use std::ops::Range; use std::sync::Arc; use crate::source_map::SourceMap; use crate::{BytePos, Pos, RelativeBytePos, SourceFile, SpanData}; /// A `SourceMap` wrapper that caches info about a single recent code position. This gives a good /// speedup when hashing spans, because often multiple spans within a single line are hashed in /// succession, and this avoids expensive `SourceMap` lookups each time the cache is hit. We used /// to cache multiple code positions, but caching a single position ended up being simpler and /// faster. #[derive(Clone)] pub struct CachingSourceMapView<'sm> { source_map: &'sm SourceMap, file: Arc, // The line's byte position range in the `SourceMap`. This range will fail to contain a valid // position in certain edge cases. Spans often start/end one past something, and when that // something is the last character of a file (this can happen when a file doesn't end in a // newline, for example), we'd still like for the position to be considered within the last // line. However, it isn't according to the exclusive upper bound of this range. We cannot // change the upper bound to be inclusive, because for most lines, the upper bound is the same // as the lower bound of the next line, so there would be an ambiguity. // // Since the containment aspect of this range is only used to see whether or not the cache // entry contains a position, the only ramification of the above is that we will get cache // misses for these rare positions. A line lookup for the position via `SourceMap::lookup_line` // after a cache miss will produce the last line number, as desired. line_bounds: Range, line_number: usize, } impl<'sm> CachingSourceMapView<'sm> { pub fn new(source_map: &'sm SourceMap) -> CachingSourceMapView<'sm> { let files = source_map.files(); let first_file = Arc::clone(&files[0]); CachingSourceMapView { source_map, file: first_file, line_bounds: BytePos(0)..BytePos(0), line_number: 0, } } #[inline] fn pos_to_line(&self, pos: BytePos) -> (Range, usize) { let pos = self.file.relative_position(pos); let line_index = self.file.lookup_line(pos).unwrap(); let line_bounds = self.file.line_bounds(line_index); let line_number = line_index + 1; (line_bounds, line_number) } #[inline] fn update(&mut self, new_file: Option>, pos: BytePos) { if let Some(file) = new_file { self.file = file; } (self.line_bounds, self.line_number) = self.pos_to_line(pos); } pub fn byte_pos_to_line_and_col( &mut self, pos: BytePos, ) -> Option<(Arc, usize, RelativeBytePos)> { if self.line_bounds.contains(&pos) { // Cache hit: do nothing. } else { // Cache miss. If the entry doesn't point to the correct file, get the new file and // index. let new_file = if !file_contains(&self.file, pos) { Some(self.file_for_position(pos)?) } else { None }; self.update(new_file, pos); }; let col = RelativeBytePos(pos.to_u32() - self.line_bounds.start.to_u32()); Some((Arc::clone(&self.file), self.line_number, col)) } pub fn span_data_to_lines_and_cols( &mut self, span_data: &SpanData, ) -> Option<(&SourceFile, usize, BytePos, usize, BytePos)> { let lo_hit = self.line_bounds.contains(&span_data.lo); let hi_hit = self.line_bounds.contains(&span_data.hi); if lo_hit && hi_hit { // span_data.lo and span_data.hi are cached (i.e. both in the same line). return Some(( &self.file, self.line_number, span_data.lo - self.line_bounds.start, self.line_number, span_data.hi - self.line_bounds.start, )); } // If the cached file is wrong, update it. Return early if the span lo and hi are in // different files. let new_file = if !file_contains(&self.file, span_data.lo) { let new_file = self.file_for_position(span_data.lo)?; if !file_contains(&new_file, span_data.hi) { return None; } Some(new_file) } else { if !file_contains(&self.file, span_data.hi) { return None; } None }; // If we reach here, lo and hi are in the same file. if !lo_hit { // We cache the lo information. self.update(new_file, span_data.lo); } let lo_line_bounds = &self.line_bounds; let lo_line_number = self.line_number.clone(); let (hi_line_bounds, hi_line_number) = if !self.line_bounds.contains(&span_data.hi) { // hi and lo are in different lines. We compute but don't cache the hi information. self.pos_to_line(span_data.hi) } else { // hi and lo are in the same line. (self.line_bounds.clone(), self.line_number) }; // Span lo and hi may equal line end when last line doesn't // end in newline, hence the inclusive upper bounds below. assert!(span_data.lo >= lo_line_bounds.start); assert!(span_data.lo <= lo_line_bounds.end); assert!(span_data.hi >= hi_line_bounds.start); assert!(span_data.hi <= hi_line_bounds.end); assert!(self.file.contains(span_data.lo)); assert!(self.file.contains(span_data.hi)); Some(( &self.file, lo_line_number, span_data.lo - lo_line_bounds.start, hi_line_number, span_data.hi, )) } fn file_for_position(&self, pos: BytePos) -> Option> { if !self.source_map.files().is_empty() { let file_idx = self.source_map.lookup_source_file_idx(pos); let file = &self.source_map.files()[file_idx]; if file_contains(file, pos) { return Some(Arc::clone(file)); } } None } } #[inline] fn file_contains(file: &SourceFile, pos: BytePos) -> bool { // `SourceMap::lookup_source_file_idx` and `SourceFile::contains` both consider the position // one past the end of a file to belong to it. Normally, that's what we want. But for the // purposes of converting a byte position to a line and column number, we can't come up with a // line and column number if the file is empty, because an empty file doesn't contain any // lines. So for our purposes, we don't consider empty files to contain any byte position. file.contains(pos) && !file.is_empty() }