | ||
| i -= 1u; | ||
| } | ||
| if i != 0 && is_eof(rdr) { | ||
| fatal_span(rdr, start_bpos, rdr.last_pos.get(), | ||
| ~"unterminated numeric characer escape"); |
|
LGTM, r=me after squash |
Avoid using -1 as a char sentinel, when Option<char> is the perfect thing.
The transmute was unsound.
There are many instances of .unwrap_or('\x00') for "ignoring" EOF which
either do not make the situation worse than it was (well, actually make
it better, since it's easy to grep for places that don't handle EOF) or
can never ever be read.
Fixes rust-lang#8971.
A `char` is a Unicode codepoint, and so ranges from 0--0x10FFFF (with the surrogate gaps): we may as well inform LLVM of this.
Apparently loading them signed will break if/when they become i1.
The lexer and json were using `transmute(-1): char` as a sentinel value for EOF, which is invalid since `char` is strictly a unicode codepoint. Fixing this allows for range asserts on chars since they always lie between 0 and 0x10FFFF.
|
What’s a range assert and when is it used? Should it also exclude the 0xD800 to 0xDFFF range (surrogate code points), matching |
|
@SimonSapin A range assert tells LLVM that a Excluding the surrogate range is possible, but it seems unlikely that there will be much performance gained from just excluding an interior range, while telling LLVM that |
|
Perfect answer, thanks @huonw! |
infer from RPIT bounds of _this_ function
Collect obligations from RPITs (Return Position `impl Trait`) of a function which is being inferred.
This allows inferring {unknown}s from RPIT bounds.
Closes rust-lang#8403
Update copyright year for Clippy (2024 edition) Our license was outdated by a year (+ change in the main README.md) changelog:none
The lexer and json were using
transmute(-1): charas a sentinel value for EOF, which is invalid sincecharis strictly a unicode codepoint.Fixing this allows for range asserts on chars since they always lie between 0 and 0x10FFFF.