Moving some transliteration datagen logic into the crate

Cargo.toml

@@ -179,7 +179,6 @@ icu_compactdecimal_data = { version = "~1.2.0", path = "experimental/compactdeci
 icu_displaynames_data = { version = "~1.2.0", path = "experimental/displaynames/data", default-features = false }
 icu_relativetime_data = { version = "1.2.0", path = "experimental/relativetime/data", default-features = false }
 icu_singlenumberformatter_data = { version = "~1.2.0", path = "experimental/single_number_formatter/data", default-features = false }
-icu_transliterate_data = { version = "0.0.0", path = "experimental/transliterate/data", default-features = false }
 
 # FFI
 icu_capi = { version = "~1.2.0", path = "ffi/diplomat", default-features = false }

docs/tutorials/testing/patch.toml

@@ -48,7 +48,6 @@ icu_compactdecimal_data = { path = "../../experimental/compactdecimal/data" }
 icu_displaynames_data = { path = "../../experimental/displaynames/data" }
 icu_relativetime_data = { path = "../../experimental/relativetime/data" }
 icu_singlenumberformatter_data = { path = "../../experimental/single_number_formatter/data"}
-icu_transliterate_data = { path = "../../experimental/transliterate/data" }
 
 # FFI
 icu_capi = { path = "../../ffi/diplomat" }

experimental/transliterate/Cargo.toml

@@ -36,8 +36,6 @@ log = "0.4"
 serde = { version = "1.0", features = ["derive"], optional = true, default-features = false }
 displaydoc = { version = "0.2.3", default-features = false }
 
-icu_transliterate_data = { workspace = true, optional = true }
-
 [target.'cfg(not(target_arch = "wasm32"))'.dev-dependencies]
 criterion = "0.4.0"
 
@@ -48,8 +46,8 @@ bench = false
 default = ["compiled_data"]
 std = []
 serde = ["dep:serde", "zerovec/serde", "icu_collections/serde"]
-datagen = ["serde", "dep:databake", "zerovec/databake", "icu_collections/databake", "std", "icu_properties", "icu_unicodeset_parser"]
-compiled_data = ["dep:icu_transliterate_data", "icu_normalizer/compiled_data", "icu_properties/compiled_data"]
+datagen = ["serde", "dep:databake", "zerovec/databake", "icu_collections/databake", "std", "icu_properties", "icu_unicodeset_parser", "icu_provider/datagen"]
+compiled_data = ["icu_normalizer/compiled_data", "icu_properties/compiled_data"]
 bench = []
 
 [[bench]]

experimental/transliterate/benches/bench.rs

@@ -8,6 +8,8 @@ use icu_locid::Locale;
 
 use icu_transliterate::Transliterator;
 
+include!("../tests/data/baked/mod.rs");
+
 struct BenchDataContent {
     pub num: usize,
     pub name: String,
@@ -32,7 +34,7 @@ fn bench_data_from_sources(locale_str: &str, source: &str) -> Vec<BenchDataConte
         .map(|(idx, input)| BenchDataContent {
             num: idx + 1,
             name: locale_str.to_string(),
-            translit: Transliterator::try_new(locale.clone()).unwrap(),
+            translit: Transliterator::try_new_unstable(locale.clone(), &BakedDataProvider).unwrap(),
             input,
         })
         .collect()
@@ -43,7 +45,7 @@ fn bench_data() -> Vec<BenchDataContent> {
     let mut bench_data = Vec::new();
     bench_data.extend(bench_data_from_sources(
         "und-Arab-t-und-beng",
-        include_str!("../tests/cldr_testData/und-Arab-t-und-beng.txt"),
+        include_str!("../tests/data/fixtures/und-Arab-t-und-beng.txt"),
     ));
 
     bench_data.into_iter().take(5).collect()