move ascii_reduce to its own module crate

Cargo.lock

@@ -80,6 +80,15 @@ dependencies = [
  "windows-sys",
 ]
 
+[[package]]
+name = "ascii_reduce"
+version = "1.0.0"
+dependencies = [
+ "phf",
+ "phf_codegen",
+ "serde_json",
+]
+
 [[package]]
 name = "autocfg"
 version = "1.1.0"
@@ -656,6 +665,7 @@ dependencies = [
 name = "musicutil"
 version = "0.1.0"
 dependencies = [
+ "ascii_reduce",
  "bytes",
  "clap",
  "html-escape",
@@ -726,6 +736,44 @@ version = "0.1.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "19b17cddbe7ec3f8bc800887bab5e717348c95ea2ca0b1bf0837fb964dc67099"
 
+[[package]]
+name = "phf"
+version = "0.11.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ade2d8b8f33c7333b51bcf0428d37e217e9f32192ae4772156f65063b8ce03dc"
+dependencies = [
+ "phf_shared",
+]
+
+[[package]]
+name = "phf_codegen"
+version = "0.11.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e8d39688d359e6b34654d328e262234662d16cc0f60ec8dcbe5e718709342a5a"
+dependencies = [
+ "phf_generator",
+ "phf_shared",
+]
+
+[[package]]
+name = "phf_generator"
+version = "0.11.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "48e4cc64c2ad9ebe670cb8fd69dd50ae301650392e81c05f9bfcb2d5bdbc24b0"
+dependencies = [
+ "phf_shared",
+ "rand",
+]
+
+[[package]]
+name = "phf_shared"
+version = "0.11.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "90fcb95eef784c2ac79119d1dd819e162b5da872ce6f3c3abe1e8ca1c082f72b"
+dependencies = [
+ "siphasher",
+]
+
 [[package]]
 name = "pkg-config"
 version = "0.3.27"
@@ -766,6 +814,21 @@ dependencies = [
  "proc-macro2",
 ]
 
+[[package]]
+name = "rand"
+version = "0.8.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "34af8d1a0e25924bc5b7c43c079c942339d8f0a8b57c39049bef581b46327404"
+dependencies = [
+ "rand_core",
+]
+
+[[package]]
+name = "rand_core"
+version = "0.6.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ec0be4795e2f6a28069bec0b5ff3e2ac9bafc99e6a9a7dc3547996c5c816922c"
+
 [[package]]
 name = "redox_syscall"
 version = "0.3.5"
@@ -917,6 +980,12 @@ version = "1.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "a7cee0529a6d40f580e7a5e6c495c8fbfe21b7b52795ed4bb5e62cdf92bc6380"
 
+[[package]]
+name = "siphasher"
+version = "0.3.11"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "38b58827f4464d87d377d175e90bf58eb00fd8716ff0a62f80356b5e61555d0d"
+
 [[package]]
 name = "string-error"
 version = "0.1.0"

Cargo.toml

@@ -21,7 +21,13 @@ serde_with = "3"
 # argument parsing
 clap = { version = "4", features = ["derive"] }
 
-# ascii_reduce constants & transcode presets
+# for reducing filenames to ascii
+# useful for when storing music files on mp3 players with broken unicode support
+ascii_reduce = { path = "./modules/ascii_reduce" }
+
+
+
+# transcode presets & format handlers
 lazy_static = "1"
 
 # for scan_for_music

flake.lock

@@ -18,11 +18,11 @@
     },
     "nixpkgs": {
       "locked": {
-        "lastModified": 1697456312,
+        "lastModified": 1698318101,
-        "narHash": "sha256-roiSnrqb5r+ehnKCauPLugoU8S36KgmWraHgRqVYndo=",
+        "narHash": "sha256-gUihHt3yPD7bVqg+k/UVHgngyaJ3DMEBchbymBMvK1E=",
         "owner": "NixOS",
         "repo": "nixpkgs",
-        "rev": "ca012a02bf8327be9e488546faecae5e05d7d749",
+        "rev": "63678e9f3d3afecfeafa0acead6239cdb447574c",
         "type": "github"
       },
       "original": {

modules/ascii_reduce/Cargo.toml

@@ -0,0 +1,14 @@
+[package]
+name = "ascii_reduce"
+version = "1.0.0"
+edition = "2021"
+
+# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+
+[dependencies]
+phf = { version = "0.11", default-features = false }
+
+[build-dependencies]
+serde_json = "1.0"
+phf_codegen = "0.11"
+

modules/ascii_reduce/build.rs

@@ -0,0 +1,48 @@
+use std::collections::HashMap;
+use std::env;
+use std::fs::File;
+use std::io::{BufWriter, Write};
+use std::path::Path;
+
+const MAPPINGS_DATA: &str = include_str!("src/mappings.json");
+
+fn main() {
+	let data: HashMap<String, String> =
+		serde_json::from_str(MAPPINGS_DATA).expect("mapping data invalid");
+
+	let mut replacement_map: HashMap<char, String> = HashMap::new();
+	for (chr, repl) in &data {
+		match chr.parse::<u32>() {
+			Ok(n) => {
+				let b = char::from_u32(n).expect("invalid char in string");
+
+				replacement_map.insert(b, repl.to_string());
+			}
+			Err(e) => {
+				panic!(
+					"mapping data broken, could not parse char {} with error {}",
+					chr, e
+				);
+			}
+		}
+	}
+
+	let mut map: &mut phf_codegen::Map<char> = &mut phf_codegen::Map::new();
+	for replacement in replacement_map.into_iter() {
+		let r_char = replacement.1;
+		map = map.entry(
+			replacement.0,
+			format!("\"{}\"", r_char.escape_debug()).as_str(),
+		)
+	}
+
+	let path = Path::new(&env::var("OUT_DIR").unwrap()).join("codegen.rs");
+	let mut file = BufWriter::new(File::create(&path).unwrap());
+	write!(
+		&mut file,
+		"static MAPPINGS: phf::Map<char, &'static str> = {}",
+		map.build()
+	)
+	.unwrap();
+	write!(&mut file, ";\n").unwrap();
+}

modules/ascii_reduce/src/lib.rs

@@ -0,0 +1,21 @@
+include!(concat!(env!("OUT_DIR"), "/codegen.rs"));
+
+pub fn reduce(input: String) -> String {
+	if input.is_ascii() {
+		return input;
+	}
+
+	let mut output = String::with_capacity(input.len());
+	for c in input.chars() {
+		if c.is_ascii() {
+			output.push(c);
+			continue;
+		}
+
+		if let Some(replacement) = MAPPINGS.get(&c) {
+			output.push_str(replacement);
+		}
+	}
+
+	output
+}

src/commands/process.rs

@@ -5,12 +5,13 @@ use std::thread::scope;
 use crate::args::CLIArgs;
 use crate::types::AudioFileInfo;
 use crate::types::File;
-use crate::utils::ascii_reduce::reduce_to_ascii;
 use crate::utils::formats::get_format_handler;
 #[cfg(feature = "replaygain")]
 use crate::utils::replaygain::analyze_replaygain;
 use crate::utils::scan_for_music;
 
+use ascii_reduce::reduce;
+
 #[derive(Debug, Clone, clap::Args)]
 pub struct ProcessCommandArgs {
 	pub source: String,
@@ -40,8 +41,8 @@ fn rename_file(process_args: &ProcessCommandArgs, file: &mut File) {
 	let artist = artist.replace('\n', "");
 
 	// Step 2: Strip ASCII
-	let title = reduce_to_ascii(title);
+	let title = reduce(title);
-	let artist = reduce_to_ascii(artist);
+	let artist = reduce(artist);
 
 	// Step 3: Remove File Seperators
 	let title = title.replace('\\', &replace_char);

src/utils/ascii_reduce/mod.rs

@@ -1,50 +0,0 @@
-use lazy_static::lazy_static;
-use std::collections::HashMap;
-
-const MAPPINGS_DATA: &str = include_str!("mappings.json");
-
-lazy_static! {
-	static ref MAPPINGS: HashMap<char, String> = {
-		let data: HashMap<String, String> =
-			serde_json::from_str(MAPPINGS_DATA).expect("mapping data invalid");
-
-		let mut replacement_map: HashMap<char, String> = HashMap::new();
-		for (chr, repl) in &data {
-			match chr.parse::<u32>() {
-				Ok(n) => {
-					let b = char::from_u32(n).expect("invalid char in string");
-
-					replacement_map.insert(b, repl.to_string());
-				}
-				Err(e) => {
-					panic!(
-						"mapping data broken, could not parse char {} with error {}",
-						chr, e
-					);
-				}
-			}
-		}
-
-		replacement_map
-	};
-}
-
-pub fn reduce_to_ascii(input: String) -> String {
-	if input.is_ascii() {
-		return input;
-	}
-
-	let mut output = String::with_capacity(input.len());
-	for c in input.chars() {
-		if c.is_ascii() {
-			output.push(c);
-			continue;
-		}
-
-		if let Some(replacement) = MAPPINGS.get(&c) {
-			output.push_str(replacement);
-		}
-	}
-
-	output
-}

src/utils/mod.rs

@@ -1,4 +1,3 @@
-pub mod ascii_reduce;
 pub mod ffprobe;
 pub mod format_detection;
 #[cfg(feature = "replaygain")]