From 1549944789e5f8ec3c53c25f28f828ad710a1b40 Mon Sep 17 00:00:00 2001
From: LeoniePhiline <22329650+LeoniePhiline@users.noreply.github.com>
Date: Thu, 4 Jul 2024 22:39:08 +0200
Subject: [PATCH] feat: determine system locale to use in locale-aware unicode
 sorting

Prerequisite to an implementation of #1006[1]
following the example of tidy's `sort_carefully`[2].

[1]: https://github.com/eza-community/eza/issues/1006
[2]: https://github.com/sts10/tidy/blob/main/src/list_manipulations.rs#L24
---
 Cargo.lock  | 45 +++++++++++++++++++++++++++++++++++++++++++++
 Cargo.toml  |  1 +
 src/main.rs | 23 +++++++++++++++++++++++
 3 files changed, 69 insertions(+)

diff --git a/Cargo.lock b/Cargo.lock
index e4bd8220b..167df0892 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -346,6 +346,17 @@ dependencies = [
  "powerfmt",
 ]
 
+[[package]]
+name = "displaydoc"
+version = "0.2.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "97369cbbc041bc366949bc74d34658d6cda5621039731c6310521892a3a20ae0"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
 [[package]]
 name = "dunce"
 version = "1.0.4"
@@ -394,6 +405,7 @@ dependencies = [
  "criterion",
  "git2",
  "glob",
+ "icu_locid",
  "libc",
  "locale",
  "log",
@@ -529,6 +541,18 @@ dependencies = [
  "cc",
 ]
 
+[[package]]
+name = "icu_locid"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "13acbb8371917fc971be86fc8057c41a64b521c184808a698c02acc242dbf637"
+dependencies = [
+ "displaydoc",
+ "litemap",
+ "tinystr",
+ "writeable",
+]
+
 [[package]]
 name = "idna"
 version = "0.2.3"
@@ -637,6 +661,12 @@ version = "0.4.11"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "969488b55f8ac402214f3f5fd243ebb7206cf82de60d3172994707a4bcc2b829"
 
+[[package]]
+name = "litemap"
+version = "0.7.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "643cb0b8d4fcc284004d5fd0d67ccf61dfffadb7f75e1e71bc420f4688a3a704"
+
 [[package]]
 name = "locale"
 version = "0.2.2"
@@ -1220,6 +1250,15 @@ version = "0.4.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "a1710e589de0a76aaf295cd47a6699f6405737dbfd3cf2b75c92d000b548d0e6"
 
+[[package]]
+name = "tinystr"
+version = "0.7.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9117f5d4db391c1cf6927e7bea3db74b9a1c1add8f7eda9ffd5364f40f57b82f"
+dependencies = [
+ "displaydoc",
+]
+
 [[package]]
 name = "tinytemplate"
 version = "1.2.1"
@@ -1621,6 +1660,12 @@ dependencies = [
  "memchr",
 ]
 
+[[package]]
+name = "writeable"
+version = "0.5.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1e9df38ee2d2c3c5948ea468a8406ff0db0b29ae1ffde1bcf20ef305bcc95c51"
+
 [[package]]
 name = "zoneinfo_compiled"
 version = "0.5.1"
diff --git a/Cargo.toml b/Cargo.toml
index 1603d7e12..cd70bd43e 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -74,6 +74,7 @@ name = "eza"
 nu-ansi-term = "0.50.0"
 chrono = { version = "0.4.34", default-features = false, features = ["clock"] }
 glob = "0.3"
+icu_locid = "1.5.0"
 libc = "0.2"
 locale = "0.2"
 log = "0.4"
diff --git a/src/main.rs b/src/main.rs
index 162e950cc..e2da914c4 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -81,6 +81,25 @@ fn main() {
                 }
             }
 
+            let lang = if let Ok(lang) = std::env::var("LANG") {
+                // Split e.g. "en_US.UTF-8", extracting "en_US".
+                match lang.split_once('.') {
+                    Some((lang, _)) => lang.to_owned(),
+                    None => lang,
+                }
+            } else {
+                // Fall back to "en_US" if the `LANG` environment variable is not set.
+                warn!("`LANG` is not set. Falling back to 'en_US'.");
+                String::from("en_US")
+            };
+
+            let locale: icu_locid::Locale = if let Ok(locale) = lang.parse() {
+                locale
+            } else {
+                warn!("Your locale '{lang}' could not be parsed. Falling back to 'en_US'.");
+                "en_US".parse().unwrap()
+            };
+
             let git = git_options(&options, &input_paths);
             let writer = io::stdout();
             let git_repos = git_repos(&options, &input_paths);
@@ -91,6 +110,7 @@ fn main() {
                 options,
                 writer,
                 input_paths,
+                locale,
                 theme,
                 console_width,
                 git,
@@ -149,6 +169,9 @@ pub struct Exa<'args> {
     /// names (anything that isn’t an option).
     pub input_paths: Vec<&'args OsStr>,
 
+    /// The system locale, used for locale-aware unicode string sorting.
+    pub locale: icu_locid::Locale,
+
     /// The theme that has been configured from the command-line options and
     /// environment variables. If colours are disabled, this is a theme with
     /// every style set to the default.