fixed stopwords elimination, flaky build test

tomfran · Jan 15, 2024 · 6484a9d · 6484a9d
1 parent 0ac456f
commit 6484a9d
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 2 deletions.
diff --git a/data/index_unit_test/docs/3.txt b/data/index_unit_test/docs/3.txt
@@ -0,0 +1 @@
+man world
diff --git a/src/index/builder.rs b/src/index/builder.rs
@@ -18,7 +18,7 @@ use super::{
 };
 
 const PROGRESS_STYLE: &str =
-    " Documents per second: {per_sec:<3}\n\n [{elapsed_precise}] [{bar:50}] {pos}/{len} ({eta})";
+    "Documents per second: {per_sec:<3}\n\n[{elapsed_precise}] [{bar:50}] {pos}/{len} [{eta_precise}]";
 const PROGRESS_CHARS: &str = "=> ";
 
 const CUTOFF_THRESHOLD: f64 = 0.8;
@@ -93,7 +93,7 @@ fn build_in_memory(input_dir: &str, tokenizer: &Tokenizer, stemmer: &Stemmer) ->
 
     let final_postings = postings.into_inner().unwrap();
 
-    let frequency_threshold = (final_postings.len() as f64 * CUTOFF_THRESHOLD) as u32;
+    let frequency_threshold = (doc_id_mutex.into_inner().unwrap() as f64 * CUTOFF_THRESHOLD) as u32;
 
     let sorted_term_index_map: BTreeMap<String, usize> = term_index_map
         .into_inner()