instant-labs
/
instant-segment
mirror of https://github.com/instant-labs/instant-segment.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Use range for previous argument as well

This commit is contained in:
Dirkjan Ochtman 2020-11-26 11:07:14 +01:00
parent bc20e39c1e
commit ae3896b47b
1 changed files with 9 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
src/lib.rs
View File

@ -116,7 +116,12 @@ impl<'a> SegmentState<'a> {
} }
/// Score `word` in the context of `previous` word /// Score `word` in the context of `previous` word
fn search(&mut self, level: usize, range: Range<usize>, previous: Option<&str>) -> (f64, bool) { fn search(
&mut self,
level: usize,
range: Range<usize>,
previous: Option<Range<usize>>,
) -> (f64, bool) {
if range.is_empty() { if range.is_empty() {
return (0.0, false); return (0.0, false);
} }
@ -124,15 +129,15 @@ impl<'a> SegmentState<'a> {
let mut best = f64::MIN; let mut best = f64::MIN;
for split in 1..(range.len().min(self.data.limit) + 1) { for split in 1..(range.len().min(self.data.limit) + 1) {
let (start, split, end) = (range.start, range.start + split, range.end); let (start, split, end) = (range.start, range.start + split, range.end);
let prefix = &self.text[start..split]; let previous = previous.clone().map(|range| &self.text[range]);
let prefix_score = self.data.score(prefix, previous).log10(); let prefix_score = self.data.score(&self.text[start..split], previous).log10();
let pair = (split..end, start..split); let pair = (split..end, start..split);
let (suffix_score, suffix_splits) = match self.memo.get(&pair) { let (suffix_score, suffix_splits) = match self.memo.get(&pair) {
Some((score, splits)) => (*score, &self.split_cache[splits.start..splits.end]), Some((score, splits)) => (*score, &self.split_cache[splits.start..splits.end]),
None => { None => {
let (suffix_score, has_splits) = let (suffix_score, has_splits) =
self.search(level + 1, split..end, Some(prefix)); self.search(level + 1, split..end, Some(start..split));
let start = self.split_cache.len(); let start = self.split_cache.len();
self.split_cache.extend(if has_splits { self.split_cache.extend(if has_splits {
&self.best[level + 1][..] &self.best[level + 1][..]