Auto merge of #145494 - cjgillot:span-decode-once, r=lqd

Only unpack span data once to compute end_point and next_point.

Split from https://github.com/rust-lang/rust/pull/144930
This commit is contained in:
bors 2025-08-22 21:43:06 +00:00
commit 6545b05638
2 changed files with 9 additions and 10 deletions

View File

@ -911,12 +911,13 @@ impl SourceMap {
/// Returns a new span representing just the last character of this span. /// Returns a new span representing just the last character of this span.
pub fn end_point(&self, sp: Span) -> Span { pub fn end_point(&self, sp: Span) -> Span {
let pos = sp.hi().0; let sp = sp.data();
let pos = sp.hi.0;
let width = self.find_width_of_character_at_span(sp, false); let width = self.find_width_of_character_at_span(sp, false);
let corrected_end_position = pos.checked_sub(width).unwrap_or(pos); let corrected_end_position = pos.checked_sub(width).unwrap_or(pos);
let end_point = BytePos(cmp::max(corrected_end_position, sp.lo().0)); let end_point = BytePos(cmp::max(corrected_end_position, sp.lo.0));
sp.with_lo(end_point) sp.with_lo(end_point)
} }
@ -930,8 +931,9 @@ impl SourceMap {
if sp.is_dummy() { if sp.is_dummy() {
return sp; return sp;
} }
let start_of_next_point = sp.hi().0;
let sp = sp.data();
let start_of_next_point = sp.hi.0;
let width = self.find_width_of_character_at_span(sp, true); let width = self.find_width_of_character_at_span(sp, true);
// If the width is 1, then the next span should only contain the next char besides current ending. // If the width is 1, then the next span should only contain the next char besides current ending.
// However, in the case of a multibyte character, where the width != 1, the next span should // However, in the case of a multibyte character, where the width != 1, the next span should
@ -940,7 +942,7 @@ impl SourceMap {
start_of_next_point.checked_add(width).unwrap_or(start_of_next_point); start_of_next_point.checked_add(width).unwrap_or(start_of_next_point);
let end_of_next_point = BytePos(cmp::max(start_of_next_point + 1, end_of_next_point)); let end_of_next_point = BytePos(cmp::max(start_of_next_point + 1, end_of_next_point));
Span::new(BytePos(start_of_next_point), end_of_next_point, sp.ctxt(), None) Span::new(BytePos(start_of_next_point), end_of_next_point, sp.ctxt, None)
} }
/// Check whether span is followed by some specified expected string in limit scope /// Check whether span is followed by some specified expected string in limit scope
@ -963,9 +965,7 @@ impl SourceMap {
/// Finds the width of the character, either before or after the end of provided span, /// Finds the width of the character, either before or after the end of provided span,
/// depending on the `forwards` parameter. /// depending on the `forwards` parameter.
#[instrument(skip(self, sp))] #[instrument(skip(self, sp))]
fn find_width_of_character_at_span(&self, sp: Span, forwards: bool) -> u32 { fn find_width_of_character_at_span(&self, sp: SpanData, forwards: bool) -> u32 {
let sp = sp.data();
if sp.lo == sp.hi && !forwards { if sp.lo == sp.hi && !forwards {
debug!("early return empty span"); debug!("early return empty span");
return 1; return 1;

View File

@ -74,9 +74,8 @@ use crate::{BytePos, SPAN_TRACK, SpanData};
/// because `parent` isn't currently used by default. /// because `parent` isn't currently used by default.
/// ///
/// In order to reliably use parented spans in incremental compilation, /// In order to reliably use parented spans in incremental compilation,
/// the dependency to the parent definition's span. This is performed /// accesses to `lo` and `hi` must introduce a dependency to the parent definition's span.
/// using the callback `SPAN_TRACK` to access the query engine. /// This is performed using the callback `SPAN_TRACK` to access the query engine.
///
#[derive(Clone, Copy, Eq, PartialEq, Hash)] #[derive(Clone, Copy, Eq, PartialEq, Hash)]
#[rustc_pass_by_value] #[rustc_pass_by_value]
pub struct Span { pub struct Span {