Extract language definitions into data file

.github/CODEOWNERS

@@ -1,7 +1,7 @@
 /.github/ @spenserblack @o2sh
 /src/ui/image_backends/ @CephalonRho @yoichi
 /src/info/repo.rs @Byron
-/src/info/langs/ @spenserblack
+/languages.yaml @spenserblack @o2sh
+/templates/language.rs @spenserblack
 /src/info/deps/ @HallerPatrick @o2sh
 /src/cli.rs @spenserblack @o2sh
-*.ascii @o2sh

CONTRIBUTING.md

@@ -1,53 +1,55 @@
-# How to Contribute
-
-- [How to Contribute](#how-to-contribute)
-  - [Contributing via Pull Requests](#contributing-via-pull-requests)
-    - [Adding support for a new language](#adding-support-for-a-new-language)
-      - [Ascii logo](#ascii-logo)
-    - [Adding support for a new package manager](#adding-support-for-a-new-package-manager)
-    - [Adding translation for README.md](#adding-translation-for-readmemd)
-
-## Contributing via Pull Requests
-
 ### Adding support for a new language
 
-Adding support for a new Language consists in adding a new entry to the `define_language!` macro in [language.rs](src/info/langs/language.rs) and filling it in with the right data.
+Adding support for a new Language consists in adding a new entry to [language.yaml](./languages.yaml) and filling it in with the right data.
 
 **Example**:
 
-`{ TypeScript, Programming, "typescript.ascii", define_colors!([Cyan, White] : [Rgb(0, 122, 204), Rgb(255, 255, 255)]), Rgb(43, 116, 137) }`
-
-The first item `TypeScript` corresponds to the name of the language as defined in [tokei](https://github.com/XAMPPRocky/tokei). The second item refers to the language type as specified by [linguist](https://github.com/github/linguist/blob/master/lib/linguist/languages.yml), only four values are possible: Programming, Markup, Prose and Data. The third item `typescript.ascii` is the name of the file containing the ascii logo: this file has to be placed in the _./resources_ folder (more info below). Then we have the colors used to customize the look of the ascii logo when displayed to the screen. Finally, the circle color used in the language distribution whose hex value can be found in [linguist](https://github.com/github/linguist/blob/master/lib/linguist/languages.yml).
-
-#### Ascii logo
-
-```text
-{0}TSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTS
-{0}TSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTS
-{0}TSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTS
-{0}TSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTS
-{0}TSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTS
-{0}TSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTS
-{0}TSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTS
-{0}TSTSTSTS{1}TSTSTSTSTSTSTS{0}TSTS{1}TSTSTS{0}TSTSTS
-{0}TSTSTSTS{1}TSTSTSTSTSTSTS{0}TS{1}TSTSTSTSTS{0}TSTS
-{0}TSTSTSTSTSTST{1}TSTS{0}TSTSTST{1}TSTST{0}TSTSTSTST
-{0}TSTSTSTSTSTST{1}TSTS{0}TSTSTSTS{1}TSTST{0}STSTSTST
-{0}TSTSTSTSTSTST{1}TSTS{0}TSTSTSTSTS{1}TSTST{0}STSTST
-{0}TSTSTSTSTSTST{1}TSTS{0}TSTSTSTSTSTS{1}TSTST{0}TSTS
-{0}TSTSTSTSTSTST{1}TSTS{0}TSTSTSTSTSTST{1}TSTST{0}TST
-{0}TSTSTSTSTSTST{1}TSTS{0}TSTSTSTS{1}TSTSTSTSTS{0}TST
-{0}TSTSTSTSTSTST{1}TSTS{0}TSTSTSTSTS{1}TSTSTS{0}STSTS
-{0}TSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTS
-{0}TSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTSTS
+```
+CSharp: # required, this will be the name of the enum variant for the language as specified by tokei (link 1)
+  type: programming # required, can be programming, data, markup, or prose as specified by linguist (link 2)
+
+  # required, this is the logo. If it's not within 25x40 bounds, you will get a compiler error. Use `{i}` to color the ascii with `i` the color index.
+  ascii: |
+    {0}                 ++++++
+    {0}              ++++++++++++
+    {0}          ++++++++++++++++++++
+    {0}       ++++++++++++++++++++++++++
+    {0}    ++++++++++++++++++++++++++++++++
+    {0} +++++++++++++{3}************{0}+++++++++++++
+    {0}+++++++++++{3}******************{0}++++++++{2};;;
+    {0}+++++++++{3}**********************{0}++{2};;;;;;;
+    {0}++++++++{3}*********{0}++++++{3}******{2};;;;;;;;;;;
+    {0}+++++++{3}********{0}++++++++++{3}**{2};;;{3}**{2};;;{3}**{2};;;
+    {0}+++++++{3}*******{0}+++++++++{2};;;;;;{3}*********{2}::
+    {0}+++++++{3}******{0}+++++++{2};;;;;;;;;;{3}**{2};;;{3}**{2};;;
+    {0}+++++++{3}*******{0}+++{1}:::::{2};;;;;;;{3}*********{2};;
+    {0}+++++++{3}********{1}::::::::::{3}**{2};;;{3}**{2};;;{3}**{2};;;
+    {0}++++++++{3}*********{1}::::::{3}******{2};;;;;;;;;;;
+    {0}++++++{1}:::{3}**********************{1}::{2};;;;;;;
+    {0}+++{1}::::::::{3}******************{1}::::::::{2};;;
+    {1} :::::::::::::{3}************{1}:::::::::::::
+    {1}    ::::::::::::::::::::::::::::::::
+    {1}       ::::::::::::::::::::::::::
+    {1}          ::::::::::::::::::::
+    {1}              ::::::::::::
+    {1}                 ::::::
+  colors:
+    ansi: # required, a list of the ANSI colors used to colorize the logo
+      - blue
+      - magenta
+      - magenta
+      - white
+    hex: # optional, alternative to basic colors for terminals that support true colour.
+      - '#9B4F97'
+      - '#67217A'
+      - '#803788'
+      - '#FFFFFF'
+    chip: '#178600' # required, this is used for the language breakdown bar, its value can be found in linguist (link 2).
+  serialization: c# # required only if the Enum name `CSharp` doesn't match the display name `C#`
 ```
 
-Remarks:
-
-- The ascii logo's dimensions must fall below `25*40` (height\*width); The CI will fail otherwise.
-- Use `{i}` to color the ascii with `i` the color index from the `vec!` of colors defined in `define_language!`.
-- Make sure to trim any unnecessary trailing whitespaces.
-- Optionally, you may provide a `vec!` of colors in `rgb` format as an alternative to basic colors for terminals that support [true colour](https://gist.github.com/XVilka/8346728).
+- link 1: https://github.com/XAMPPRocky/tokei#supported-languages
+- link 2: https://github.com/github/linguist/blob/master/lib/linguist/languages.yml
 
 ### Adding support for a new package manager
 

Cargo.toml

@@ -59,9 +59,12 @@ ansi_term = "0.12.1"
 base64 = "0.13.0"
 libc = "0.2.126"
 
-[dev-dependencies]
-more-asserts = "0.3.0"
-paste = "1.0.7"
+[build-dependencies]
+lazy_static = "1"
+regex = "1"
+serde_json = "1"
+serde_yaml = "0.8"
+tera = { version = "1", default-features = false }
 
 [features]
 fail-on-deprecated = []

build.rs

@@ -0,0 +1,74 @@
+use lazy_static::lazy_static;
+use regex::Regex;
+use std::collections::HashMap;
+use std::env;
+use std::error::Error;
+use std::fs::{self, File};
+use std::path::Path;
+use tera::{Context, Tera};
+
+fn main() -> Result<(), Box<dyn Error>> {
+    let out_dir = env::var("OUT_DIR").expect("No OUT_DIR variable.");
+    let mut tera = Tera::default();
+    tera.register_filter("strip_color_tokens", strip_color_tokens_filter);
+    tera.register_filter("hex_to_rgb", hex_to_rgb_filter);
+
+    let lang_data: serde_json::Value = serde_yaml::from_reader(File::open("languages.yaml")?)?;
+
+    let output_path = Path::new(&out_dir).join("language.rs");
+
+    let rust_code = tera.render_str(
+        &std::fs::read_to_string("src/info/langs/language.tera.rs")?,
+        &Context::from_value(serde_json::json!({ "languages": lang_data, }))?,
+    )?;
+    fs::write(&output_path, rust_code)?;
+
+    Ok(())
+}
+
+/// Strips out `{n}` from the given string.
+fn strip_color_tokens_filter(
+    value: &tera::Value,
+    _args: &HashMap<String, tera::Value>,
+) -> tera::Result<tera::Value> {
+    lazy_static! {
+        static ref COLOR_INDEX_REGEX: Regex = Regex::new(r"\{\d+\}").unwrap();
+    }
+    let s = match value {
+        tera::Value::String(s) => s,
+        _ => return Err(tera::Error::msg("expected string")),
+    };
+    return Ok(tera::Value::String(
+        COLOR_INDEX_REGEX.replace_all(s, "").to_string(),
+    ));
+}
+
+fn hex_to_rgb_filter(
+    value: &tera::Value,
+    _args: &HashMap<String, tera::Value>,
+) -> tera::Result<tera::Value> {
+    let hex_string = match value {
+        tera::Value::String(s) => s,
+        _ => return Err(tera::Error::msg("expected string")),
+    };
+    let hex_string = match hex_string.strip_prefix('#') {
+        Some(s) => s,
+        None => return Err(tera::Error::msg("expected hex string starting with `#`")),
+    };
+    if hex_string.len() != 6 {
+        return Err(tera::Error::msg("expected a 6 digit hex string"));
+    }
+    let channel_bytes = match u32::from_str_radix(hex_string, 16) {
+        Ok(n) => n,
+        Err(_) => return Err(tera::Error::msg("expected a valid hex string")),
+    };
+    let r = (channel_bytes >> 16) & 0xFF;
+    let g = (channel_bytes >> 8) & 0xFF;
+    let b = channel_bytes & 0xFF;
+
+    Ok(serde_json::json!({
+        "r": r,
+        "g": g,
+        "b": b,
+    }))
+}