Filesize formatting (#456)

* configure the format of filesize

* type-o

* removed some comments

* updated tests

* accomodated filesize_metric better, added test

Cargo.lock

@@ -103,6 +103,15 @@ version = "0.3.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "a4c527152e37cf757a3f78aae5a06fbeefdb07ccc535c980a3208ee3060dd544"
 
+[[package]]
+name = "arrayvec"
+version = "0.4.12"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "cd9fd44efafa8690358b7408d253adf110036b88f55672a933f01d616ad9b1b9"
+dependencies = [
+ "nodrop",
+]
+
 [[package]]
 name = "arrayvec"
 version = "0.5.2"
@@ -269,7 +278,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "afa748e348ad3be8263be728124b24a24f268266f6f5d58af9d75f6a40b5c587"
 dependencies = [
  "arrayref",
- "arrayvec",
+ "arrayvec 0.5.2",
  "constant_time_eq",
 ]
 
@@ -306,6 +315,12 @@ dependencies = [
  "serde",
 ]
 
+[[package]]
+name = "bumpalo"
+version = "3.8.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8f1e260c3a9040a7c19a12468758f4c16f31a81a1fe087482be9570ec864bb6c"
+
 [[package]]
 name = "byte-unit"
 version = "4.0.13"
@@ -548,6 +563,16 @@ version = "0.2.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "7a81dae078cea95a014a339291cec439d2f232ebe854a9d672b796c6afafa9b7"
 
+[[package]]
+name = "cstr_core"
+version = "0.2.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "917ba9efe9e1e736671d5a03f006afc4e7e3f32503e2077e0bcaf519c0c8c1d3"
+dependencies = [
+ "cty",
+ "memchr",
+]
+
 [[package]]
 name = "csv"
 version = "1.1.6"
@@ -590,6 +615,12 @@ dependencies = [
  "winapi",
 ]
 
+[[package]]
+name = "cty"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b365fabc795046672053e29c954733ec3b05e4be654ab130fe8f1f94d7051f35"
+
 [[package]]
 name = "dialoguer"
 version = "0.9.0"
@@ -1137,6 +1168,15 @@ dependencies = [
  "libc",
 ]
 
+[[package]]
+name = "js-sys"
+version = "0.3.55"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7cc9ffccd38c451a86bf13657df244e9c3f37493cce8e5e21e940963777acc84"
+dependencies = [
+ "wasm-bindgen",
+]
+
 [[package]]
 name = "lazy_static"
 version = "1.4.0"
@@ -1455,6 +1495,12 @@ dependencies = [
  "memoffset",
 ]
 
+[[package]]
+name = "nodrop"
+version = "0.1.14"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "72ef4a56884ca558e5ddb05a1d1e7e1bfd9a68d9ed024c21704cc98872dae1bb"
+
 [[package]]
 name = "nom"
 version = "1.2.4"
@@ -1631,8 +1677,10 @@ dependencies = [
  "indexmap",
  "miette",
  "nu-json",
+ "num-format",
  "serde",
  "serde_json",
+ "sys-locale",
  "thiserror",
  "typetag",
 ]
@@ -1754,6 +1802,16 @@ dependencies = [
  "num-traits",
 ]
 
+[[package]]
+name = "num-format"
+version = "0.4.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "bafe4179722c2894288ee77a9f044f02811c86af699344c498b0840c698a2465"
+dependencies = [
+ "arrayvec 0.4.12",
+ "itoa",
+]
+
 [[package]]
 name = "num-integer"
 version = "0.1.44"
@@ -2660,6 +2718,19 @@ dependencies = [
  "unicode-xid",
 ]
 
+[[package]]
+name = "sys-locale"
+version = "0.1.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "91f89ebb59fa30d4f65fafc2d68e94f6975256fd87e812dd99cb6e020c8563df"
+dependencies = [
+ "cc",
+ "cstr_core",
+ "libc",
+ "web-sys",
+ "winapi",
+]
+
 [[package]]
 name = "sysinfo"
 version = "0.20.5"
@@ -2967,7 +3038,7 @@ version = "0.10.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "6cbce692ab4ca2f1f3047fcf732430249c0e971bfdd2b234cf2c47ad93af5983"
 dependencies = [
- "arrayvec",
+ "arrayvec 0.5.2",
  "utf8parse",
  "vte_generate_state_changes",
 ]
@@ -3003,6 +3074,70 @@ version = "0.10.0+wasi-snapshot-preview1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "1a143597ca7c7793eff794def352d41792a93c481eb1042423ff7ff72ba2c31f"
 
+[[package]]
+name = "wasm-bindgen"
+version = "0.2.78"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "632f73e236b219150ea279196e54e610f5dbafa5d61786303d4da54f84e47fce"
+dependencies = [
+ "cfg-if",
+ "wasm-bindgen-macro",
+]
+
+[[package]]
+name = "wasm-bindgen-backend"
+version = "0.2.78"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a317bf8f9fba2476b4b2c85ef4c4af8ff39c3c7f0cdfeed4f82c34a880aa837b"
+dependencies = [
+ "bumpalo",
+ "lazy_static",
+ "log",
+ "proc-macro2",
+ "quote",
+ "syn",
+ "wasm-bindgen-shared",
+]
+
+[[package]]
+name = "wasm-bindgen-macro"
+version = "0.2.78"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d56146e7c495528bf6587663bea13a8eb588d39b36b679d83972e1a2dbbdacf9"
+dependencies = [
+ "quote",
+ "wasm-bindgen-macro-support",
+]
+
+[[package]]
+name = "wasm-bindgen-macro-support"
+version = "0.2.78"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7803e0eea25835f8abdc585cd3021b3deb11543c6fe226dcd30b228857c5c5ab"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+ "wasm-bindgen-backend",
+ "wasm-bindgen-shared",
+]
+
+[[package]]
+name = "wasm-bindgen-shared"
+version = "0.2.78"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0237232789cf037d5480773fe568aac745bfe2afbc11a863e97901780a6b47cc"
+
+[[package]]
+name = "web-sys"
+version = "0.3.55"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "38eb105f1c59d9eaa6b5cdc92b859d85b926e82cb2e0945cd0c9259faa6fe9fb"
+dependencies = [
+ "js-sys",
+ "wasm-bindgen",
+]
+
 [[package]]
 name = "winapi"
 version = "0.3.9"

crates/nu-protocol/Cargo.toml

@@ -17,6 +17,8 @@ im = "15.0.0"
 serde_json = { version = "1.0", optional = true }
 nu-json = { path = "../nu-json" }
 typetag = "0.1.8"
+num-format = "0.4.0"
+sys-locale = "0.1.0"
 
 [features]
 plugin = ["serde_json"]

crates/nu-protocol/src/config.rs

@@ -14,6 +14,7 @@ pub struct Config {
     pub footer_mode: FooterMode,
     pub animate_prompt: bool,
     pub float_precision: i64,
+    pub filesize_format: String,
     pub without_color: bool,
 }
 
@@ -28,6 +29,7 @@ impl Default for Config {
             footer_mode: FooterMode::Never,
             animate_prompt: ANIMATE_PROMPT_DEFAULT,
             float_precision: 4,
+            filesize_format: "auto".into(),
             without_color: false,
         }
     }
@@ -74,7 +76,7 @@ impl Value {
                     config.use_grid_icons = value.as_bool()?;
                 }
                 "footer_mode" => {
-                    let val_str = value.as_string()?;
+                    let val_str = value.as_string()?.to_lowercase();
                     config.footer_mode = match val_str.as_ref() {
                         "auto" => FooterMode::Auto,
                         "never" => FooterMode::Never,
@@ -94,6 +96,9 @@ impl Value {
                 "without_color" => {
                     config.without_color = value.as_bool()?;
                 }
+                "filesize_format" => {
+                    config.filesize_format = value.as_string()?.to_lowercase();
+                }
                 _ => {}
             }
         }

crates/nu-protocol/src/value/mod.rs

@@ -5,13 +5,16 @@ mod range;
 mod stream;
 mod unit;
 
+use byte_unit::ByteUnit;
 use chrono::{DateTime, FixedOffset};
 use chrono_humanize::HumanTime;
 pub use from_value::FromValue;
 use indexmap::map::IndexMap;
+use num_format::{Locale, ToFormattedString};
 pub use range::*;
 use serde::{Deserialize, Serialize};
 pub use stream::*;
+use sys_locale::get_locale;
 pub use unit::*;
 
 use std::collections::HashMap;
@@ -1474,16 +1477,114 @@ pub fn format_duration(duration: i64) -> String {
 }
 
 fn format_filesize(num_bytes: i64, config: &Config) -> String {
+    // Allow the user to specify how they want their numbers formatted
+    let filesize_format_var = get_config_filesize_format(config);
+
     let byte = byte_unit::Byte::from_bytes(num_bytes as u128);
+    let adj_byte =
+        if filesize_format_var.0 == byte_unit::ByteUnit::B && filesize_format_var.1 == "auto" {
+            byte.get_appropriate_unit(!config.filesize_metric)
+        } else {
+            byte.get_adjusted_unit(filesize_format_var.0)
+        };
 
-    if byte.get_bytes() == 0u128 {
+    match adj_byte.get_unit() {
-        return "—".to_string();
+        byte_unit::ByteUnit::B => {
-    }
+            let locale_string = get_locale().unwrap_or_else(|| String::from("en-US"));
-
+            // Since get_locale() and Locale::from_name() don't always return the same items
-    let byte = byte.get_appropriate_unit(config.filesize_metric);
+            // we need to try and parse it to match. For instance, a valid locale is de_DE
-
+            // however Locale::from_name() wants only de so we split and parse it out.
-    match byte.get_unit() {
+            let locale_string = locale_string.replace("_", "-"); // en_AU -> en-AU
-        byte_unit::ByteUnit::B => format!("{} B ", byte.get_value()),
+            let locale = match Locale::from_name(&locale_string) {
-        _ => byte.format(1),
+                Ok(loc) => loc,
+                _ => {
+                    let all = num_format::Locale::available_names();
+                    let locale_prefix = &locale_string.split('-').collect::<Vec<&str>>();
+                    if all.contains(&locale_prefix[0]) {
+                        // eprintln!("Found alternate: {}", &locale_prefix[0]);
+                        Locale::from_name(locale_prefix[0]).unwrap_or(Locale::en)
+                    } else {
+                        // eprintln!("Unable to find matching locale. Defaulting to en-US");
+                        Locale::en
                     }
                 }
+            };
+            let locale_byte = adj_byte.get_value() as u64;
+            let locale_byte_string = locale_byte.to_formatted_string(&locale);
+
+            if filesize_format_var.1 == "auto" {
+                format!("{} B", locale_byte_string)
+            } else {
+                locale_byte_string
+            }
+        }
+        _ => adj_byte.format(1),
+    }
+}
+
+fn get_config_filesize_format(config: &Config) -> (ByteUnit, &str) {
+    // We need to take into account config.filesize_metric so, if someone asks for KB
+    // filesize_metric is true, return KiB
+    let filesize_format = match config.filesize_format.as_str() {
+        "b" => (byte_unit::ByteUnit::B, ""),
+        "kb" => {
+            if config.filesize_metric {
+                (byte_unit::ByteUnit::KiB, "")
+            } else {
+                (byte_unit::ByteUnit::KB, "")
+            }
+        }
+        "kib" => (byte_unit::ByteUnit::KiB, ""),
+        "mb" => {
+            if config.filesize_metric {
+                (byte_unit::ByteUnit::MiB, "")
+            } else {
+                (byte_unit::ByteUnit::MB, "")
+            }
+        }
+        "mib" => (byte_unit::ByteUnit::MiB, ""),
+        "gb" => {
+            if config.filesize_metric {
+                (byte_unit::ByteUnit::GiB, "")
+            } else {
+                (byte_unit::ByteUnit::GB, "")
+            }
+        }
+        "gib" => (byte_unit::ByteUnit::GiB, ""),
+        "tb" => {
+            if config.filesize_metric {
+                (byte_unit::ByteUnit::TiB, "")
+            } else {
+                (byte_unit::ByteUnit::TB, "")
+            }
+        }
+        "tib" => (byte_unit::ByteUnit::TiB, ""),
+        "pb" => {
+            if config.filesize_metric {
+                (byte_unit::ByteUnit::PiB, "")
+            } else {
+                (byte_unit::ByteUnit::PB, "")
+            }
+        }
+        "pib" => (byte_unit::ByteUnit::PiB, ""),
+        "eb" => {
+            if config.filesize_metric {
+                (byte_unit::ByteUnit::EiB, "")
+            } else {
+                (byte_unit::ByteUnit::EB, "")
+            }
+        }
+        "eib" => (byte_unit::ByteUnit::EiB, ""),
+        "zb" => {
+            if config.filesize_metric {
+                (byte_unit::ByteUnit::ZiB, "")
+            } else {
+                (byte_unit::ByteUnit::ZB, "")
+            }
+        }
+        "zib" => (byte_unit::ByteUnit::ZiB, ""),
+        _ => (byte_unit::ByteUnit::B, "auto"),
+    };
+
+    filesize_format
+}

src/tests.rs

@@ -1148,19 +1148,28 @@ fn multi_word_imports() -> TestResult {
 }
 
 #[test]
-fn config_var_1() -> TestResult {
+fn config_filesize_format_with_metric_true() -> TestResult {
     // Note: this tests both the config variable and that it is properly captured into a block
     run_test(
-        r#"let config = {"filesize_metric": $true }; do { 40kb | into string } "#,
+        r#"let config = {"filesize_metric": $true "filesize_format": "kib" }; do { 40kb | into string } "#,
         "39.1 KiB",
     )
 }
 
 #[test]
-fn config_var_2() -> TestResult {
+fn config_filesize_format_with_metric_false_kib() -> TestResult {
     // Note: this tests both the config variable and that it is properly captured into a block
     run_test(
-        r#"let config = {"filesize_metric": $false }; do { 40kb | into string } "#,
+        r#"let config = {"filesize_metric": $false "filesize_format": "kib" }; do { 40kb | into string } "#,
+        "39.1 KiB",
+    )
+}
+
+#[test]
+fn config_filesize_format_with_metric_false_kb() -> TestResult {
+    // Note: this tests both the config variable and that it is properly captured into a block
+    run_test(
+        r#"let config = {"filesize_metric": $false "filesize_format": "kb" }; do { 40kb | into string } "#,
         "40.0 KB",
     )
 }