From ca93eedec6a1c9af237282daa1b445af4919b8c3 Mon Sep 17 00:00:00 2001
From: Ryan Peach <rgpeach10@gmail.com>
Date: Sat, 2 Nov 2024 11:53:38 -0400
Subject: [PATCH] Now only compare ngrams of the same size

---
 src/config.rs                 |  2 +-
 src/lib.rs                    | 10 ++++++----
 src/rules/similar_filename.rs | 13 +++++++++++++
 3 files changed, 20 insertions(+), 5 deletions(-)

diff --git a/src/config.rs b/src/config.rs
index b606f29..7d3cccd 100644
--- a/src/config.rs
+++ b/src/config.rs
@@ -46,7 +46,7 @@ pub struct Config {
     #[builder(default=r"-|_|\s".to_owned())]
     pub filename_spacing_pattern: String,
     /// See [`self::cli::Config::filename_match_threshold`]
-    #[builder(default = 100)]
+    #[builder(default = 95)]
     pub filename_match_threshold: i64,
     /// See [`self::cli::Config::exclude`]
     #[builder(default=vec![])]
diff --git a/src/lib.rs b/src/lib.rs
index 2669fe2..55fbc98 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -56,14 +56,16 @@ pub fn lib(config: &config::Config) -> Result<OutputReport> {
 
     // All our reports
     // NOTE: Always use `filter_by_excludes` and `dedupe_by_code` on the reports
-    let similar_filenames = SimilarFilename::calculate(
+    let mut similar_filenames = SimilarFilename::calculate(
         &file_ngrams,
         config.filename_match_threshold,
         &filename_spacing_regex,
     )
-    .map_err(|e| miette!("From SimilarFilename: {e}"))?
-    .filter_by_excludes(config.exclude.clone())
-    .dedupe_by_code();
+    .map_err(|e| miette!("From SimilarFilename: {e}"))?;
+    similar_filenames.sort_by(|b, a| a.partial_cmp(b).expect("This never fails"));
+    let similar_filenames = similar_filenames
+        .filter_by_excludes(config.exclude.clone())
+        .dedupe_by_code();
 
     let duplicate_aliases =
         DuplicateAlias::calculate(get_files(config.directories.clone()), config)
diff --git a/src/rules/similar_filename.rs b/src/rules/similar_filename.rs
index 3b29001..61060eb 100644
--- a/src/rules/similar_filename.rs
+++ b/src/rules/similar_filename.rs
@@ -16,6 +16,8 @@ pub struct SimilarFilename {
     /// Used to identify the diagnostic and exclude it if needed
     id: String,
 
+    score: i64,
+
     #[source_code]
     filepaths: String,
 
@@ -29,6 +31,12 @@ pub struct SimilarFilename {
     advice: String,
 }
 
+impl PartialOrd for SimilarFilename {
+    fn partial_cmp(&self, other: &Self) -> Option<std::cmp::Ordering> {
+        self.score.partial_cmp(&other.score)
+    }
+}
+
 impl PartialEq for SimilarFilename {
     fn eq(&self, other: &Self) -> bool {
         self.id == other.id
@@ -109,6 +117,7 @@ impl SimilarFilename {
         );
         Ok(Self {
             id,
+            score,
             filepaths,
             file1_ngram,
             file2_ngram,
@@ -134,6 +143,10 @@ impl SimilarFilename {
         for (i, (ngram, filepath)) in file_ngrams.clone().iter().enumerate() {
             // We can start at i + 1 because we've already checked all previous files
             for (other_ngram, other_filepath) in file_ngrams.iter().skip(i + 1) {
+                if ngram.nb_words() != other_ngram.nb_words() {
+                    continue;
+                }
+
                 file_crosscheck_bar.inc(1);
 
                 // Skip if the same file