ok/jj
1
0
Fork 0
forked from mirrors/jj
Code Pull requests Activity

diff: omit construction of count-to-words map for right-side histogram

Browse source 
This also allows us to borrow Vec<WordPositions> from &self.
This commit is contained in:
Yuya Nishihara 2024-09-26 16:50:16 +09:00
parent 493f610fd5
commit 5c52b4ec13
1 changed files with 13 additions and 13 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

26
lib/src/diff.rs
View file

@ -103,13 +103,12 @@ impl<'input, 'aux> DiffSource<'input, 'aux> {
} }
} }
struct Histogram<'a> { struct Histogram<'input> {
word_to_positions: HashMap<&'a BStr, Vec<WordPosition>>, word_to_positions: HashMap<&'input BStr, Vec<WordPosition>>,
count_to_words: BTreeMap<usize, Vec<&'a BStr>>,
} }
impl Histogram<'_> { impl<'input> Histogram<'input> {
fn calculate<'a>(source: &DiffSource<'a, '_>, max_occurrences: usize) -> Histogram<'a> { fn calculate(source: &DiffSource<'input, '_>, max_occurrences: usize) -> Self {
let mut word_to_positions: HashMap<&BStr, Vec<WordPosition>> = HashMap::new(); let mut word_to_positions: HashMap<&BStr, Vec<WordPosition>> = HashMap::new();
for (i, range) in source.ranges.iter().enumerate() { for (i, range) in source.ranges.iter().enumerate() {
let word = &source.text[range.clone()]; let word = &source.text[range.clone()];
@ -120,14 +119,15 @@ impl Histogram<'_> {
positions.push(WordPosition(i)); positions.push(WordPosition(i));
} }
} }
Histogram { word_to_positions }
}
fn build_count_to_words(&self) -> BTreeMap<usize, Vec<&'input BStr>> {
let mut count_to_words: BTreeMap<usize, Vec<&BStr>> = BTreeMap::new(); let mut count_to_words: BTreeMap<usize, Vec<&BStr>> = BTreeMap::new();
for (word, ranges) in &word_to_positions { for (word, ranges) in &self.word_to_positions {
count_to_words.entry(ranges.len()).or_default().push(word); count_to_words.entry(ranges.len()).or_default().push(word);
} }
Histogram { count_to_words
word_to_positions,
count_to_words,
}
} }
} }
@ -233,7 +233,8 @@ fn unchanged_ranges_lcs(
) -> Vec<(Range<usize>, Range<usize>)> { ) -> Vec<(Range<usize>, Range<usize>)> {
let max_occurrences = 100; let max_occurrences = 100;
let left_histogram = Histogram::calculate(left, max_occurrences); let left_histogram = Histogram::calculate(left, max_occurrences);
if *left_histogram.count_to_words.keys().next().unwrap() > max_occurrences { let left_count_to_words = left_histogram.build_count_to_words();
if *left_count_to_words.keys().next().unwrap() > max_occurrences {
// If there are very many occurrences of all words, then we just give up. // If there are very many occurrences of all words, then we just give up.
return vec![]; return vec![];
} }
@ -241,8 +242,7 @@ fn unchanged_ranges_lcs(
// Look for words with few occurrences in `left` (could equally well have picked // Look for words with few occurrences in `left` (could equally well have picked
// `right`?). If any of them also occur in `right`, then we add the words to // `right`?). If any of them also occur in `right`, then we add the words to
// the LCS. // the LCS.
let Some(uncommon_shared_words) = left_histogram let Some(uncommon_shared_words) = left_count_to_words
.count_to_words
.iter() .iter()
.map(|(left_count, left_words)| -> Vec<&BStr> { .map(|(left_count, left_words)| -> Vec<&BStr> {
left_words left_words