Pre-generate src/tables.rs: Rm all build-deps & build.rs

.github/workflows/test.yml

@@ -13,5 +13,18 @@ jobs:
           toolchain: stable
           profile: minimal
           override: true
+
+      - name: Configure src/tables.rs cache
+        id: cache-tables
+        uses: actions/cache@v3
+        with:
+          path: src/tables.rs
+          key: ${{ hashFiles('LineBreak.txt', 'gen-tables/**') }}
+
+      - name: Generates src/tables.rs
+        run: cargo run
+        working-directory: ./gen-tables
+        if: steps.cache-tables.outputs.cache-hit != 'true'
+
       - name: Run tests
-        run: cargo test
+        run: cargo test

.gitignore

@@ -1,2 +1,3 @@
 /target
-Cargo.lock
+/Cargo.lock
+/src/tables.rs

Cargo.toml

@@ -9,10 +9,6 @@ readme = "README.md"
 keywords = ["unicode", "text", "layout"]
 categories = ["internationalization"]
 license = "Apache-2.0"
-include = ["src/**/*", "build.rs", "LineBreak.txt", "LICENSE"]
+include = ["src/**/*", "LICENSE"]
 edition = "2021"
 rust-version = "1.56"
-
-[build-dependencies]
-regex = "1"
-hashbrown = "0.12.3"

README.md

@@ -21,4 +21,16 @@ assert!(linebreaks(text).eq([
 ]));
 ```
 
+## Development
+
+After cloning the repository or modifying `LineBreak.txt` the tables
+have to be (re-)generated:
+
+```sh
+# Generate src/tables.rs
+(cd gen-tables && cargo run)
+# Run tests to make sure it was successful
+cargo test
+```
+
 [UAX14]: https://www.unicode.org/reports/tr14/

gen-tables/.gitignore

@@ -0,0 +1 @@
+target

gen-tables/Cargo.toml

@@ -0,0 +1,13 @@
+[package]
+name = "gen-tables"
+version = "0.0.0"
+edition = "2021"
+publish = false
+
+[dependencies]
+regex = "1"
+hashbrown = "0.14"
+
+# Prevent this from interfering with workspaces
+[workspace]
+members = ["."]

build.rs → gen-tables/src/main.rs

@@ -20,11 +20,10 @@ use std::hash::{BuildHasher, Hash, Hasher};
 use std::io::{BufRead, BufReader, BufWriter, Write};
 use std::marker::PhantomData;
 use std::ops::Range;
-use std::path::Path;
 use std::str::FromStr;
-use std::{env, error, iter};
+use std::{error, iter};
 
-include!("src/shared.rs");
+include!("../../src/shared.rs");
 
 impl FromStr for BreakClass {
     type Err = &'static str;
@@ -88,6 +87,7 @@ static BREAK_CLASS_TABLE: [&str; NUM_CLASSES] = [
 
 #[derive(Copy, Clone)]
 #[repr(u8)]
+#[allow(clippy::upper_case_acronyms)]
 enum ExtraState {
     ZWSP = sot + 1,
     OPSP,
@@ -683,8 +683,8 @@ struct CpTrie<T> {
 }
 
 fn main() -> Result<(), Box<dyn error::Error>> {
-    println!("cargo:rerun-if-changed=LineBreak.txt");
-    debug_assert!(NUM_STATES <= 0x3F, "too many states");
+    #[allow(clippy::assertions_on_constants)]
+    const _: () = debug_assert!(NUM_STATES <= 0x3F, "too many states");
 
     let pair_table = rules2table! {
         // Non-tailorable Line Breaking Rules
@@ -795,7 +795,7 @@ fn main() -> Result<(), Box<dyn error::Error>> {
 ;
 (?P<lb>\w{2,3}) # Line_Break property",
     )?;
-    let prop_ranges = BufReader::new(File::open("LineBreak.txt")?)
+    let prop_ranges = BufReader::new(File::open("../LineBreak.txt")?)
         .lines()
         .map(Result::unwrap)
         .filter(|l| !(l.starts_with('#') || l.is_empty()))
@@ -828,9 +828,7 @@ fn main() -> Result<(), Box<dyn error::Error>> {
         builder.build()
     };
 
-    let out_dir = env::var("OUT_DIR")?;
-    let dest_path = Path::new(&out_dir).join("tables.rs");
-    let mut stream = BufWriter::new(File::create(&dest_path)?);
+    let mut stream = BufWriter::new(File::create("../src/tables.rs")?);
     writeln!(
         stream,
         "const BREAK_PROP_TRIE_HIGH_START: u32 = {};

src/lib.rs

@@ -27,8 +27,7 @@ use core::iter::once;
 pub const UNICODE_VERSION: (u8, u8, u8) = (15, 0, 0);
 
 include!("shared.rs");
-
-include!(concat!(env!("OUT_DIR"), "/tables.rs"));
+include!("tables.rs");
 
 /// Returns the line break property of the specified code point.
 ///