release: 0.10.0

Author: joshstoik1

CHANGELOG.md

@@ -2,6 +2,19 @@
 
 
 
+## [0.10.0](https://github.com/Blobfolio/dowser/releases/tag/v0.10.0) - 2024-10-22
+
+### New
+
+* `Dowser::read_paths_from_file`
+
+### Changed
+
+* Bump MSRV to `1.81`
+* Miscellaneous code cleanup and lints
+
+
+
 ## [0.9.3](https://github.com/Blobfolio/dowser/releases/tag/v0.9.3) - 2024-09-05
 
 ### Changed

CREDITS.md

@@ -1,11 +1,11 @@
 # Project Dependencies
     Package:   dowser
-    Version:   0.9.3
-    Generated: 2024-09-05 19:25:18 UTC
+    Version:   0.10.0
+    Generated: 2024-10-23 04:07:24 UTC
 
 | Package | Version | Author(s) | License |
 | ---- | ---- | ---- | ---- |
 | [ahash](https://github.com/tkaitchuck/ahash) | 0.8.11 | [Tom Kaitchuck](mailto:[email protected]) | Apache-2.0 or MIT |
 | [cfg-if](https://github.com/alexcrichton/cfg-if) | 1.0.0 | [Alex Crichton](mailto:[email protected]) | Apache-2.0 or MIT |
-| [dactyl](https://github.com/Blobfolio/dactyl) | 0.7.3 | [Blobfolio, LLC.](mailto:[email protected]) | WTFPL |
+| [dactyl](https://github.com/Blobfolio/dactyl) | 0.7.4 | [Blobfolio, LLC.](mailto:[email protected]) | WTFPL |
 | [zerocopy](https://github.com/google/zerocopy) | 0.7.35 | [Joshua Liebow-Feeser](mailto:[email protected]) | Apache-2.0, BSD-2-Clause, or MIT |

Cargo.toml

@@ -1,9 +1,9 @@
 [package]
 name = "dowser"
-version = "0.9.3"
+version = "0.10.0"
 authors = ["Blobfolio, LLC. <[email protected]>"]
 edition = "2021"
-rust-version = "1.72"
+rust-version = "1.81"
 description = "A recursive, canonicalizing file finding library for Unix."
 license = "WTFPL"
 repository = "https://github.com/Blobfolio/dowser"

src/entry.rs

@@ -2,10 +2,8 @@
 # Dowser: Obligatory `DirEntry` Replacement.
 */
 
-use ahash::AHasher;
 use std::{
 	fs::DirEntry,
-	hash::Hasher,
 	io::Result,
 	path::{
 		Path,
@@ -15,6 +13,18 @@ use std::{
 
 
 
+/// # Static Hasher.
+///
+/// This is used for cheap collision detection. No need to get fancy with it.
+const AHASHER: ahash::RandomState = ahash::RandomState::with_seeds(
+	0x8596_cc44_bef0_1aa0,
+	0x98d4_0948_da60_19ae,
+	0x49f1_3013_c503_a6aa,
+	0xc4d7_82ff_3c9f_7bef,
+);
+
+
+
 /// # File Entry.
 ///
 /// This holds a pre-computed hash, whether or not the path points to a
@@ -69,27 +79,22 @@ impl Entry {
 
 	#[cfg(unix)]
 	#[must_use]
+	#[inline]
 	/// # Hash Path.
 	///
 	/// Since all paths are canonical, we can test for uniqueness by simply
 	/// hashing them.
 	pub(super) fn hash_path(path: &Path) -> u64 {
 		use std::os::unix::ffi::OsStrExt;
-		let mut hasher = AHasher::default();
-		hasher.write(path.as_os_str().as_bytes());
-		hasher.finish()
+		AHASHER.hash_one(path.as_os_str().as_bytes())
 	}
 
 	#[cfg(not(unix))]
 	#[must_use]
+	#[inline]
 	/// # Hash Path.
 	///
 	/// Since all paths are canonical, we can test for uniqueness by simply
 	/// hashing them.
-	pub(super) fn hash_path(path: &Path) -> u64 {
-		use std::hash::Hash;
-		let mut hasher = AHasher::default();
-		path.hash(&mut hasher);
-		hasher.finish()
-	}
+	pub(super) fn hash_path(path: &Path) -> u64 { AHASHER.hash_one(path) }
 }

src/ext.rs

@@ -545,7 +545,7 @@ impl Extension {
 
 /// # Codegen Helpers.
 impl Extension {
-	#[allow(clippy::needless_doctest_main)] // For demonstration.
+	#[expect(clippy::needless_doctest_main, reason = "For demonstration.")]
 	#[must_use]
 	/// # Codegen Helper.
 	///

src/iter.rs

@@ -228,6 +228,56 @@ impl Dowser {
 	}
 }
 
+impl Dowser {
+	/// # Load Paths From File.
+	///
+	/// Queue up multiple file and/or directory paths from a text file, one
+	/// entry per line.
+	///
+	/// Lines are trimmed and ignored if empty, but otherwise resolved the
+	/// same as if passed directly to methods like [`Dowser::with_path`], i.e.
+	/// relative to the current working directory (not the text file itself).
+	///
+	/// For that reason, it is recommended that all paths stored in text files
+	/// be absolute to avoid any ambiguity.
+	///
+	/// ## Examples
+	///
+	/// ```no_run
+	/// use dowser::Dowser;
+	/// use std::path::PathBuf;
+	///
+	/// // Read the paths from list.txt.
+	/// let mut crawler = Dowser::default();
+	/// crawler.read_paths_from_file("list.txt").unwrap();
+	///
+	/// // Crunch into a vec.
+	/// let files: Vec::<PathBuf> = crawler.collect();
+	/// ```
+	///
+	/// ## Errors
+	///
+	/// This method will bubble up any errors encountered while trying to read
+	/// the text file.
+	pub fn read_paths_from_file<P: AsRef<Path>>(&mut self, src: P)
+	-> Result<(), std::io::Error> {
+		let raw = std::fs::read_to_string(src)?;
+		for line in raw.lines() {
+			let line = line.trim();
+			if ! line.is_empty() {
+				if let Some(e) = Entry::from_path(line) {
+					if self.seen.insert(e.hash) {
+						if e.is_dir { self.dirs.push(e.path); }
+						else { self.files.push(e.path); }
+					}
+				}
+			}
+		}
+
+		Ok(())
+	}
+}
+
 impl Dowser {
 	#[must_use]
 	#[inline]
@@ -581,4 +631,55 @@ mod tests {
 		let _res = Dowser::default().without_paths([path]);
 		let _res = Dowser::default().without_paths(&[path.to_path_buf()]);
 	}
+
+	#[test]
+	fn t_read_paths_from_file() {
+		use std::collections::BTreeSet;
+		use std::fs::File;
+		use std::io::Write;
+
+		// Find the temporary directory.
+		let tmp = std::env::temp_dir();
+		if ! tmp.is_dir() { return; }
+
+		// Declare a few paths to test crawl.
+		let asset_dir = std::fs::canonicalize("tests/assets")
+			.expect("Missing dowser assets dir");
+		let link01 = std::fs::canonicalize("tests/links/01")
+			.expect("Missing dowser links/01");
+
+		// Mock up a text file containing those entries.
+		let text_file = tmp.join("dowser.test.txt");
+		let text = format!(
+			"{}\n{}\n",
+			asset_dir.as_os_str().to_str().expect("Asset dir cannot be represented as a string."),
+			link01.as_os_str().to_str().expect("Link01 cannot be represented as a string."),
+		);
+
+		// Try to save the text file.
+		let res = File::create(&text_file)
+			.and_then(|mut file|
+				file.write_all(text.as_bytes()).and_then(|()| file.flush())
+			);
+
+		// Not all environments will allow that; only proceed with the testing
+		// if it worked.
+		if res.is_ok() && text_file.is_file() {
+			// Feed the text file to dowser, collect the results.
+			let mut crawl = Dowser::default();
+			crawl.read_paths_from_file(&text_file)
+				.expect("Loading text file failed.");
+			let found: BTreeSet<PathBuf> = crawl.collect();
+
+			// We don't need the text file anymore.
+			let _res = std::fs::remove_file(text_file);
+
+			// It should have found the following!
+			assert!(found.len() == 4);
+			assert!(found.contains(&link01));
+			assert!(found.contains(&asset_dir.join("file.txt")));
+			assert!(found.contains(&asset_dir.join("functioning.JPEG")));
+			assert!(found.contains(&asset_dir.join("is-executable.sh")));
+		}
+	}
 }

src/lib.rs

@@ -74,7 +74,7 @@ assert_eq!(files1.len(), files2.len());
 #![forbid(unsafe_code)]
 
 #![deny(
-	// TODO: clippy::allow_attributes_without_reason,
+	clippy::allow_attributes_without_reason,
 	clippy::correctness,
 	unreachable_pub,
 )]
@@ -86,7 +86,7 @@ assert_eq!(files1.len(), files2.len());
 	clippy::perf,
 	clippy::style,
 
-	// TODO: clippy::allow_attributes,
+	clippy::allow_attributes,
 	clippy::clone_on_ref_ptr,
 	clippy::create_dir,
 	clippy::filetype_is_file,
@@ -120,7 +120,7 @@ assert_eq!(files1.len(), files2.len());
 	unused_import_braces,
 )]
 
-#![allow(clippy::redundant_pub_crate)] // Unresolvable.
+#![expect(clippy::redundant_pub_crate, reason = "Unresolvable.")]
 
 mod entry;
 mod ext;