Add und-only baked testdata to repo and make cargo make quick quick…

… again (#4486)

With some basic instructions on how to use it with VSCode.

.github/workflows/build-test.yml

@@ -288,6 +288,8 @@ jobs:
       with:
         key: download-cache
         path: /tmp/icu4x-source-cache
+    - name: Install rustfmt
+      run: rustup component add rustfmt
 
     # Actual job
     - name: Run `cargo make ci-job-testdata`

CONTRIBUTING.md

@@ -10,7 +10,9 @@ Issues are open to everyone to discuss and can be used to jump-start Pull Reques
 
 In most cases, the first step is to find or file a new issue related to your planned contribution, discuss it, and once you received a feedback indicating that the pull request would be welcomed you can start working on it.
 
-## Installing dependencies
+## Development Environment
+
+### Installing dependencies
 
 To build ICU4X, you will need the following dependencies:
 
@@ -20,6 +22,26 @@ To build ICU4X, you will need the following dependencies:
 
 Certain tests may need further dependencies, these are documented below in the [Testing](#testing) section.
 
+### IDE setup
+
+ICU4X can be edited using any text editor capable of editing Rust code.
+
+#### Visual Studio Code
+
+Many ICU4X engineers use [Visual Studio Code](https://code.visualstudio.com/) with the [rust-analyzer](https://marketplace.visualstudio.com/items?itemName=rust-lang.rust-analyzer) extension.
+
+To build all code paths and improve build times in VSCode, we recommend the following settings. (Note: The second setting causes VSCode to build ICU4X with only the `und` locale, which reduces build times but also makes some tests fail; to run them normally, run `cargo test --all-features` on the command line.) To add these settings, choose "Preferences: Open Workspace Settings (JSON)" from the command palette (Ctrl+Shift+P):
+
+```javascript
+"settings": {
+	"rust-analyzer.cargo.features": "all",
+	"rust-analyzer.cargo.extraEnv": {
+		// Path relative to `provider/baked/*/src/lib.rs`
+		"ICU4X_DATA_DIR": "../../../../provider/datagen/tests/data/baked"
+	}
+}
+```
+
 ## Contributing a Pull Request
 
 The first step is to fork the repository to your namespace and create a branch off of the `main` branch to work with.

Makefile.toml

@@ -26,13 +26,12 @@ ICU4X_BUILDING_WITH_FORCED_NIGHTLY = { value = "1", condition = { env_set = ["IC
 [tasks.quick]
 description = "Run quick version of all lints and builds (useful before pushing to GitHub)"
 category = "ICU4X Development"
-dependencies = [
+env = { "ICU4X_DATA_DIR" = "${CARGO_MAKE_WORKSPACE_WORKING_DIRECTORY}/provider/datagen/tests/data/baked" }
+run_task.name = [
     "ci-job-fmt",
     "ci-job-clippy",
-    "ci-job-nostd",
     "check-no-features",
-    "check-tutorials-local",
-    "ci-job-tidy",
+    "depcheck",
     "ci-job-doc",
 ]
 

components/calendar/src/provider.rs

@@ -39,6 +39,7 @@ pub struct Baked;
 const _: () = {
     pub mod icu {
         pub use crate as calendar;
+        #[allow(unused_imports)] // baked data may or may not need this
         pub use icu_locid_transform as locid_transform;
     }
     icu_calendar_data::make_provider!(Baked);

components/collator/src/provider.rs

@@ -54,6 +54,7 @@ const _: () = {
     pub mod icu {
         pub use crate as collator;
         pub use icu_collections as collections;
+        #[allow(unused_imports)] // baked data may or may not need this
         pub use icu_locid_transform as locid_transform;
     }
     icu_collator_data::make_provider!(Baked);

components/datetime/src/provider/mod.rs

@@ -48,6 +48,7 @@ pub struct Baked;
 const _: () = {
     pub mod icu {
         pub use crate as datetime;
+        #[allow(unused_imports)] // baked data may or may not need this
         pub use icu_locid_transform as locid_transform;
     }
     icu_datetime_data::make_provider!(Baked);

components/decimal/src/provider.rs

@@ -34,6 +34,7 @@ pub struct Baked;
 const _: () = {
     pub mod icu {
         pub use crate as decimal;
+        #[allow(unused_imports)] // baked data may or may not need this
         pub use icu_locid_transform as locid_transform;
     }
     icu_decimal_data::make_provider!(Baked);

components/list/src/provider/mod.rs

@@ -38,6 +38,7 @@ pub struct Baked;
 const _: () = {
     pub mod icu {
         pub use crate as list;
+        #[allow(unused_imports)] // baked data may or may not need this
         pub use icu_locid_transform as locid_transform;
     }
     icu_list_data::make_provider!(Baked);

components/plurals/src/provider.rs

@@ -34,6 +34,7 @@ pub struct Baked;
 const _: () = {
     pub mod icu {
         pub use crate as plurals;
+        #[allow(unused_imports)] // baked data may or may not need this
         pub use icu_locid_transform as locid_transform;
     }
     icu_plurals_data::make_provider!(Baked);

components/properties/src/provider.rs

@@ -47,6 +47,7 @@ const _: () = {
     pub mod icu {
         pub use crate as properties;
         pub use icu_collections as collections;
+        #[allow(unused_imports)] // baked data may or may not need this
         pub use icu_locid_transform as locid_transform;
     }
     icu_properties_data::make_provider!(Baked);

experimental/compactdecimal/src/provider.rs

@@ -31,6 +31,7 @@ pub struct Baked;
 const _: () = {
     pub mod icu {
         pub use crate as compactdecimal;
+        #[allow(unused_imports)] // baked data may or may not need this
         pub use icu_locid_transform as locid_transform;
     }
     icu_compactdecimal_data::make_provider!(Baked);

experimental/displaynames/src/provider.rs

@@ -29,6 +29,7 @@ pub struct Baked;
 const _: () = {
     pub mod icu {
         pub use crate as displaynames;
+        #[allow(unused_imports)] // baked data may or may not need this
         pub use icu_locid_transform as locid_transform;
     }
     icu_displaynames_data::make_provider!(Baked);

experimental/relativetime/src/provider.rs

@@ -28,6 +28,7 @@ pub struct Baked;
 const _: () = {
     pub mod icu {
         pub use crate as relativetime;
+        #[allow(unused_imports)] // baked data may or may not need this
         pub use icu_locid_transform as locid_transform;
     }
     icu_relativetime_data::make_provider!(Baked);

provider/datagen/src/tests/make_testdata.rs

@@ -2,6 +2,7 @@
 // called LICENSE at the top level of the ICU4X source tree
 // (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
 
+use crate::baked_exporter;
 use crate::prelude::*;
 use crlify::BufWriterWithLineEndingFix;
 use icu_provider::datagen::*;
@@ -46,6 +47,18 @@ fn generate_json_and_verify_postcard() {
         ),
     });
 
+    let stubdata_out = Box::new(BakedStubdataExporter(
+        baked_exporter::BakedExporter::new(
+            "tests/data/baked".into(),
+            baked_exporter::Options {
+                overwrite: true,
+                pretty: true,
+                ..Default::default()
+            },
+        )
+        .unwrap(),
+    ));
+
     DatagenDriver::new()
         .with_keys(crate::all_keys())
         .with_locales(LOCALES.iter().cloned())
@@ -55,11 +68,67 @@ fn generate_json_and_verify_postcard() {
         ])
         .export(
             &DatagenProvider::new_testing(),
-            MultiExporter::new(vec![json_out, postcard_out]),
+            MultiExporter::new(vec![json_out, postcard_out, stubdata_out]),
         )
         .unwrap();
 }
 
+/// Generates a stub data directory that can be used with `ICU4X_DATA_DIR`
+/// for faster development and debugging. For example, put the following in
+/// VSCode settings.json:
+///
+/// ```javascript
+/// "rust-analyzer.cargo.extraEnv": {
+///   // Relative to provider/baked/x/src
+///   "ICU4X_DATA_DIR": "../../../datagen/tests/data/stub"
+/// },
+/// ```
+struct BakedStubdataExporter(baked_exporter::BakedExporter);
+
+impl DataExporter for BakedStubdataExporter {
+    fn put_payload(
+        &self,
+        key: DataKey,
+        locale: &DataLocale,
+        payload: &DataPayload<ExportMarker>,
+    ) -> Result<(), DataError> {
+        // put `und-*` but not any other locales
+        if locale.is_langid_und() {
+            self.0.put_payload(key, locale, payload)
+        } else {
+            Ok(())
+        }
+    }
+
+    fn flush_singleton(
+        &self,
+        key: DataKey,
+        payload: &DataPayload<ExportMarker>,
+    ) -> Result<(), DataError> {
+        self.0.flush_singleton(key, payload)
+    }
+
+    fn flush(&self, key: DataKey) -> Result<(), DataError> {
+        self.0.flush(key)
+    }
+
+    fn flush_with_built_in_fallback(
+        &self,
+        key: DataKey,
+        fallback_mode: BuiltInFallbackMode,
+    ) -> Result<(), DataError> {
+        self.0.flush_with_built_in_fallback(key, fallback_mode)
+    }
+
+    fn close(&mut self) -> Result<(), DataError> {
+        self.0.close()
+    }
+
+    fn supports_built_in_fallback(&self) -> bool {
+        self.0.supports_built_in_fallback()
+    }
+}
+
 struct PostcardTestingExporter {
     size_hash: Mutex<BTreeMap<(DataKey, String), (usize, u64)>>,
     zero_copy_violations: Mutex<BTreeSet<DataKey>>,