feat: allow ignoring hidden files and files matched by .gitignore fil…

…es (#245)

Cargo.toml

@@ -24,11 +24,11 @@ lzzzz = "0.8.0"
 once_cell = "1.8.0"
 snap = "1.0.5"
 tar = "0.4.37"
-walkdir = "2.3.2"
 xz2 = "0.1.6"
 zip = { version = "0.5.13", default-features = false }
 zstd = { version = "0.9.0", default-features = false }
 tempfile = "3.2.0"
+ignore = "0.4.18"
 indicatif = "0.16.2"
 
 [build-dependencies]

src/archive/tar.rs

@@ -10,13 +10,12 @@ use std::{
 
 use fs_err as fs;
 use tar;
-use walkdir::WalkDir;
 
 use crate::{
     error::FinalError,
     info,
     list::FileInArchive,
-    utils::{self, Bytes},
+    utils::{self, Bytes, FileVisibilityPolicy},
 };
 
 /// Unpacks the archive given by `archive` into the folder given by `into`.
@@ -79,7 +78,12 @@ pub fn list_archive(
 }
 
 /// Compresses the archives given by `input_filenames` into the file given previously to `writer`.
-pub fn build_archive_from_paths<W, D>(input_filenames: &[PathBuf], writer: W, mut display_handle: D) -> crate::Result<W>
+pub fn build_archive_from_paths<W, D>(
+    input_filenames: &[PathBuf],
+    writer: W,
+    file_visibility_policy: FileVisibilityPolicy,
+    mut display_handle: D,
+) -> crate::Result<W>
 where
     W: Write,
     D: Write,
@@ -92,7 +96,7 @@ where
         // Safe unwrap, input shall be treated before
         let filename = filename.file_name().unwrap();
 
-        for entry in WalkDir::new(&filename) {
+        for entry in file_visibility_policy.build_walker(filename) {
             let entry = entry?;
             let path = entry.path();
 

src/archive/zip.rs

@@ -9,14 +9,16 @@ use std::{
 };
 
 use fs_err as fs;
-use walkdir::WalkDir;
 use zip::{self, read::ZipFile, ZipArchive};
 
 use crate::{
     error::FinalError,
     info,
     list::FileInArchive,
-    utils::{self, cd_into_same_dir_as, concatenate_os_str_list, get_invalid_utf8_paths, strip_cur_dir, to_utf, Bytes},
+    utils::{
+        self, cd_into_same_dir_as, concatenate_os_str_list, get_invalid_utf8_paths, strip_cur_dir, to_utf, Bytes,
+        FileVisibilityPolicy,
+    },
 };
 
 /// Unpacks the archive given by `archive` into the folder given by `output_folder`.
@@ -119,7 +121,12 @@ where
 }
 
 /// Compresses the archives given by `input_filenames` into the file given previously to `writer`.
-pub fn build_archive_from_paths<W, D>(input_filenames: &[PathBuf], writer: W, mut display_handle: D) -> crate::Result<W>
+pub fn build_archive_from_paths<W, D>(
+    input_filenames: &[PathBuf],
+    writer: W,
+    file_visibility_policy: FileVisibilityPolicy,
+    mut display_handle: D,
+) -> crate::Result<W>
 where
     W: Write + Seek,
     D: Write,
@@ -144,7 +151,7 @@ where
         // Safe unwrap, input shall be treated before
         let filename = filename.file_name().unwrap();
 
-        for entry in WalkDir::new(filename) {
+        for entry in file_visibility_policy.build_walker(filename) {
             let entry = entry?;
             let path = entry.path();
 

src/cli.rs

@@ -10,7 +10,7 @@ use clap::Parser;
 use fs_err as fs;
 use once_cell::sync::OnceCell;
 
-use crate::{Opts, QuestionPolicy, Subcommand};
+use crate::{utils::FileVisibilityPolicy, Opts, QuestionPolicy, Subcommand};
 
 /// Whether to enable accessible output (removes info output and reduces other
 /// output, removes visual markers like '[' and ']').
@@ -23,7 +23,7 @@ impl Opts {
     /// And:
     ///   1. Make paths absolute.
     ///   2. Checks the QuestionPolicy.
-    pub fn parse_args() -> crate::Result<(Self, QuestionPolicy)> {
+    pub fn parse_args() -> crate::Result<(Self, QuestionPolicy, FileVisibilityPolicy)> {
         let mut opts = Self::parse();
 
         ACCESSIBLE.set(opts.accessible).unwrap();
@@ -40,7 +40,14 @@ impl Opts {
             (true, true) => unreachable!(),
         };
 
-        Ok((opts, skip_questions_positively))
+        // TODO: change this to be just a single function call?
+        let file_visibility_policy = FileVisibilityPolicy::new()
+            .read_git_exclude(opts.gitignore)
+            .read_ignore(opts.gitignore)
+            .read_git_ignore(opts.gitignore)
+            .read_hidden(opts.hidden);
+
+        Ok((opts, skip_questions_positively, file_visibility_policy))
     }
 }
 

src/commands.rs

@@ -24,7 +24,7 @@ use crate::{
     progress::Progress,
     utils::{
         self, concatenate_os_str_list, dir_is_empty, nice_directory_display, to_utf, try_infer_extension,
-        user_wants_to_continue,
+        user_wants_to_continue, FileVisibilityPolicy,
     },
     warning, Opts, QuestionAction, QuestionPolicy, Subcommand,
 };
@@ -43,7 +43,11 @@ fn represents_several_files(files: &[PathBuf]) -> bool {
 }
 
 /// Entrypoint of ouch, receives cli options and matches Subcommand to decide what to do
-pub fn run(args: Opts, question_policy: QuestionPolicy) -> crate::Result<()> {
+pub fn run(
+    args: Opts,
+    question_policy: QuestionPolicy,
+    file_visibility_policy: FileVisibilityPolicy,
+) -> crate::Result<()> {
     match args.cmd {
         Subcommand::Compress { mut files, output: output_path } => {
             // If the output_path file exists and is the same as some of the input files, warn the user and skip those inputs (in order to avoid compression recursion)
@@ -156,7 +160,8 @@ pub fn run(args: Opts, question_policy: QuestionPolicy) -> crate::Result<()> {
                     formats = new_formats;
                 }
             }
-            let compress_result = compress_files(files, formats, output_file, &output_path, question_policy);
+            let compress_result =
+                compress_files(files, formats, output_file, &output_path, question_policy, file_visibility_policy);
 
             // If any error occurred, delete incomplete file
             if compress_result.is_err() {
@@ -285,6 +290,7 @@ fn compress_files(
     output_file: fs::File,
     output_dir: &Path,
     question_policy: QuestionPolicy,
+    file_visibility_policy: FileVisibilityPolicy,
 ) -> crate::Result<()> {
     // The next lines are for displaying the progress bar
     // If the input files contain a directory, then the total size will be underestimated
@@ -348,6 +354,7 @@ fn compress_files(
             archive::tar::build_archive_from_paths(
                 &files,
                 &mut writer,
+                file_visibility_policy,
                 progress.as_mut().map(Progress::display_handle).unwrap_or(&mut io::stdout()),
             )?;
             writer.flush()?;
@@ -385,6 +392,7 @@ fn compress_files(
             archive::zip::build_archive_from_paths(
                 &files,
                 &mut vec_buffer,
+                file_visibility_policy,
                 progress.as_mut().map(Progress::display_handle).unwrap_or(&mut io::stdout()),
             )?;
             let vec_buffer = vec_buffer.into_inner();

src/error.rs

@@ -172,8 +172,8 @@ impl From<zip::result::ZipError> for Error {
     }
 }
 
-impl From<walkdir::Error> for Error {
-    fn from(err: walkdir::Error) -> Self {
+impl From<ignore::Error> for Error {
+    fn from(err: ignore::Error) -> Self {
         Self::WalkdirError { reason: err.to_string() }
     }
 }

src/main.rs

@@ -28,6 +28,6 @@ fn main() {
 }
 
 fn run() -> Result<()> {
-    let (args, skip_questions_positively) = Opts::parse_args()?;
-    commands::run(args, skip_questions_positively)
+    let (args, skip_questions_positively, file_visibility_policy) = Opts::parse_args()?;
+    commands::run(args, skip_questions_positively, file_visibility_policy)
 }

src/opts.rs

@@ -23,6 +23,14 @@ pub struct Opts {
     #[clap(short = 'A', long, env = "ACCESSIBLE", global = true)]
     pub accessible: bool,
 
+    /// Ignores hidden files
+    #[clap(short = 'H', long)]
+    pub hidden: bool,
+
+    /// Ignores files matched by git's ignore files
+    #[clap(short = 'g', long)]
+    pub gitignore: bool,
+
     /// Ouch and claps subcommands
     #[clap(subcommand)]
     pub cmd: Subcommand,

src/utils/file_visibility.rs

@@ -0,0 +1,68 @@
+use std::path::Path;
+
+/// Determines which files should be read or ignored during directory walking
+pub struct FileVisibilityPolicy {
+    /// Enables reading .ignore files.
+    ///
+    /// Disabled by default.
+    pub read_ignore: bool,
+
+    /// If enabled, ignores hidden files.
+    ///
+    /// Disabled by default
+    pub read_hidden: bool,
+
+    /// Enables reading .gitignore files.
+    ///
+    /// This is enabled by default.
+    pub read_git_ignore: bool,
+
+    /// Enables reading `.git/info/exclude` files.
+    pub read_git_exclude: bool,
+}
+
+impl Default for FileVisibilityPolicy {
+    fn default() -> Self {
+        Self { read_ignore: false, read_hidden: true, read_git_ignore: false, read_git_exclude: false }
+    }
+}
+
+impl FileVisibilityPolicy {
+    pub fn new() -> Self {
+        Self::default()
+    }
+
+    #[must_use]
+    /// Enables reading .ignore files.
+    pub fn read_ignore(self, read_ignore: bool) -> Self {
+        Self { read_ignore, ..self }
+    }
+
+    #[must_use]
+    /// Enables reading .gitignore files.
+    pub fn read_git_ignore(self, read_git_ignore: bool) -> Self {
+        Self { read_git_ignore, ..self }
+    }
+
+    #[must_use]
+    /// Enables reading `.git/info/exclude` files.
+    pub fn read_git_exclude(self, read_git_exclude: bool) -> Self {
+        Self { read_git_exclude, ..self }
+    }
+
+    #[must_use]
+    /// Enables reading `.git/info/exclude` files.
+    pub fn read_hidden(self, read_hidden: bool) -> Self {
+        Self { read_hidden, ..self }
+    }
+
+    /// Walks through a directory using [`ignore::Walk`]
+    pub fn build_walker(&self, path: impl AsRef<Path>) -> ignore::Walk {
+        ignore::WalkBuilder::new(path)
+            .git_exclude(self.read_git_exclude)
+            .git_ignore(self.read_git_ignore)
+            .ignore(self.read_ignore)
+            .hidden(self.read_hidden)
+            .build()
+    }
+}

src/utils/mod.rs

@@ -4,10 +4,12 @@
 //! stdin interaction helpers.
 
 pub mod colors;
+mod file_visibility;
 mod formatting;
 mod fs;
 mod question;
 
+pub use file_visibility::FileVisibilityPolicy;
 pub use formatting::{concatenate_os_str_list, nice_directory_display, strip_cur_dir, to_utf, Bytes};
 pub use fs::{
     cd_into_same_dir_as, clear_path, create_dir_if_non_existent, dir_is_empty, is_symlink, try_infer_extension,