fmt

1 files changed, 49 insertions, 50 deletions

diff --git a/feat_extractor/src/lib.rs b/feat_extractor/src/lib.rs
index fae0c4a..1cd8150 100644
--- a/feat_extractor/src/lib.rs
+++ b/feat_extractor/src/lib.rs
@@ -2,7 +2,6 @@ use std::collections::HashSet;
 use rich_crate::ManifestExt;
 use rich_crate::Manifest;
 
-
 lazy_static::lazy_static! {
     /// ignore these as keywords
     pub(crate) static ref STOPWORDS: HashSet<&'static str> = [
@@ -35,60 +34,60 @@ lazy_static::lazy_static! {
     ].iter().copied().collect();
 }
 
-    // returns an array of lowercase phrases
-    fn extract_text_phrases(manifest: &Manifest, github_description: Option<&str>, readme_text: Option<&str>) -> Vec<(f64, String)> {
-        let mut out = Vec::new();
-        let mut len = 0;
-        if let Some(s) = &manifest.package().description {
-            let s = s.to_lowercase();
-            len += s.len();
-            out.push((1., s));
-        }
-        if let Some(s) = github_description {
-            let s = s.to_lowercase();
-            len += s.len();
-            out.push((1., s));
-        }
-        if let Some(sub) = &readme_text {
-            // render readme to DOM, extract nodes
-            for par in sub.split('\n') {
-                if len > 200 {
-                    break;
-                }
-                let par = par.trim_start_matches(|c: char| c.is_whitespace() || c == '#' || c == '=' || c == '*' || c == '-');
-                let par = par.replace("http://", " ").replace("https://", " ");
-                if !par.is_empty() {
-                    let par = par.to_lowercase();
-                    len += par.len();
-                    out.push((0.4, par));
-                }
+// returns an array of lowercase phrases
+fn extract_text_phrases(manifest: &Manifest, github_description: Option<&str>, readme_text: Option<&str>) -> Vec<(f64, String)> {
+    let mut out = Vec::new();
+    let mut len = 0;
+    if let Some(s) = &manifest.package().description {
+        let s = s.to_lowercase();
+        len += s.len();
+        out.push((1., s));
+    }
+    if let Some(s) = github_description {
+        let s = s.to_lowercase();
+        len += s.len();
+        out.push((1., s));
+    }
+    if let Some(sub) = &readme_text {
+        // render readme to DOM, extract nodes
+        for par in sub.split('\n') {
+            if len > 200 {
+                break;
+            }
+            let par = par.trim_start_matches(|c: char| c.is_whitespace() || c == '#' || c == '=' || c == '*' || c == '-');
+            let par = par.replace("http://", " ").replace("https://", " ");
+            if !par.is_empty() {
+                let par = par.to_lowercase();
+                len += par.len();
+                out.push((0.4, par));
             }
         }
-        out
     }
+    out
+}
 
-    pub fn auto_keywords(manifest: &Manifest, github_description: Option<&str>, readme_text: Option<&str>) -> Vec<(f32, String)> {
-            let d = extract_text_phrases(manifest, github_description, readme_text);
-            let mut sw = rake::StopWords::new();
-            sw.reserve(STOPWORDS.len());
-            sw.extend(STOPWORDS.iter().map(|s| s.to_string())); // TODO: use real stopwords, THEN filter via STOPWORDS again, because multiple Rust-y words are fine
-            // normalize space and _ to -
-            let r = rake::Rake::new(sw);
-            let rake_keywords = r.run_sentences(d.iter().map(|(_, s)| s.as_str()));
-            let rake_keywords = rake_keywords.iter()
-                .map(|k| (
-                    k.score.min(1.1), //
-                    chop3words(k.keyword.as_str()) // rake generates very long setences sometimes
-                ));
-            // split on / and punctuation too
-            let keywords = d.iter().flat_map(|&(w2, ref d)| d.split_whitespace().map(move |s| (w2, s.trim_end_matches("'s"))))
-                .filter(|&(_, k)| k.len() >= 2)
-                .filter(|&(_, k)| STOPWORDS.get(k).is_none());
+pub fn auto_keywords(manifest: &Manifest, github_description: Option<&str>, readme_text: Option<&str>) -> Vec<(f32, String)> {
+    let d = extract_text_phrases(manifest, github_description, readme_text);
+    let mut sw = rake::StopWords::new();
+    sw.reserve(STOPWORDS.len());
+    sw.extend(STOPWORDS.iter().map(|s| s.to_string())); // TODO: use real stopwords, THEN filter via STOPWORDS again, because multiple Rust-y words are fine
+    // normalize space and _ to -
+    let r = rake::Rake::new(sw);
+    let rake_keywords = r.run_sentences(d.iter().map(|(_, s)| s.as_str()));
+    let rake_keywords = rake_keywords.iter()
+        .map(|k| (
+            k.score.min(1.1), //
+            chop3words(k.keyword.as_str()) // rake generates very long setences sometimes
+        ));
+    // split on / and punctuation too
+    let keywords = d.iter().flat_map(|&(w2, ref d)| d.split_whitespace().map(move |s| (w2, s.trim_end_matches("'s"))))
+        .filter(|&(_, k)| k.len() >= 2)
+        .filter(|&(_, k)| STOPWORDS.get(k).is_none());
 
-            // replace ' ' with '-'
-            // keep if 3 words or less
-            rake_keywords.chain(keywords).take(25).map(|(w, s)| (w as f32, s.to_owned())).collect()
-        }
+    // replace ' ' with '-'
+    // keep if 3 words or less
+    rake_keywords.chain(keywords).take(25).map(|(w, s)| (w as f32, s.to_owned())).collect()
+}
 
 fn chop3words(s: &str) -> &str {
     let mut words = 0;