Modify the Levenshtein-based suggestions to include imports

This commit is contained in:
Ravi Shankar
2015-12-14 22:36:31 +05:30
parent 35b6461b6e
commit 51ff171948
6 changed files with 95 additions and 95 deletions

View File

@@ -24,7 +24,7 @@ use parse::token;
use parse::token::{InternedString, intern, str_to_ident};
use ptr::P;
use util::small_vector::SmallVector;
use util::lev_distance::{lev_distance, max_suggestion_distance};
use util::lev_distance::find_best_match_for_name;
use ext::mtwt;
use fold::Folder;
@@ -780,15 +780,8 @@ impl<'a> ExtCtxt<'a> {
}
pub fn suggest_macro_name(&mut self, name: &str, span: Span) {
let mut min: Option<(Name, usize)> = None;
let max_dist = max_suggestion_distance(name);
for macro_name in self.syntax_env.names.iter() {
let dist = lev_distance(name, &macro_name.as_str());
if dist <= max_dist && (min.is_none() || min.unwrap().1 > dist) {
min = Some((*macro_name, dist));
}
}
if let Some((suggestion, _)) = min {
let names = &self.syntax_env.names;
if let Some(suggestion) = find_best_match_for_name(names.iter(), name, None) {
self.fileline_help(span, &format!("did you mean `{}!`?", suggestion));
}
}

View File

@@ -8,50 +8,64 @@
// option. This file may not be copied, modified, or distributed
// except according to those terms.
use ast::Name;
use std::cmp;
use parse::token::InternedString;
pub fn lev_distance(me: &str, t: &str) -> usize {
if me.is_empty() { return t.chars().count(); }
if t.is_empty() { return me.chars().count(); }
/// To find the Levenshtein distance between two strings
pub fn lev_distance(a: &str, b: &str) -> usize {
// cases which don't require further computation
if a.is_empty() {
return b.chars().count();
} else if b.is_empty() {
return a.chars().count();
}
let mut dcol: Vec<_> = (0..t.len() + 1).collect();
let mut dcol: Vec<_> = (0..b.len() + 1).collect();
let mut t_last = 0;
for (i, sc) in me.chars().enumerate() {
for (i, sc) in a.chars().enumerate() {
let mut current = i;
dcol[0] = current + 1;
for (j, tc) in t.chars().enumerate() {
for (j, tc) in b.chars().enumerate() {
let next = dcol[j + 1];
if sc == tc {
dcol[j + 1] = current;
} else {
dcol[j + 1] = cmp::min(current, next);
dcol[j + 1] = cmp::min(dcol[j + 1], dcol[j]) + 1;
}
current = next;
t_last = j;
}
}
dcol[t_last + 1]
} dcol[t_last + 1]
}
pub fn max_suggestion_distance(name: &str) -> usize {
use std::cmp::max;
// As a loose rule to avoid obviously incorrect suggestions, clamp the
// maximum edit distance we will accept for a suggestion to one third of
// the typo'd name's length.
max(name.len(), 3) / 3
/// To find the best match for a given string from an iterator of names
/// As a loose rule to avoid the obviously incorrect suggestions, it takes
/// an optional limit for the maximum allowable edit distance, which defaults
/// to one-third of the given word
pub fn find_best_match_for_name<'a, T>(iter_names: T,
lookup: &str,
dist: Option<usize>) -> Option<InternedString>
where T: Iterator<Item = &'a Name> {
let max_dist = dist.map_or_else(|| cmp::max(lookup.len(), 3) / 3, |d| d);
iter_names
.filter_map(|name| {
let dist = lev_distance(lookup, &name.as_str());
match dist <= max_dist { // filter the unwanted cases
true => Some((name.as_str(), dist)),
false => None,
}
})
.min_by_key(|&(_, val)| val) // extract the tuple containing the minimum edit distance
.map(|(s, _)| s) // and return only the string
}
#[test]
fn test_lev_distance() {
use std::char::{ from_u32, MAX };
use std::char::{from_u32, MAX};
// Test bytelength agnosticity
for c in (0..MAX as u32)
.filter_map(|i| from_u32(i))