Replace build script with private crate

Closes GH-34.
Closes GH-35.

Author: wooorm

.github/workflows/main.yml

@@ -11,6 +11,7 @@ jobs:
         with:
           toolchain: stable
           components: rustfmt, clippy
+      - run: cargo run --manifest-path generate/Cargo.toml
       - run: cargo fmt --check && cargo clippy --examples --tests --benches
       - run: cargo test
   coverage:
@@ -20,5 +21,6 @@ jobs:
       - uses: actions-rs/toolchain@v1
         with:
           toolchain: stable
+      - run: cargo run --manifest-path generate/Cargo.toml
       - run: cargo install cargo-tarpaulin && cargo tarpaulin --out Xml
       - uses: codecov/codecov-action@v3

Cargo.toml

@@ -21,7 +21,6 @@ harness = false
 log = "0.4"
 unicode-id = { version = "0.3", features = ["no_std"] }
 
-
 [dev-dependencies]
 env_logger = "0.10"
 criterion = "0.4"
@@ -33,7 +32,5 @@ swc_core = { version = "0.43.30", features = [
   "common",
 ] }
 
-[build-dependencies]
-regex = "1"
-reqwest = "0.11"
-tokio = { version = "1", features = ["full"] }
+[workspace]
+members = ["generate"]

generate/Cargo.toml

@@ -0,0 +1,11 @@
+[package]
+name = "markdown-generate"
+version = "0.0.0"
+authors = ["Titus Wormer <[email protected]>"]
+edition = "2018"
+publish = false
+
+[dependencies]
+regex = "1"
+reqwest = "0.11"
+tokio = { version = "1", features = ["full"] }

build.rs renamed to generate/src/main.rs

@@ -1,3 +1,9 @@
+// To regenerate, run the following from the repository root:
+//
+// ```sh
+// cargo run --manifest-path generate/Cargo.toml
+// ```
+
 use regex::Regex;
 use std::fs;
 
@@ -63,7 +69,7 @@ async fn commonmark() {
     let doc = format!(
         "//! `CommonMark` test suite.
 
-// > 👉 **Important**: this module is generated by `build.rs`.
+// > 👉 **Important**: this module is generated by `generate/src/main.rs`.
 // > It is generate from the latest CommonMark website.
 
 use markdown::{{to_html_with_options, CompileOptions, Options}};
@@ -130,7 +136,7 @@ async fn punctuation() {
 
 /// List of characters that are considered punctuation.
 ///
-/// > 👉 **Important**: this module is generated by `build.rs`.
+/// > 👉 **Important**: this module is generated by `generate/src/main.rs`.
 /// > It is generate from the latest Unicode data.
 ///
 /// Rust does not contain an `is_punctuation` method on `char`, while it does

readme.md

@@ -241,6 +241,10 @@ Fuzz testing is used to check for things that might fall through coverage.
 
 The following bash scripts are useful when working on this project:
 
+*   generate code (latest CM tests and Unicode info):
+    ```sh
+    cargo run --manifest-path generate/Cargo.toml
+    ```
 *   run examples:
     ```sh
     RUST_BACKTRACE=1 RUST_LOG=debug cargo run --example lib

src/util/unicode.rs

@@ -2,7 +2,7 @@
 
 /// List of characters that are considered punctuation.
 ///
-/// > 👉 **Important**: this module is generated by `build.rs`.
+/// > 👉 **Important**: this module is generated by `generate/src/main.rs`.
 /// > It is generate from the latest Unicode data.
 ///
 /// Rust does not contain an `is_punctuation` method on `char`, while it does

tests/commonmark.rs

@@ -1,6 +1,6 @@
 //! `CommonMark` test suite.
 
-// > 👉 **Important**: this module is generated by `build.rs`.
+// > 👉 **Important**: this module is generated by `generate/src/main.rs`.
 // > It is generate from the latest CommonMark website.
 
 use markdown::{to_html_with_options, CompileOptions, Options};