Release version 0.4 (#41)

* fix false positives for unicode substring matcher * add note to docs about using the high level nucleo crate * expose number of active injectors from matcher * fix clippy lint * release version 0.4
helix-editor · Feb 20, 2024 · 9e234de · 9e234de
1 parent c7893db
commit 9e234de
Show file tree

Hide file tree

Showing 11 changed files with 149 additions and 60 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,10 +1,15 @@
 # Changelog
 
-# [0.3.1] - 2023-12-22
+# [0.4.0] - 2023-12-22
+
+## Added
+
+* `active_injectors()` to retrieve the number of injectors that can potentially add new items to the matcher in the future.
 
 ## Bugfixes
 
 * fix Unicode substring matcher expecting an exact match (rejecting trailing characters)
+* fix crashes and false positives in unicode substring matcher
 
 # [0.3.0] - 2023-12-22
 

diff --git a/Cargo.lock b/Cargo.lock
diff --git a/Cargo.toml b/Cargo.toml
@@ -1,8 +1,8 @@
 [package]
 name = "nucleo"
 description = "plug and play high performance fuzzy matcher"
-authors = ["Pascal Kuthe <pascal.kuthe@semimod.de>"]
-version = "0.3.0"
+authors = ["Pascal Kuthe <pascalkuthe@pm.me>"]
+version = "0.4.0"
 edition = "2021"
 license = "MPL-2.0"
 repository = "https://github.com/helix-editor/nucleo"
@@ -12,7 +12,7 @@ exclude = ["/typos.toml", "/tarpaulin.toml"]
 [lib]
 
 [dependencies]
-nucleo-matcher = { version = "0.3.0", path = "matcher" }
+nucleo-matcher = { version = "0.3.1", path = "matcher" }
 parking_lot = { version = "0.12.1", features = ["send_guard", "arc_lock"]}
 rayon = "1.7.0"
 

diff --git a/bench/Cargo.toml b/bench/Cargo.toml
@@ -6,7 +6,7 @@ edition = "2021"
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
 
 [dependencies]
-nucleo = { version = "0.3", path = "../" }
+nucleo = { version = "*", path = "../" }
 brunch = "0.5.0"
 fuzzy-matcher = "0.3.7"
 walkdir = "2"
diff --git a/matcher/Cargo.toml b/matcher/Cargo.toml
@@ -1,8 +1,8 @@
 [package]
 name = "nucleo-matcher"
 description = "plug and play high performance fuzzy matcher"
-authors = ["Pascal Kuthe <pascal.kuthe@semimod.de>"]
-version = "0.3.0"
+authors = ["Pascal Kuthe <pascalkuthe@pm.me>"]
+version = "0.3.1"
 edition = "2021"
 license = "MPL-2.0"
 repository = "https://github.com/helix-editor/nucleo"

diff --git a/matcher/src/exact.rs b/matcher/src/exact.rs
@@ -236,7 +236,8 @@ impl Matcher {
             .checked_sub(1)
             .map(|i| haystack[i].char_class(&self.config))
             .unwrap_or(self.config.initial_char_class);
-        for (i, &c) in haystack[start..].iter().enumerate() {
+        let end = haystack.len() - needle.len();
+        for (i, &c) in haystack[start..end].iter().enumerate() {
             let (c, char_class) = c.char_class_and_normalize(&self.config);
             if c != needle[0] {
                 continue;
@@ -265,8 +266,8 @@ impl Matcher {
         let score = self.calculate_score::<INDICES, _, _>(
             haystack,
             needle,
-            max_pos,
-            max_pos + needle.len(),
+            start + max_pos,
+            start + max_pos + needle.len(),
             indices,
         );
         Some(score)

diff --git a/matcher/src/lib.rs b/matcher/src/lib.rs
@@ -2,6 +2,11 @@
 `nucleo_matcher` is a low level crate that contains the matcher implementation
 used by the high level `nucleo` crate.
 
+**NOTE**: If you are building an fzf-like interactive fuzzy finder that is
+meant to match a reasonably large number of items (> 100) using the high level
+`nucleo` crate is highly recommended. Using `nucleo-matcher` directly in you ui
+loop will be very slow. Implementing this logic yourself is very complex.
+
 The matcher is hightly optimized and can significantly outperform `fzf` and
 `skim` (the `fuzzy-matcher` crate). However some of these optimizations require
 a slightly less convenient API. Be sure to carefully read the documentation of

diff --git a/matcher/src/prefilter.rs b/matcher/src/prefilter.rs
@@ -78,6 +78,9 @@ impl Matcher {
             .position(|c| c.normalize(&self.config) == needle_char)?;
         let needle_char = needle.last();
         if only_greedy {
+            if haystack.len() - start < needle.len() {
+                return None;
+            }
             Some((start, start + 1))
         } else {
             let end = haystack.len()

diff --git a/matcher/src/tests.rs b/matcher/src/tests.rs
@@ -84,20 +84,17 @@ fn assert_matches(
     }
 }
 
-pub fn assert_not_matches(
+fn assert_not_matches_with(
     normalize: bool,
     case_sensitive: bool,
-    path: bool,
+    algorithm: &[Algorithm],
     cases: &[(&str, &str)],
 ) {
-    let mut config = Config {
+    let config = Config {
         normalize,
         ignore_case: !case_sensitive,
         ..Config::DEFAULT
     };
-    if path {
-        config.set_match_paths();
-    }
     let mut matcher = Matcher::new(config);
     let mut needle_buf = Vec::new();
     let mut haystack_buf = Vec::new();
@@ -110,31 +107,32 @@ pub fn assert_not_matches(
         let needle = Utf32Str::new(&needle, &mut needle_buf);
         let haystack = Utf32Str::new(haystack, &mut haystack_buf);
 
-        let res = matcher.fuzzy_match(haystack, needle);
-        assert_eq!(res, None, "{needle:?} should not match {haystack:?}");
-        let res = matcher.fuzzy_match_greedy(haystack, needle);
-        assert_eq!(
-            res, None,
-            "{needle:?} should not match {haystack:?} (greedy)"
-        );
-        let res = matcher.substring_match(haystack, needle);
-        assert_eq!(
-            res, None,
-            "{needle:?} should not match {haystack:?} (substring)"
-        );
-        let res = matcher.prefix_match(haystack, needle);
-        assert_eq!(
-            res, None,
-            "{needle:?} should not match {haystack:?} (prefix)"
-        );
-        let res = matcher.postfix_match(haystack, needle);
-        assert_eq!(
-            res, None,
-            "{needle:?} should not match {haystack:?} (postfix)"
-        );
+        for algo in algorithm {
+            let res = match algo {
+                FuzzyOptimal => matcher.fuzzy_match(haystack, needle),
+                FuzzyGreedy => matcher.fuzzy_match_greedy(haystack, needle),
+                Substring => matcher.substring_match(haystack, needle),
+                Prefix => matcher.prefix_match(haystack, needle),
+                Postfix => matcher.postfix_match(haystack, needle),
+                Exact => matcher.exact_match(haystack, needle),
+            };
+            assert_eq!(
+                res, None,
+                "{needle:?} should not match {haystack:?} {algo:?}"
+            );
+        }
     }
 }
 
+pub fn assert_not_matches(normalize: bool, case_sensitive: bool, cases: &[(&str, &str)]) {
+    assert_not_matches_with(
+        normalize,
+        case_sensitive,
+        &[FuzzyOptimal, FuzzyGreedy, Substring, Prefix, Postfix, Exact],
+        cases,
+    )
+}
+
 const BONUS_BOUNDARY_WHITE: u16 = Config::DEFAULT.bonus_boundary_white;
 const BONUS_BOUNDARY_DELIMITER: u16 = Config::DEFAULT.bonus_boundary_delimiter;
 
@@ -377,6 +375,15 @@ fn test_substring() {
             ),
         ],
     );
+    assert_not_matches_with(
+        true,
+        false,
+        &[Prefix, Substring, Postfix, Exact],
+        &[(
+            "At the Road’s End - Seeming - SOL: A Self-Banishment Ritual",
+            "adi",
+        )],
+    )
 }
 
 #[test]
@@ -476,12 +483,20 @@ fn test_unicode() {
         false,
         false,
         false,
-        &[(
-            "你好世界",
-            "你好",
-            &[0, 1],
-            BONUS_BOUNDARY_WHITE * (BONUS_FIRST_CHAR_MULTIPLIER + 1),
-        )],
+        &[
+            (
+                "你好世界",
+                "你好",
+                &[0, 1],
+                BONUS_BOUNDARY_WHITE * (BONUS_FIRST_CHAR_MULTIPLIER + 1),
+            ),
+            (
+                " 你好世界",
+                "你好",
+                &[1, 2],
+                BONUS_BOUNDARY_WHITE * (BONUS_FIRST_CHAR_MULTIPLIER + 1),
+            ),
+        ],
     );
     assert_matches(
         &[FuzzyGreedy, FuzzyOptimal],
@@ -497,7 +512,6 @@ fn test_unicode() {
         )],
     );
     assert_not_matches(
-        false,
         false,
         false,
         &[("Flibbertigibbet / イタズラっ子たち", "lying")],
@@ -614,7 +628,6 @@ fn test_reject() {
     assert_not_matches(
         true,
         false,
-        false,
         &[
             ("你好界", "abc"),
             ("你好界", "a"),
@@ -627,7 +640,6 @@ fn test_reject() {
     assert_not_matches(
         true,
         true,
-        false,
         &[
             ("你好界", "abc"),
             ("abc", "你"),
@@ -643,14 +655,13 @@ fn test_reject() {
     assert_not_matches(
         false,
         true,
-        false,
         &[
             ("Só Danço Samba", "sod"),
             ("Só Danço Samba", "soc"),
             ("Só Danç", "So"),
         ],
     );
-    assert_not_matches(false, false, false, &[("ۂۂfoۂۂ", "foo")]);
+    assert_not_matches(false, false, &[("ۂۂfoۂۂ", "foo")]);
 }
 
 #[test]