exclude files from churn (#1120)

* exclude file from churn

* remove print

Cargo.toml

@@ -31,6 +31,7 @@ gix = { version = "0.48.0", default-features = false, features = [
     "max-performance-safe",
 ] }
 git2 = { version = "0.17.2", default-features = false }
+globset = "0.4.11"
 human-panic = "1.1.5"
 image = "0.24.6"
 num-format = "0.4.4"

src/cli.rs

@@ -73,8 +73,8 @@ pub struct InfoCliOptions {
     #[arg(long, value_name = "NUM")]
     pub churn_pool_size: Option<usize>,
     /// Ignore all files & directories matching EXCLUDE
-    #[arg(long, short, num_args = 1.., value_hint = ValueHint::AnyPath)]
-    pub exclude: Vec<PathBuf>,
+    #[arg(long, short, num_args = 1..)]
+    pub exclude: Vec<String>,
     /// Exclude [bot] commits. Use <REGEX> to override the default pattern
     #[arg(long, value_name = "REGEX")]
     pub no_bots: Option<Option<MyRegex>>,

src/info/author.rs

@@ -6,7 +6,7 @@ use crate::{
 use serde::Serialize;
 use std::fmt::Write;
 
-#[derive(Serialize, Clone)]
+#[derive(Serialize, Clone, Debug, PartialEq)]
 #[serde(rename_all = "camelCase")]
 pub struct Author {
     pub name: String,

src/info/churn.rs

@@ -3,7 +3,7 @@ use crate::{cli::NumberSeparator, info::format_number};
 use serde::Serialize;
 use std::fmt::Write;
 
-#[derive(Serialize, Clone)]
+#[derive(Serialize, Clone, Debug, PartialEq)]
 #[serde(rename_all = "camelCase")]
 pub struct FileChurn {
     pub file_path: String,

src/info/git/metrics.rs

@@ -5,6 +5,7 @@ use crate::info::churn::FileChurn;
 use anyhow::Result;
 use gix::bstr::BString;
 use gix::date::Time;
+use globset::{Glob, GlobSetBuilder};
 use std::collections::HashMap;
 
 pub struct GitMetrics {
@@ -39,7 +40,8 @@ impl GitMetrics {
             number_of_commits_by_file_path,
             options.info.number_of_file_churns,
             options.text_formatting.number_separator,
-        );
+            &options.info.exclude,
+        )?;
 
         // This could happen if a branch pointed to non-commit object, so no traversal actually happens.
         let (time_of_first_commit, time_of_most_recent_commit) = time_of_first_commit
@@ -62,19 +64,33 @@ fn compute_file_churns(
     number_of_commits_by_file_path: HashMap<BString, usize>,
     number_of_file_churns_to_display: usize,
     number_separator: NumberSeparator,
-) -> Vec<FileChurn> {
+    globs_to_exclude: &[String],
+) -> Result<Vec<FileChurn>> {
+    let mut builder = GlobSetBuilder::new();
+    for glob in globs_to_exclude {
+        builder.add(Glob::new(glob)?);
+    }
+    let glob_set = builder.build()?;
     let mut number_of_commits_by_file_path_sorted = Vec::from_iter(number_of_commits_by_file_path);
 
     number_of_commits_by_file_path_sorted
         .sort_by(|(_, a_count), (_, b_count)| b_count.cmp(a_count));
 
-    number_of_commits_by_file_path_sorted
+    Ok(number_of_commits_by_file_path_sorted
         .into_iter()
-        .map(|(file_path, nbr_of_commits)| {
-            FileChurn::new(file_path.to_string(), nbr_of_commits, number_separator)
+        .filter_map(|(file_path, nbr_of_commits)| {
+            if !glob_set.is_match(file_path.to_string()) {
+                Some(FileChurn::new(
+                    file_path.to_string(),
+                    nbr_of_commits,
+                    number_separator,
+                ))
+            } else {
+                None
+            }
         })
         .take(number_of_file_churns_to_display)
-        .collect()
+        .collect())
 }
 
 fn compute_authors(
@@ -124,14 +140,14 @@ mod tests {
                 name: "John Doe".into(),
                 email: "johndoe@example.com".into(),
             },
-            10,
+            30,
         );
         number_of_commits_by_signature.insert(
             Sig {
                 name: "Jane Doe".into(),
                 email: "janedoe@example.com".into(),
             },
-            5,
+            20,
         );
         number_of_commits_by_signature.insert(
             Sig {
@@ -140,12 +156,12 @@ mod tests {
             },
             50,
         );
-        let total_number_of_commits = 15;
+        let total_number_of_commits = 100;
         let number_of_authors_to_display = 2;
         let show_email = false;
         let number_separator = NumberSeparator::Comma;
 
-        let (authors, total_number_of_authors) = compute_authors(
+        let (actual, total_number_of_authors) = compute_authors(
             number_of_commits_by_signature,
             total_number_of_commits,
             number_of_authors_to_display,
@@ -154,31 +170,41 @@ mod tests {
         );
 
         assert_eq!(total_number_of_authors, 3);
-        assert_eq!(authors.len(), 2);
-        assert_eq!(authors.get(0).unwrap().name, "Ellen Smith".to_string());
+        let expected = vec![
+            Author::new(String::from("Ellen Smith"), None, 50, 100, number_separator),
+            Author::new(String::from("John Doe"), None, 30, 100, number_separator),
+        ];
+        assert_eq!(actual, expected);
     }
 
     #[test]
-    fn test_compute_file_churns() {
+    fn test_compute_file_churns() -> Result<()> {
         let mut number_of_commits_by_file_path = HashMap::new();
         number_of_commits_by_file_path.insert("path/to/file1.txt".into(), 2);
         number_of_commits_by_file_path.insert("path/to/file2.txt".into(), 5);
         number_of_commits_by_file_path.insert("path/to/file3.txt".into(), 3);
         number_of_commits_by_file_path.insert("path/to/file4.txt".into(), 7);
+        number_of_commits_by_file_path.insert("foo/x/y/file.txt".into(), 70);
+        number_of_commits_by_file_path.insert("foo/x/file.txt".into(), 10);
 
         let number_of_file_churns_to_display = 3;
         let number_separator = NumberSeparator::Comma;
-        let file_churns = compute_file_churns(
+        let globs_to_exclude = vec![
+            "foo/**/file.txt".to_string(),
+            "path/to/file2.txt".to_string(),
+        ];
+        let actual = compute_file_churns(
             number_of_commits_by_file_path,
             number_of_file_churns_to_display,
             number_separator,
-        );
-
-        assert_eq!(file_churns.len(), 3);
-        assert_eq!(
-            file_churns.get(0).unwrap().file_path,
-            "path/to/file4.txt".to_string()
-        );
-        assert_eq!(file_churns.get(0).unwrap().nbr_of_commits, 7);
+            &globs_to_exclude,
+        )?;
+        let expected = vec![
+            FileChurn::new(String::from("path/to/file4.txt"), 7, number_separator),
+            FileChurn::new(String::from("path/to/file3.txt"), 3, number_separator),
+            FileChurn::new(String::from("path/to/file1.txt"), 2, number_separator),
+        ];
+        assert_eq!(actual, expected);
+        Ok(())
     }
 }

src/info/langs/mod.rs

@@ -1,9 +1,7 @@
 use anyhow::{Context, Result};
 use language::{Language, LanguageType};
-use regex::Regex;
 use std::collections::HashMap;
 use std::path::Path;
-use std::path::PathBuf;
 use strum::IntoEnumIterator;
 
 pub mod language;
@@ -17,11 +15,11 @@ pub fn get_main_language(loc_by_language: &[(Language, usize)]) -> Language {
 /// The vector is sorted by loc in descending order.
 pub fn get_loc_by_language_sorted(
     dir: &Path,
-    ignored_directories: &[PathBuf],
+    globs_to_exclude: &[String],
     language_types: &[LanguageType],
     include_hidden: bool,
 ) -> Result<Vec<(Language, usize)>> {
-    let stats = get_statistics(dir, ignored_directories, language_types, include_hidden);
+    let stats = get_statistics(dir, globs_to_exclude, language_types, include_hidden);
 
     let loc_by_language =
         get_loc_by_language(&stats).context("Could not find any source code in this repository")?;
@@ -65,7 +63,7 @@ pub fn get_total_loc(loc_by_language: &[(Language, usize)]) -> usize {
 
 fn get_statistics(
     dir: &Path,
-    ignored_directories: &[PathBuf],
+    globs_to_exclude: &[String],
     language_types: &[LanguageType],
     include_hidden: bool,
 ) -> tokei::Languages {
@@ -77,8 +75,7 @@ fn get_statistics(
         hidden: Some(include_hidden),
         ..tokei::Config::default()
     };
-    let user_ignored = get_ignored_directories(ignored_directories);
-    let ignored: Vec<&str> = user_ignored.iter().map(AsRef::as_ref).collect();
+    let ignored: Vec<&str> = globs_to_exclude.iter().map(AsRef::as_ref).collect();
     languages.get_statistics(&[&dir], &ignored, &tokei_config);
     languages
 }
@@ -90,23 +87,6 @@ fn filter_languages_on_type(types: &[LanguageType]) -> Vec<tokei::LanguageType>
         .collect()
 }
 
-fn get_ignored_directories(user_ignored_directories: &[PathBuf]) -> Vec<String> {
-    let mut ignored_directories = Vec::new();
-    if !user_ignored_directories.is_empty() {
-        let re = Regex::new(r"((.*)+/)+(.*)").unwrap();
-        for user_ignored_directory in user_ignored_directories {
-            let dir = user_ignored_directory.display().to_string();
-            if re.is_match(&dir) {
-                let prefix = if dir.starts_with('/') { "**" } else { "**/" };
-                ignored_directories.push(format!("{prefix}{dir}"));
-            } else {
-                ignored_directories.push(dir);
-            }
-        }
-    }
-    ignored_directories
-}
-
 #[cfg(test)]
 mod test {
     use super::*;

src/info/mod.rs

@@ -134,14 +134,14 @@ pub fn build_info(cli_options: &CliOptions) -> Result<Info> {
     let repo_path = get_work_dir(&repo)?;
 
     let loc_by_language_sorted_handle = std::thread::spawn({
-        let ignored_directories = cli_options.info.exclude.clone();
+        let globs_to_exclude = cli_options.info.exclude.clone();
         let language_types = cli_options.info.r#type.clone();
         let include_hidden = cli_options.info.include_hidden;
         let workdir = repo_path.clone();
         move || {
             langs::get_loc_by_language_sorted(
                 &workdir,
-                &ignored_directories,
+                &globs_to_exclude,
                 &language_types,
                 include_hidden,
             )