0.2.0: sorta: add option to align lines, make color an optional flag

Cargo.lock

@@ -12,14 +12,115 @@ dependencies = [
 ]
 
 [[package]]
-name = "evenmoreutils"
+name = "anstream"
-version = "0.1.0"
+version = "0.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b1f58811cfac344940f1a400b6e6231ce35171f614f26439e80f8c1465c5cc0c"
 dependencies = [
+ "anstyle",
+ "anstyle-parse",
+ "anstyle-query",
+ "anstyle-wincon",
+ "colorchoice",
+ "utf8parse",
+]
+
+[[package]]
+name = "anstyle"
+version = "1.0.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "15c4c2c83f81532e5845a733998b6971faca23490340a418e9b72a3ec9de12ea"
+
+[[package]]
+name = "anstyle-parse"
+version = "0.2.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "938874ff5980b03a87c5524b3ae5b59cf99b1d6bc836848df7bc5ada9643c333"
+dependencies = [
+ "utf8parse",
+]
+
+[[package]]
+name = "anstyle-query"
+version = "1.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5ca11d4be1bab0c8bc8734a9aa7bf4ee8316d462a08c6ac5052f888fef5b494b"
+dependencies = [
+ "windows-sys",
+]
+
+[[package]]
+name = "anstyle-wincon"
+version = "2.1.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "58f54d10c6dfa51283a066ceab3ec1ab78d13fae00aa49243a45e4571fb79dfd"
+dependencies = [
+ "anstyle",
+ "windows-sys",
+]
+
+[[package]]
+name = "clap"
+version = "4.4.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6a13b88d2c62ff462f88e4a121f17a82c1af05693a2f192b5c38d14de73c19f6"
+dependencies = [
+ "clap_builder",
+ "clap_derive",
+]
+
+[[package]]
+name = "clap_builder"
+version = "4.4.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2bb9faaa7c2ef94b2743a21f5a29e6f0010dff4caa69ac8e9d6cf8b6fa74da08"
+dependencies = [
+ "anstream",
+ "anstyle",
+ "clap_lex",
+ "strsim",
+]
+
+[[package]]
+name = "clap_derive"
+version = "4.4.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0862016ff20d69b84ef8247369fabf5c008a7417002411897d40ee1f4532b873"
+dependencies = [
+ "heck",
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "clap_lex"
+version = "0.5.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "cd7cc57abe963c6d3b9d8be5b06ba7c8957a930305ca90304f24ef040aa6f961"
+
+[[package]]
+name = "colorchoice"
+version = "1.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "acbf1af155f9b9ef647e42cdc158db4b64a1b61f743629225fde6f3e0be2a7c7"
+
+[[package]]
+name = "evenmoreutils"
+version = "0.2.0"
+dependencies = [
+ "clap",
  "lazy_static",
  "regex",
  "thiserror",
 ]
 
+[[package]]
+name = "heck"
+version = "0.4.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "95505c38b4572b2d910cecb0281560f54b440a19336cbbcb27bf6ce6adc6f5a8"
+
 [[package]]
 name = "lazy_static"
 version = "1.4.0"
@@ -79,6 +180,12 @@ version = "0.7.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "e5ea92a5b6195c6ef2a0295ea818b312502c6fc94dde986c5553242e18fd4ce2"
 
+[[package]]
+name = "strsim"
+version = "0.10.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "73473c0e59e6d5812c5dfe2a064a6444949f089e20eec9a2e5506596494e4623"
+
 [[package]]
 name = "syn"
 version = "2.0.27"
@@ -115,3 +222,75 @@ name = "unicode-ident"
 version = "1.0.11"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "301abaae475aa91687eb82514b328ab47a211a533026cb25fc3e519b86adfc3c"
+
+[[package]]
+name = "utf8parse"
+version = "0.2.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "711b9620af191e0cdc7468a8d14e709c3dcdb115b36f838e601583af800a370a"
+
+[[package]]
+name = "windows-sys"
+version = "0.48.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "677d2418bec65e3338edb076e806bc1ec15693c5d0104683f2efe857f61056a9"
+dependencies = [
+ "windows-targets",
+]
+
+[[package]]
+name = "windows-targets"
+version = "0.48.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9a2fa6e2155d7247be68c096456083145c183cbbbc2764150dda45a87197940c"
+dependencies = [
+ "windows_aarch64_gnullvm",
+ "windows_aarch64_msvc",
+ "windows_i686_gnu",
+ "windows_i686_msvc",
+ "windows_x86_64_gnu",
+ "windows_x86_64_gnullvm",
+ "windows_x86_64_msvc",
+]
+
+[[package]]
+name = "windows_aarch64_gnullvm"
+version = "0.48.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2b38e32f0abccf9987a4e3079dfb67dcd799fb61361e53e2882c3cbaf0d905d8"
+
+[[package]]
+name = "windows_aarch64_msvc"
+version = "0.48.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "dc35310971f3b2dbbf3f0690a219f40e2d9afcf64f9ab7cc1be722937c26b4bc"
+
+[[package]]
+name = "windows_i686_gnu"
+version = "0.48.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a75915e7def60c94dcef72200b9a8e58e5091744960da64ec734a6c6e9b3743e"
+
+[[package]]
+name = "windows_i686_msvc"
+version = "0.48.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8f55c233f70c4b27f66c523580f78f1004e8b5a8b659e05a4eb49d4166cca406"
+
+[[package]]
+name = "windows_x86_64_gnu"
+version = "0.48.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "53d40abd2583d23e4718fddf1ebec84dbff8381c07cae67ff7768bbf19c6718e"
+
+[[package]]
+name = "windows_x86_64_gnullvm"
+version = "0.48.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0b7b52767868a23d5bab768e390dc5f5c55825b6d30b86c844ff2dc7414044cc"
+
+[[package]]
+name = "windows_x86_64_msvc"
+version = "0.48.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ed94fce61571a4006852b7389a063ab983c02eb1bb37b47f8272ce92d06d9538"

Cargo.toml

@@ -1,11 +1,12 @@
 [package]
 name = "evenmoreutils"
-version = "0.1.0"
+version = "0.2.0"
 edition = "2021"
 authors = ["xenofem <xenofem@xeno.science>"]
 license = "MIT"
 
 [dependencies]
+clap = { version = "4.3", features = ["derive"] }
 lazy_static = "1.4"
 regex = "1.7"
 thiserror = "1.0"

src/bin/sorta.rs

@@ -1,35 +1,38 @@
 use std::io::{self, BufRead, Write};
 
+use clap::Parser;
 use lazy_static::lazy_static;
-use regex::{Regex, escape};
+use regex::{escape, Regex};
+
+const COLOR_START: &str = "\x1b[01;34m";
+const COLOR_END: &str = "\x1b[0m";
 
 lazy_static! {
-    static ref PARENS: Vec<(char, char)> = vec![
+    static ref PARENS: Vec<(char, char)> = vec![('(', ')'), ('[', ']'), ('{', '}'), ('【', '】')];
-        ('(', ')'),
+    static ref ARTICLES: Vec<&'static str> = vec!["a", "an", "the"];
-        ('[', ']'),
+    static ref CHAFF: Vec<&'static str> = vec!["-", "_", ".", "『", "』"];
-        ('{', '}'),
+    static ref EXTENSIONS: Vec<&'static str> = vec!["mkv", "mp4", "webm"];
-        ('【', '】'),
-    ];
-    static ref ARTICLES: Vec<&'static str> = vec![
-        "a",
-        "an",
-        "the",
-    ];
-    static ref CHAFF: Vec<&'static str> = vec![
-        "-",
-        "_",
-        ".",
-        "『",
-        "』",
-    ];
-    static ref EXTENSIONS: Vec<&'static str> = vec![
-        "mkv",
-        "mp4",
-        "webm",
-    ];
 }
 
-fn alternator_regex(choices: &Vec<&str>) -> String {
+#[derive(Parser, Debug)]
+#[clap(author, version)]
+/// Sort lines based on the "important" part of each line
+///
+/// sorta sorts the lines of stdin by stripping off prefixes such as
+/// parenthesized descriptors, articles, etc, sorting based on the
+/// stripped versions, and printing the original lines to stdout in
+/// sorted order.
+struct Cli {
+    /// Indent lines with spaces to align the important parts of each line
+    #[arg(short, long)]
+    align: bool,
+
+    /// Color the important part of each line
+    #[arg(short, long)]
+    color: bool,
+}
+
+fn alternator_regex(choices: &[&str]) -> String {
     let mut result = String::new();
     for (i, choice) in choices.iter().enumerate() {
         if i > 0 {
@@ -44,6 +47,8 @@ fn main() -> io::Result<()> {
     let article_regex_raw = format!(r"\b({})\b", alternator_regex(&ARTICLES));
     let extension_regex_raw = format!(r"\.({})", alternator_regex(&EXTENSIONS));
 
+    let cli = Cli::parse();
+
     let mut space_paren_chaff_regex_raw = String::from(r"\s");
     for (open, close) in PARENS.iter() {
         let open = escape(&open.to_string());
@@ -55,8 +60,14 @@ fn main() -> io::Result<()> {
         space_paren_chaff_regex_raw.push_str(&escape(chaff));
     }
 
-    let strip_start_regex = Regex::new(&format!("^(?i)({article_regex_raw}|{space_paren_chaff_regex_raw})*")).unwrap();
+    let strip_start_regex = Regex::new(&format!(
-    let strip_end_regex = Regex::new(&format!("(?i)({space_paren_chaff_regex_raw})*({extension_regex_raw})?$")).unwrap();
+        "^(?i)({article_regex_raw}|{space_paren_chaff_regex_raw})*"
+    ))
+    .unwrap();
+    let strip_end_regex = Regex::new(&format!(
+        "(?i)({space_paren_chaff_regex_raw})*({extension_regex_raw})?$"
+    ))
+    .unwrap();
 
     let mut results = Vec::new();
 
@@ -64,8 +75,16 @@ fn main() -> io::Result<()> {
     for line in stdin.lock().lines() {
         let line = line?;
         let strip_start_idx = strip_start_regex.find(&line).map(|m| m.end()).unwrap_or(0);
-        let strip_end_idx = strip_end_regex.find_at(&line, strip_start_idx).map(|m| m.start()).unwrap_or(line.len());
+        let strip_end_idx = strip_end_regex
-        let item = ((&line[strip_start_idx..]).to_lowercase(), line, strip_start_idx, strip_end_idx);
+            .find_at(&line, strip_start_idx)
+            .map(|m| m.start())
+            .unwrap_or(line.len());
+        let item = (
+            line[strip_start_idx..].to_lowercase(),
+            line,
+            strip_start_idx,
+            strip_end_idx,
+        );
         let insertion_idx = match results.binary_search(&item) {
             Ok(n) => n,
             Err(n) => n,
@@ -73,9 +92,28 @@ fn main() -> io::Result<()> {
         results.insert(insertion_idx, item);
     }
 
+    let alignment_width = if cli.align {
+        results
+            .iter()
+            .map(|(_, _, start_idx, _)| start_idx)
+            .max()
+            .copied()
+            .unwrap_or_default()
+    } else {
+        0
+    };
+
     let mut stdout = io::stdout().lock();
     for (_, line, strip_start_idx, strip_end_idx) in results.into_iter() {
-        writeln!(stdout, "{}\x1b[01;34m{}\x1b[0m{}", &line[..strip_start_idx], &line[strip_start_idx..strip_end_idx], &line[strip_end_idx..])?;
+        writeln!(
+            stdout,
+            "{:>alignment_width$}{}{}{}{}",
+            &line[..strip_start_idx],
+            if cli.color { COLOR_START } else { "" },
+            &line[strip_start_idx..strip_end_idx],
+            if cli.color { COLOR_END } else { "" },
+            &line[strip_end_idx..],
+        )?;
     }
     Ok(())
 }