chore: checkpoint before Python removal

vendor/icu_normalizer/.cargo-checksum.json

@@ -0,0 +1 @@
+{"files":{".cargo_vcs_info.json":"75992dd22cc724a459b542a67a921e80396f5849eb42d9ebf059a39654f9a9d6","Cargo.lock":"6103b6ae33368ba7383a9c85d2534b26c907012434faf758006fea61706aed6a","Cargo.toml":"24273bcc1f268b0a559f70d25e7278cae2fdbb2aa8720876e713953bfb9d568b","Cargo.toml.orig":"1c8a5ef7cda3f66eb283bae506e5750329376558c29be6fefcbe60979d3fea3b","LICENSE":"f367c1b8e1aa262435251e442901da4607b4650e0e63a026f5044473ecfb90f2","README.md":"aec56e279d7e40a901b47a2eccb52197fde6c9499011b349c5ef509363bee6a9","benches/bench.rs":"c863372f937f0bf3c5cf214dd3729a88d450fa67f5d054753c2d9ea31fa84286","benches/canonical_composition.rs":"0aa91d5d400f58da61865f5fabe878c8506e60466c78503f77041ef7257e6dbe","benches/canonical_decomposition.rs":"3b44b8f832e426e8c82e449743117182ab7b138288001b621ccc9325b4c27b6c","benches/composing_normalizer_nfc.rs":"9a7aaae94e0096ccac9f3d1a83585c3f449af87f9f0f8b05615d2a010078e3e8","benches/composing_normalizer_nfkc.rs":"ad92d562a1e9aad3611521526882e1896aa436d2ac59493c8c00686c57bdf31e","benches/data/README.md":"fa79b84815a228c3fbfa5d4c6d12885036994ca8ad61e683b2113cf2b428bb85","benches/data/TestNames_Japanese_h.txt":"6522f8ed794ad348c904079082ec3aa303ae7acf3f68bbc49fa0ee90eebf31e0","benches/data/TestNames_Japanese_k.txt":"e4e18804fe742ecd27ae48bc3564c6bc653180a3c649d43a2ab4d8b7f2607627","benches/data/TestNames_Korean.txt":"9cbf54d5ee16726c0fc9477366e273ba1b82e651c9e88e6c7532df5344f03920","benches/data/TestNames_Latin.txt":"3a30d450d259a6be4a6aee8eeef08d3767d11fcc047b8f58060c542efe1182d1","benches/data/TestNames_Thai.txt":"28d76ddb62d6f47646232860fce7440544f402158443889393fd7e8bf10e9c3d","benches/data/TestRandomWordsUDHR_ar.txt":"02a775153e9746ae938a9db0b60244f2c00d911bb72b611a3593b0991fd95723","benches/data/TestRandomWordsUDHR_de.txt":"100b9502e7ddcb2fcbd055cb7ec9113245105bd1c606cace5e5bc147cc18727b","benches/data/TestRandomWordsUDHR_el.txt":"d1a2f0f9efc9ce663026ca7c285177391937c90008479a8c5b909c300dc86972","benches/data/TestRandomWordsUDHR_es.txt":"deeebda09e0ce0f80dd805317e96d1a630908601ff2a4d1ccb0021b00b55814b","benches/data/TestRandomWordsUDHR_fr.txt":"5931edc9f1af2c27a0b35c9624732e70b87b0fd72ab486710f3aa6367c7ad35f","benches/data/TestRandomWordsUDHR_he.txt":"dc77a89ffb9803e5c574d87f4789cb17624df73e40a8a92961df8ea8be103425","benches/data/TestRandomWordsUDHR_pl.txt":"26c378295ee2ef75ccacea691df0456394184a9a5c9ce48b2bada169b2402bbb","benches/data/TestRandomWordsUDHR_ru.txt":"a1c339f6d7b69cf9154e855c290ab09eeaf167ebcdf6d4bcb917de039fba10ee","benches/data/TestRandomWordsUDHR_th.txt":"3ba518be9863c85c3ac80cbb12299e3594e6f5afed3406d910d948007adaaf4e","benches/data/TestRandomWordsUDHR_tr.txt":"815c7babbc7228ef89b56f29638aeb63013aeca0003a49e58994e26b41cba01c","benches/data/wotw.txt":"8f28e68041ce75bbf75e72e186a6145e4c2de9e7e62b9b86ce0621c527a23669","benches/decomposing_normalizer_nfd.rs":"28f3d54c9af813af7ac9d0fbc9d45a7a6d27a25266bd593453eb35c1894280b5","benches/decomposing_normalizer_nfkd.rs":"cbaa2755878ee1cc90170210fddb7c79836457f89eb84f4f32fb51348f350bd5","benches/utf16_throughput.rs":"8b8065ccb2d31d7191ba4916e00c341bcc00df56fc7dac3d7a42de907a3c8d94","src/lib.rs":"46ea272d5821a752e29f6622b23b5549bf1b008775a0eb6c38bebdbe9c6191eb","src/properties.rs":"d4dcba8900c68ee08db4b00695edfbf83a7e325fb7f970bb96c40ff26afc3d80","src/provider.rs":"5850afc7ae842c7af74ce029be256944c64f5d0b51d95725a8366f5af22163e9","src/uts46.rs":"de86f63076dba6e5f95dfc99865db8ce84820190d54185b0380c737b8758247a","tests/data/NormalizationTest.txt":"1b04c22b82064adf871e76fd2148cd749129163f7d05bd7ace923516a65afe02","tests/data/README.md":"521fcd44a1f10f21629df88113fa29ca9f4e1dfbeea79fda19a7dc8ba435e24b","tests/tests.rs":"01db1c9dc1c7c71f80aed528e4309f416349af9eec887d2e438a3a11f2ee7f7c"},"package":"5f6c8828b67bf8908d82127b2054ea1b4427ff0230ee9141c54251934ab1b599"}

vendor/icu_normalizer/.cargo_vcs_info.json

@@ -0,0 +1,6 @@
+{
+  "git": {
+    "sha1": "38a49da495248dd1ded84cf306e4ca42e64d5bb3"
+  },
+  "path_in_vcs": "components/normalizer"
+}

vendor/icu_normalizer/Cargo.lock

@@ -0,0 +1,943 @@
+# This file is automatically @generated by Cargo.
+# It is not intended for manual editing.
+version = 3
+
+[[package]]
+name = "aho-corasick"
+version = "1.1.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8e60d3430d3a69478ad0993f19238d2df97c507009a52b3c10addcd7f6bcb916"
+dependencies = [
+ "memchr",
+]
+
+[[package]]
+name = "anes"
+version = "0.1.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4b46cbb362ab8752921c97e041f5e366ee6297bd428a31275b9fcf1e380f7299"
+
+[[package]]
+name = "anstyle"
+version = "1.0.13"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5192cca8006f1fd4f7237516f40fa183bb07f8fbdfedaa0036de5ea9b0b45e78"
+
+[[package]]
+name = "arraystring"
+version = "0.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4d517c467117e1d8ca795bc8cc90857ff7f79790cca0e26f6e9462694ece0185"
+dependencies = [
+ "typenum",
+]
+
+[[package]]
+name = "arrayvec"
+version = "0.7.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7c02d123df017efcdfbd739ef81735b36c5ba83ec3c59c80a9d7ecc718f92e50"
+
+[[package]]
+name = "atoi"
+version = "2.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f28d99ec8bfea296261ca1af174f24225171fea9664ba9003cbebee704810528"
+dependencies = [
+ "num-traits",
+]
+
+[[package]]
+name = "autocfg"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c08606f8c3cbf4ce6ec8e28fb0014a2c086708fe954eaa885384a6165172e7e8"
+
+[[package]]
+name = "bumpalo"
+version = "3.19.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "46c5e41b57b8bba42a04676d81cb89e9ee8e859a1a66f80a5a72e1cb76b34d43"
+
+[[package]]
+name = "cast"
+version = "0.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "37b2a672a2cb129a2e41c10b1224bb368f9f37a2b16b612598138befd7b37eb5"
+
+[[package]]
+name = "cfg-if"
+version = "1.0.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9330f8b2ff13f34540b44e946ef35111825727b38d33286ef986142615121801"
+
+[[package]]
+name = "ciborium"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "42e69ffd6f0917f5c029256a24d0161db17cea3997d185db0d35926308770f0e"
+dependencies = [
+ "ciborium-io",
+ "ciborium-ll",
+ "serde",
+]
+
+[[package]]
+name = "ciborium-io"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "05afea1e0a06c9be33d539b876f1ce3692f4afea2cb41f740e7743225ed1c757"
+
+[[package]]
+name = "ciborium-ll"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "57663b653d948a338bfb3eeba9bb2fd5fcfaecb9e199e87e1eda4d9e8b240fd9"
+dependencies = [
+ "ciborium-io",
+ "half",
+]
+
+[[package]]
+name = "clap"
+version = "4.4.18"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1e578d6ec4194633722ccf9544794b71b1385c3c027efe0c55db226fc880865c"
+dependencies = [
+ "clap_builder",
+]
+
+[[package]]
+name = "clap_builder"
+version = "4.4.18"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4df4df40ec50c46000231c914968278b1eb05098cf8f1b3a518a95030e71d1c7"
+dependencies = [
+ "anstyle",
+ "clap_lex",
+]
+
+[[package]]
+name = "clap_lex"
+version = "0.6.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "702fc72eb24e5a1e48ce58027a675bc24edd52096d5397d4aea7c6dd9eca0bd1"
+
+[[package]]
+name = "cobs"
+version = "0.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0fa961b519f0b462e3a3b4a34b64d119eeaca1d59af726fe450bbba07a9fc0a1"
+dependencies = [
+ "thiserror",
+]
+
+[[package]]
+name = "criterion"
+version = "0.5.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f2b12d017a929603d80db1831cd3a24082f8137ce19c69e6447f54f5fc8d692f"
+dependencies = [
+ "anes",
+ "cast",
+ "ciborium",
+ "clap",
+ "criterion-plot",
+ "is-terminal",
+ "itertools",
+ "num-traits",
+ "once_cell",
+ "oorandom",
+ "plotters",
+ "rayon",
+ "regex",
+ "serde",
+ "serde_derive",
+ "serde_json",
+ "tinytemplate",
+ "walkdir",
+]
+
+[[package]]
+name = "criterion-plot"
+version = "0.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6b50826342786a51a89e2da3a28f1c32b06e387201bc2d19791f622c673706b1"
+dependencies = [
+ "cast",
+ "itertools",
+]
+
+[[package]]
+name = "crossbeam-deque"
+version = "0.8.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9dd111b7b7f7d55b72c0a6ae361660ee5853c9af73f70c3c2ef6858b950e2e51"
+dependencies = [
+ "crossbeam-epoch",
+ "crossbeam-utils",
+]
+
+[[package]]
+name = "crossbeam-epoch"
+version = "0.9.18"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5b82ac4a3c2ca9c3460964f020e1402edd5753411d7737aa39c3714ad1b5420e"
+dependencies = [
+ "crossbeam-utils",
+]
+
+[[package]]
+name = "crossbeam-utils"
+version = "0.8.21"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d0a5c400df2834b80a4c3327b3aad3a4c4cd4de0629063962b03235697506a28"
+
+[[package]]
+name = "crunchy"
+version = "0.2.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "460fbee9c2c2f33933d720630a6a0bac33ba7053db5344fac858d4b8952d77d5"
+
+[[package]]
+name = "databake"
+version = "0.2.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ff6ee9e2d2afb173bcdeee45934c89ec341ab26f91c9933774fc15c2b58f83ef"
+dependencies = [
+ "databake-derive",
+ "proc-macro2",
+ "quote",
+]
+
+[[package]]
+name = "databake-derive"
+version = "0.2.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6834770958c7b84223607e49758ec0dde273c4df915e734aad50f62968a4c134"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+ "synstructure",
+]
+
+[[package]]
+name = "detone"
+version = "1.0.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5d5b580660e7375410c9199e84aa298f919925fb53d8cc9b02d8010ff5a14d09"
+
+[[package]]
+name = "displaydoc"
+version = "0.2.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "97369cbbc041bc366949bc74d34658d6cda5621039731c6310521892a3a20ae0"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "either"
+version = "1.15.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "48c757948c5ede0e46177b7add2e67155f70e33c07fea8284df6576da70b3719"
+
+[[package]]
+name = "erased-serde"
+version = "0.4.8"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "259d404d09818dec19332e31d94558aeb442fea04c817006456c24b5460bbd4b"
+dependencies = [
+ "serde",
+ "serde_core",
+ "typeid",
+]
+
+[[package]]
+name = "half"
+version = "2.4.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6dd08c532ae367adf81c312a4580bc67f1d0fe8bc9c460520283f4c0ff277888"
+dependencies = [
+ "cfg-if",
+ "crunchy",
+]
+
+[[package]]
+name = "hermit-abi"
+version = "0.5.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "fc0fef456e4baa96da950455cd02c081ca953b141298e41db3fc7e36b1da849c"
+
+[[package]]
+name = "icu_collections"
+version = "2.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4c6b649701667bbe825c3b7e6388cb521c23d88644678e83c0c4d0a621a34b43"
+dependencies = [
+ "databake",
+ "displaydoc",
+ "potential_utf",
+ "serde",
+ "yoke",
+ "zerofrom",
+ "zerovec",
+]
+
+[[package]]
+name = "icu_locale_core"
+version = "2.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "edba7861004dd3714265b4db54a3c390e880ab658fec5f7db895fae2046b5bb6"
+dependencies = [
+ "databake",
+ "displaydoc",
+ "litemap",
+ "serde",
+ "tinystr",
+ "writeable",
+ "zerovec",
+]
+
+[[package]]
+name = "icu_normalizer"
+version = "2.1.1"
+dependencies = [
+ "arraystring",
+ "arrayvec",
+ "atoi",
+ "criterion",
+ "databake",
+ "detone",
+ "icu_collections",
+ "icu_normalizer_data",
+ "icu_properties",
+ "icu_provider",
+ "serde",
+ "smallvec",
+ "utf16_iter",
+ "utf8_iter",
+ "write16",
+ "zerovec",
+]
+
+[[package]]
+name = "icu_normalizer_data"
+version = "2.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7aedcccd01fc5fe81e6b489c15b247b8b0690feb23304303a9e560f37efc560a"
+
+[[package]]
+name = "icu_properties"
+version = "2.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e93fcd3157766c0c8da2f8cff6ce651a31f0810eaa1c51ec363ef790bbb5fb99"
+dependencies = [
+ "databake",
+ "icu_collections",
+ "icu_locale_core",
+ "icu_properties_data",
+ "icu_provider",
+ "serde",
+ "zerotrie",
+ "zerovec",
+]
+
+[[package]]
+name = "icu_properties_data"
+version = "2.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "02845b3647bb045f1100ecd6480ff52f34c35f82d9880e029d329c21d1054899"
+
+[[package]]
+name = "icu_provider"
+version = "2.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "85962cf0ce02e1e0a629cc34e7ca3e373ce20dda4c4d7294bbd0bf1fdb59e614"
+dependencies = [
+ "databake",
+ "displaydoc",
+ "erased-serde",
+ "icu_locale_core",
+ "postcard",
+ "serde",
+ "stable_deref_trait",
+ "writeable",
+ "yoke",
+ "zerofrom",
+ "zerotrie",
+ "zerovec",
+]
+
+[[package]]
+name = "is-terminal"
+version = "0.4.17"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3640c1c38b8e4e43584d8df18be5fc6b0aa314ce6ebf51b53313d4306cca8e46"
+dependencies = [
+ "hermit-abi",
+ "libc",
+ "windows-sys",
+]
+
+[[package]]
+name = "itertools"
+version = "0.10.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b0fd2260e829bddf4cb6ea802289de2f86d6a7a690192fbe91b3f46e0f2c8473"
+dependencies = [
+ "either",
+]
+
+[[package]]
+name = "itoa"
+version = "1.0.15"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4a5f13b858c8d314ee3e8f639011f7ccefe71f97f96e50151fb991f267928e2c"
+
+[[package]]
+name = "js-sys"
+version = "0.3.81"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ec48937a97411dcb524a265206ccd4c90bb711fca92b2792c407f268825b9305"
+dependencies = [
+ "once_cell",
+ "wasm-bindgen",
+]
+
+[[package]]
+name = "libc"
+version = "0.2.177"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2874a2af47a2325c2001a6e6fad9b16a53b802102b528163885171cf92b15976"
+
+[[package]]
+name = "litemap"
+version = "0.8.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6373607a59f0be73a39b6fe456b8192fcc3585f602af20751600e974dd455e77"
+dependencies = [
+ "serde_core",
+]
+
+[[package]]
+name = "log"
+version = "0.4.28"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "34080505efa8e45a4b816c349525ebe327ceaa8559756f0356cba97ef3bf7432"
+
+[[package]]
+name = "memchr"
+version = "2.7.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f52b00d39961fc5b2736ea853c9cc86238e165017a493d1d5c8eac6bdc4cc273"
+
+[[package]]
+name = "num-traits"
+version = "0.2.19"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "071dfc062690e90b734c0b2273ce72ad0ffa95f0c74596bc250dcfd960262841"
+dependencies = [
+ "autocfg",
+]
+
+[[package]]
+name = "once_cell"
+version = "1.21.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "42f5e15c9953c5e4ccceeb2e7382a716482c34515315f7b03532b8b4e8393d2d"
+
+[[package]]
+name = "oorandom"
+version = "11.1.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d6790f58c7ff633d8771f42965289203411a5e5c68388703c06e14f24770b41e"
+
+[[package]]
+name = "plotters"
+version = "0.3.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5aeb6f403d7a4911efb1e33402027fc44f29b5bf6def3effcc22d7bb75f2b747"
+dependencies = [
+ "num-traits",
+ "plotters-backend",
+ "plotters-svg",
+ "wasm-bindgen",
+ "web-sys",
+]
+
+[[package]]
+name = "plotters-backend"
+version = "0.3.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "df42e13c12958a16b3f7f4386b9ab1f3e7933914ecea48da7139435263a4172a"
+
+[[package]]
+name = "plotters-svg"
+version = "0.3.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "51bae2ac328883f7acdfea3d66a7c35751187f870bc81f94563733a154d7a670"
+dependencies = [
+ "plotters-backend",
+]
+
+[[package]]
+name = "postcard"
+version = "1.1.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6764c3b5dd454e283a30e6dfe78e9b31096d9e32036b5d1eaac7a6119ccb9a24"
+dependencies = [
+ "cobs",
+ "serde",
+]
+
+[[package]]
+name = "potential_utf"
+version = "0.1.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b73949432f5e2a09657003c25bca5e19a0e9c84f8058ca374f49e0ebe605af77"
+dependencies = [
+ "serde_core",
+ "zerovec",
+]
+
+[[package]]
+name = "proc-macro2"
+version = "1.0.103"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5ee95bc4ef87b8d5ba32e8b7714ccc834865276eab0aed5c9958d00ec45f49e8"
+dependencies = [
+ "unicode-ident",
+]
+
+[[package]]
+name = "quote"
+version = "1.0.41"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ce25767e7b499d1b604768e7cde645d14cc8584231ea6b295e9c9eb22c02e1d1"
+dependencies = [
+ "proc-macro2",
+]
+
+[[package]]
+name = "rayon"
+version = "1.10.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b418a60154510ca1a002a752ca9714984e21e4241e804d32555251faf8b78ffa"
+dependencies = [
+ "either",
+ "rayon-core",
+]
+
+[[package]]
+name = "rayon-core"
+version = "1.12.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1465873a3dfdaa8ae7cb14b4383657caab0b3e8a0aa9ae8e04b044854c8dfce2"
+dependencies = [
+ "crossbeam-deque",
+ "crossbeam-utils",
+]
+
+[[package]]
+name = "regex"
+version = "1.12.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "843bc0191f75f3e22651ae5f1e72939ab2f72a4bc30fa80a066bd66edefc24d4"
+dependencies = [
+ "aho-corasick",
+ "memchr",
+ "regex-automata",
+ "regex-syntax",
+]
+
+[[package]]
+name = "regex-automata"
+version = "0.4.13"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5276caf25ac86c8d810222b3dbb938e512c55c6831a10f3e6ed1c93b84041f1c"
+dependencies = [
+ "aho-corasick",
+ "memchr",
+ "regex-syntax",
+]
+
+[[package]]
+name = "regex-syntax"
+version = "0.8.8"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7a2d987857b319362043e95f5353c0535c1f58eec5336fdfcf626430af7def58"
+
+[[package]]
+name = "rustversion"
+version = "1.0.22"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b39cdef0fa800fc44525c84ccb54a029961a8215f9619753635a9c0d2538d46d"
+
+[[package]]
+name = "ryu"
+version = "1.0.20"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "28d3b2b1366ec20994f1fd18c3c594f05c5dd4bc44d8bb0c1c632c8d6829481f"
+
+[[package]]
+name = "same-file"
+version = "1.0.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "93fc1dc3aaa9bfed95e02e6eadabb4baf7e3078b0bd1b4d7b6b0b68378900502"
+dependencies = [
+ "winapi-util",
+]
+
+[[package]]
+name = "serde"
+version = "1.0.228"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9a8e94ea7f378bd32cbbd37198a4a91436180c5bb472411e48b5ec2e2124ae9e"
+dependencies = [
+ "serde_core",
+ "serde_derive",
+]
+
+[[package]]
+name = "serde_core"
+version = "1.0.228"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "41d385c7d4ca58e59fc732af25c3983b67ac852c1a25000afe1175de458b67ad"
+dependencies = [
+ "serde_derive",
+]
+
+[[package]]
+name = "serde_derive"
+version = "1.0.228"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d540f220d3187173da220f885ab66608367b6574e925011a9353e4badda91d79"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "serde_json"
+version = "1.0.145"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "402a6f66d8c709116cf22f558eab210f5a50187f702eb4d7e5ef38d9a7f1c79c"
+dependencies = [
+ "itoa",
+ "memchr",
+ "ryu",
+ "serde",
+ "serde_core",
+]
+
+[[package]]
+name = "smallvec"
+version = "1.15.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "67b1b7a3b5fe4f1376887184045fcf45c69e92af734b7aaddc05fb777b6fbd03"
+
+[[package]]
+name = "stable_deref_trait"
+version = "1.2.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6ce2be8dc25455e1f91df71bfa12ad37d7af1092ae736f3a6cd0e37bc7810596"
+
+[[package]]
+name = "syn"
+version = "2.0.108"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "da58917d35242480a05c2897064da0a80589a2a0476c9a3f2fdc83b53502e917"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "unicode-ident",
+]
+
+[[package]]
+name = "synstructure"
+version = "0.13.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "728a70f3dbaf5bab7f0c4b1ac8d7ae5ea60a4b5549c8a5914361c99147a709d2"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "thiserror"
+version = "2.0.17"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f63587ca0f12b72a0600bcba1d40081f830876000bb46dd2337a3051618f4fc8"
+dependencies = [
+ "thiserror-impl",
+]
+
+[[package]]
+name = "thiserror-impl"
+version = "2.0.17"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3ff15c8ecd7de3849db632e14d18d2571fa09dfc5ed93479bc4485c7a517c913"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "tinystr"
+version = "0.8.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "42d3e9c45c09de15d06dd8acf5f4e0e399e85927b7f00711024eb7ae10fa4869"
+dependencies = [
+ "displaydoc",
+ "serde_core",
+ "zerovec",
+]
+
+[[package]]
+name = "tinytemplate"
+version = "1.2.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "be4d6b5f19ff7664e8c98d03e2139cb510db9b0a60b55f8e8709b689d939b6bc"
+dependencies = [
+ "serde",
+ "serde_json",
+]
+
+[[package]]
+name = "typeid"
+version = "1.0.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "bc7d623258602320d5c55d1bc22793b57daff0ec7efc270ea7d55ce1d5f5471c"
+
+[[package]]
+name = "typenum"
+version = "1.19.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "562d481066bde0658276a35467c4af00bdc6ee726305698a55b86e61d7ad82bb"
+
+[[package]]
+name = "unicode-ident"
+version = "1.0.20"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "462eeb75aeb73aea900253ce739c8e18a67423fadf006037cd3ff27e82748a06"
+
+[[package]]
+name = "utf16_iter"
+version = "1.0.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c8232dd3cdaed5356e0f716d285e4b40b932ac434100fe9b7e0e8e935b9e6246"
+
+[[package]]
+name = "utf8_iter"
+version = "1.0.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b6c140620e7ffbb22c2dee59cafe6084a59b5ffc27a8859a5f0d494b5d52b6be"
+
+[[package]]
+name = "walkdir"
+version = "2.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "29790946404f91d9c5d06f9874efddea1dc06c5efe94541a7d6863108e3a5e4b"
+dependencies = [
+ "same-file",
+ "winapi-util",
+]
+
+[[package]]
+name = "wasm-bindgen"
+version = "0.2.104"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c1da10c01ae9f1ae40cbfac0bac3b1e724b320abfcf52229f80b547c0d250e2d"
+dependencies = [
+ "cfg-if",
+ "once_cell",
+ "rustversion",
+ "wasm-bindgen-macro",
+ "wasm-bindgen-shared",
+]
+
+[[package]]
+name = "wasm-bindgen-backend"
+version = "0.2.104"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "671c9a5a66f49d8a47345ab942e2cb93c7d1d0339065d4f8139c486121b43b19"
+dependencies = [
+ "bumpalo",
+ "log",
+ "proc-macro2",
+ "quote",
+ "syn",
+ "wasm-bindgen-shared",
+]
+
+[[package]]
+name = "wasm-bindgen-macro"
+version = "0.2.104"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7ca60477e4c59f5f2986c50191cd972e3a50d8a95603bc9434501cf156a9a119"
+dependencies = [
+ "quote",
+ "wasm-bindgen-macro-support",
+]
+
+[[package]]
+name = "wasm-bindgen-macro-support"
+version = "0.2.104"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9f07d2f20d4da7b26400c9f4a0511e6e0345b040694e8a75bd41d578fa4421d7"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+ "wasm-bindgen-backend",
+ "wasm-bindgen-shared",
+]
+
+[[package]]
+name = "wasm-bindgen-shared"
+version = "0.2.104"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "bad67dc8b2a1a6e5448428adec4c3e84c43e561d8c9ee8a9e5aabeb193ec41d1"
+dependencies = [
+ "unicode-ident",
+]
+
+[[package]]
+name = "web-sys"
+version = "0.3.81"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9367c417a924a74cae129e6a2ae3b47fabb1f8995595ab474029da749a8be120"
+dependencies = [
+ "js-sys",
+ "wasm-bindgen",
+]
+
+[[package]]
+name = "winapi-util"
+version = "0.1.11"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c2a7b1c03c876122aa43f3020e6c3c3ee5c05081c9a00739faf7503aeba10d22"
+dependencies = [
+ "windows-sys",
+]
+
+[[package]]
+name = "windows-link"
+version = "0.2.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f0805222e57f7521d6a62e36fa9163bc891acd422f971defe97d64e70d0a4fe5"
+
+[[package]]
+name = "windows-sys"
+version = "0.61.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ae137229bcbd6cdf0f7b80a31df61766145077ddf49416a728b02cb3921ff3fc"
+dependencies = [
+ "windows-link",
+]
+
+[[package]]
+name = "write16"
+version = "1.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d1890f4022759daae28ed4fe62859b1236caebfc61ede2f63ed4e695f3f6d936"
+dependencies = [
+ "arrayvec",
+ "smallvec",
+]
+
+[[package]]
+name = "writeable"
+version = "0.6.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9edde0db4769d2dc68579893f2306b26c6ecfbe0ef499b013d731b7b9247e0b9"
+
+[[package]]
+name = "yoke"
+version = "0.8.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "72d6e5c6afb84d73944e5cedb052c4680d5657337201555f9f2a16b7406d4954"
+dependencies = [
+ "stable_deref_trait",
+ "yoke-derive",
+ "zerofrom",
+]
+
+[[package]]
+name = "yoke-derive"
+version = "0.8.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b659052874eb698efe5b9e8cf382204678a0086ebf46982b79d6ca3182927e5d"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+ "synstructure",
+]
+
+[[package]]
+name = "zerofrom"
+version = "0.1.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "50cc42e0333e05660c3587f3bf9d0478688e15d870fab3346451ce7f8c9fbea5"
+dependencies = [
+ "zerofrom-derive",
+]
+
+[[package]]
+name = "zerofrom-derive"
+version = "0.1.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d71e5d6e06ab090c67b5e44993ec16b72dcbaabc526db883a360057678b48502"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+ "synstructure",
+]
+
+[[package]]
+name = "zerotrie"
+version = "0.2.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2a59c17a5562d507e4b54960e8569ebee33bee890c70aa3fe7b97e85a9fd7851"
+dependencies = [
+ "databake",
+ "displaydoc",
+ "litemap",
+ "serde_core",
+ "yoke",
+ "zerofrom",
+ "zerovec",
+]
+
+[[package]]
+name = "zerovec"
+version = "0.11.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6c28719294829477f525be0186d13efa9a3c602f7ec202ca9e353d310fb9a002"
+dependencies = [
+ "databake",
+ "serde",
+ "yoke",
+ "zerofrom",
+ "zerovec-derive",
+]
+
+[[package]]
+name = "zerovec-derive"
+version = "0.11.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "eadce39539ca5cb3985590102671f2567e659fca9666581ad3411d59207951f3"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]

vendor/icu_normalizer/Cargo.toml

@@ -0,0 +1,210 @@
+# THIS FILE IS AUTOMATICALLY GENERATED BY CARGO
+#
+# When uploading crates to the registry Cargo will automatically
+# "normalize" Cargo.toml files for maximal compatibility
+# with all versions of Cargo and also rewrite `path` dependencies
+# to registry (e.g., crates.io) dependencies.
+#
+# If you are reading this file be aware that the original Cargo.toml
+# will likely look very different (and much more reasonable).
+# See Cargo.toml.orig for the original contents.
+
+[package]
+edition = "2021"
+rust-version = "1.83"
+name = "icu_normalizer"
+version = "2.1.1"
+authors = ["The ICU4X Project Developers"]
+build = false
+include = [
+    "data/**/*",
+    "src/**/*",
+    "examples/**/*",
+    "benches/**/*",
+    "tests/**/*",
+    "Cargo.toml",
+    "LICENSE",
+    "README.md",
+    "build.rs",
+]
+autolib = false
+autobins = false
+autoexamples = false
+autotests = false
+autobenches = false
+description = "API for normalizing text into Unicode Normalization Forms"
+homepage = "https://icu4x.unicode.org"
+readme = "README.md"
+categories = ["internationalization"]
+license = "Unicode-3.0"
+repository = "https://github.com/unicode-org/icu4x"
+
+[package.metadata.docs.rs]
+all-features = true
+
+[features]
+compiled_data = [
+    "dep:icu_normalizer_data",
+    "icu_properties?/compiled_data",
+    "icu_provider/baked",
+]
+datagen = [
+    "serde",
+    "dep:databake",
+    "icu_properties",
+    "icu_collections/databake",
+    "zerovec/databake",
+    "icu_properties?/datagen",
+    "icu_provider/export",
+]
+default = [
+    "compiled_data",
+    "utf8_iter",
+    "utf16_iter",
+]
+experimental = []
+icu_properties = ["dep:icu_properties"]
+serde = [
+    "dep:serde",
+    "icu_collections/serde",
+    "zerovec/serde",
+    "icu_properties?/serde",
+    "icu_provider/serde",
+]
+utf16_iter = [
+    "dep:utf16_iter",
+    "dep:write16",
+]
+utf8_iter = ["dep:utf8_iter"]
+write16 = []
+
+[lib]
+name = "icu_normalizer"
+path = "src/lib.rs"
+
+[[test]]
+name = "tests"
+path = "tests/tests.rs"
+
+[[bench]]
+name = "bench"
+path = "benches/bench.rs"
+harness = false
+required-features = [
+    "utf16_iter",
+    "utf8_iter",
+]
+
+[[bench]]
+name = "canonical_composition"
+path = "benches/canonical_composition.rs"
+
+[[bench]]
+name = "canonical_decomposition"
+path = "benches/canonical_decomposition.rs"
+
+[[bench]]
+name = "composing_normalizer_nfc"
+path = "benches/composing_normalizer_nfc.rs"
+
+[[bench]]
+name = "composing_normalizer_nfkc"
+path = "benches/composing_normalizer_nfkc.rs"
+
+[[bench]]
+name = "decomposing_normalizer_nfd"
+path = "benches/decomposing_normalizer_nfd.rs"
+
+[[bench]]
+name = "decomposing_normalizer_nfkd"
+path = "benches/decomposing_normalizer_nfkd.rs"
+
+[[bench]]
+name = "utf16_throughput"
+path = "benches/utf16_throughput.rs"
+
+[dependencies.databake]
+version = "0.2.0"
+features = ["derive"]
+optional = true
+default-features = false
+
+[dependencies.icu_collections]
+version = "~2.1.1"
+default-features = false
+
+[dependencies.icu_normalizer_data]
+version = "~2.1.1"
+optional = true
+default-features = false
+
+[dependencies.icu_properties]
+version = "~2.1.1"
+optional = true
+default-features = false
+
+[dependencies.icu_provider]
+version = "2.1.1"
+default-features = false
+
+[dependencies.serde]
+version = "1.0.220"
+features = [
+    "derive",
+    "alloc",
+]
+optional = true
+default-features = false
+
+[dependencies.smallvec]
+version = "1.10.0"
+default-features = false
+
+[dependencies.utf16_iter]
+version = "1.0.2"
+optional = true
+default-features = false
+
+[dependencies.utf8_iter]
+version = "1.0.2"
+optional = true
+default-features = false
+
+[dependencies.write16]
+version = "1.0.0"
+features = ["alloc"]
+optional = true
+default-features = false
+
+[dependencies.zerovec]
+version = "0.11.3"
+default-features = false
+
+[dev-dependencies.arraystring]
+version = "0.3.0"
+
+[dev-dependencies.arrayvec]
+version = "0.7.2"
+default-features = false
+
+[dev-dependencies.atoi]
+version = "2.0.0"
+
+[dev-dependencies.detone]
+version = "1.0.0"
+
+[dev-dependencies.write16]
+version = "1.0.0"
+features = [
+    "arrayvec",
+    "smallvec",
+]
+default-features = false
+
+[target.'cfg(not(target_arch = "wasm32"))'.dev-dependencies.criterion]
+version = "0.5.0"
+
+[lints.rust.unexpected_cfgs]
+level = "warn"
+priority = 0
+check-cfg = ["cfg(icu4x_unstable_fast_trie_only)"]

vendor/icu_normalizer/LICENSE

@@ -0,0 +1,46 @@
+UNICODE LICENSE V3
+
+COPYRIGHT AND PERMISSION NOTICE
+
+Copyright © 2020-2024 Unicode, Inc.
+
+NOTICE TO USER: Carefully read the following legal agreement. BY
+DOWNLOADING, INSTALLING, COPYING OR OTHERWISE USING DATA FILES, AND/OR
+SOFTWARE, YOU UNEQUIVOCALLY ACCEPT, AND AGREE TO BE BOUND BY, ALL OF THE
+TERMS AND CONDITIONS OF THIS AGREEMENT. IF YOU DO NOT AGREE, DO NOT
+DOWNLOAD, INSTALL, COPY, DISTRIBUTE OR USE THE DATA FILES OR SOFTWARE.
+
+Permission is hereby granted, free of charge, to any person obtaining a
+copy of data files and any associated documentation (the "Data Files") or
+software and any associated documentation (the "Software") to deal in the
+Data Files or Software without restriction, including without limitation
+the rights to use, copy, modify, merge, publish, distribute, and/or sell
+copies of the Data Files or Software, and to permit persons to whom the
+Data Files or Software are furnished to do so, provided that either (a)
+this copyright and permission notice appear with all copies of the Data
+Files or Software, or (b) this copyright and permission notice appear in
+associated Documentation.
+
+THE DATA FILES AND SOFTWARE ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY
+KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT OF
+THIRD PARTY RIGHTS.
+
+IN NO EVENT SHALL THE COPYRIGHT HOLDER OR HOLDERS INCLUDED IN THIS NOTICE
+BE LIABLE FOR ANY CLAIM, OR ANY SPECIAL INDIRECT OR CONSEQUENTIAL DAMAGES,
+OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS,
+WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION,
+ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THE DATA
+FILES OR SOFTWARE.
+
+Except as contained in this notice, the name of a copyright holder shall
+not be used in advertising or otherwise to promote the sale, use or other
+dealings in these Data Files or Software without prior written
+authorization of the copyright holder.
+
+SPDX-License-Identifier: Unicode-3.0
+
+—
+
+Portions of ICU4X may have been adapted from ICU4C and/or ICU4J.
+ICU 1.8.1 to ICU 57.1 © 1995-2016 International Business Machines Corporation and others.

vendor/icu_normalizer/README.md

@@ -0,0 +1,48 @@
+# icu_normalizer [![crates.io](https://img.shields.io/crates/v/icu_normalizer)](https://crates.io/crates/icu_normalizer)
+
+<!-- cargo-rdme start -->
+
+Normalizing text into Unicode Normalization Forms.
+
+This module is published as its own crate ([`icu_normalizer`](https://docs.rs/icu_normalizer/latest/icu_normalizer/))
+and as part of the [`icu`](https://docs.rs/icu/latest/icu/) crate. See the latter for more details on the ICU4X project.
+
+## Functionality
+
+The top level of the crate provides normalization of input into the four normalization forms defined in [UAX #15: Unicode
+Normalization Forms](https://www.unicode.org/reports/tr15/): NFC, NFD, NFKC, and NFKD.
+
+Three kinds of contiguous inputs are supported: known-well-formed UTF-8 (`&str`), potentially-not-well-formed UTF-8,
+and potentially-not-well-formed UTF-16. Additionally, an iterator over `char` can be wrapped in a normalizing iterator.
+
+The `uts46` module provides the combination of mapping and normalization operations for [UTS #46: Unicode IDNA
+Compatibility Processing](https://www.unicode.org/reports/tr46/). This functionality is not meant to be used by
+applications directly. Instead, it is meant as a building block for a full implementation of UTS #46, such as the
+[`idna`](https://docs.rs/idna/latest/idna/) crate.
+
+The `properties` module provides the non-recursive canonical decomposition operation on a per `char` basis and
+the canonical compositon operation given two `char`s. It also provides access to the Canonical Combining Class
+property. These operations are primarily meant for [HarfBuzz](https://harfbuzz.github.io/) via the
+[`icu_harfbuzz`](https://docs.rs/icu_harfbuzz/latest/icu_harfbuzz/) crate.
+
+Notably, this normalizer does _not_ provide the normalization “quick check” that can result in “maybe” in
+addition to “yes” and “no”. The normalization checks provided by this crate always give a definitive
+non-“maybe” answer.
+
+## Examples
+
+```rust
+let nfc = icu_normalizer::ComposingNormalizerBorrowed::new_nfc();
+assert_eq!(nfc.normalize("a\u{0308}"), "ä");
+assert!(nfc.is_normalized("ä"));
+
+let nfd = icu_normalizer::DecomposingNormalizerBorrowed::new_nfd();
+assert_eq!(nfd.normalize("ä"), "a\u{0308}");
+assert!(!nfd.is_normalized("ä"));
+```
+
+<!-- cargo-rdme end -->
+
+## More Information
+
+For more information on development, authorship, contributing etc. please visit [`ICU4X home page`](https://github.com/unicode-org/icu4x).

vendor/icu_normalizer/benches/bench.rs

@@ -0,0 +1,26 @@
+// This file is part of ICU4X. For terms of use, please see the file
+// called LICENSE at the top level of the ICU4X source tree
+// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+use criterion::{criterion_group, criterion_main};
+
+mod canonical_composition;
+mod canonical_decomposition;
+mod composing_normalizer_nfc;
+mod composing_normalizer_nfkc;
+mod decomposing_normalizer_nfd;
+mod decomposing_normalizer_nfkd;
+mod utf16_throughput;
+
+criterion_group!(
+    benches,
+    canonical_composition::criterion_benchmark,
+    canonical_decomposition::criterion_benchmark,
+    composing_normalizer_nfc::criterion_benchmark,
+    composing_normalizer_nfkc::criterion_benchmark,
+    decomposing_normalizer_nfd::criterion_benchmark,
+    decomposing_normalizer_nfkd::criterion_benchmark,
+    utf16_throughput::criterion_benchmark,
+);
+
+criterion_main!(benches);

vendor/icu_normalizer/benches/canonical_composition.rs

@@ -0,0 +1,188 @@
+// This file is part of ICU4X. For terms of use, please see the file
+// called LICENSE at the top level of the ICU4X source tree
+// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+use criterion::{black_box, BenchmarkId, Criterion};
+use detone::IterDecomposeVietnamese;
+
+use icu_normalizer::properties::{
+    CanonicalCompositionBorrowed, CanonicalDecompositionBorrowed, Decomposed,
+};
+use icu_normalizer::ComposingNormalizerBorrowed;
+
+struct BenchDataContent {
+    pub file_name: String,
+    pub pairs: Vec<(char, char)>,
+}
+
+fn strip_headers(content: &str) -> String {
+    content
+        .lines()
+        .filter(|&s| !s.starts_with('#'))
+        .map(|s| s.to_owned())
+        .collect::<Vec<String>>()
+        .join("\n")
+}
+
+fn normalizer_bench_data() -> [BenchDataContent; 16] {
+    let nfc_normalizer = ComposingNormalizerBorrowed::new_nfc();
+
+    [
+        BenchDataContent {
+            file_name: "TestNames_Latin".to_owned(),
+            pairs: decompose_data(
+                &nfc_normalizer
+                    .normalize(&strip_headers(include_str!("./data/TestNames_Latin.txt"))),
+            ),
+        },
+        BenchDataContent {
+            file_name: "TestNames_Japanese_h".to_owned(),
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestNames_Japanese_h.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "TestNames_Japanese_k".to_owned(),
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestNames_Japanese_k.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "TestNames_Korean".to_owned(),
+            pairs: decompose_data(
+                &nfc_normalizer
+                    .normalize(&strip_headers(include_str!("./data/TestNames_Korean.txt"))),
+            ),
+        },
+        BenchDataContent {
+            file_name: "TestRandomWordsUDHR_ar".to_owned(),
+            #[cfg(debug_assertions)]
+            pairs: Vec::new(),
+            #[cfg(not(debug_assertions))]
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestRandomWordsUDHR_ar.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "TestRandomWordsUDHR_de".to_owned(),
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestRandomWordsUDHR_de.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "TestRandomWordsUDHR_el".to_owned(),
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestRandomWordsUDHR_el.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "TestRandomWordsUDHR_es".to_owned(),
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestRandomWordsUDHR_es.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "TestRandomWordsUDHR_fr".to_owned(),
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestRandomWordsUDHR_fr.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "TestRandomWordsUDHR_he".to_owned(),
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestRandomWordsUDHR_he.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "TestRandomWordsUDHR_pl".to_owned(),
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestRandomWordsUDHR_pl.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "TestRandomWordsUDHR_ru".to_owned(),
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestRandomWordsUDHR_ru.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "TestRandomWordsUDHR_th".to_owned(),
+            #[cfg(debug_assertions)]
+            pairs: Vec::new(),
+            #[cfg(not(debug_assertions))]
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestRandomWordsUDHR_th.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "TestRandomWordsUDHR_tr".to_owned(),
+            pairs: decompose_data(&nfc_normalizer.normalize(&strip_headers(include_str!(
+                "./data/TestRandomWordsUDHR_tr.txt"
+            )))),
+        },
+        BenchDataContent {
+            file_name: "udhr_vie".to_owned(),
+            pairs: decompose_data(
+                &nfc_normalizer.normalize(&strip_headers(include_str!("data/wotw.txt"))),
+            ),
+        },
+        BenchDataContent {
+            file_name: "udhr_vie_detone".to_owned(),
+            pairs: {
+                let result: Vec<(char, char)> = nfc_normalizer
+                    .normalize(&strip_headers(include_str!("data/wotw.txt")))
+                    .chars()
+                    .filter_map(|c| {
+                        let mut iter = std::iter::once(c).decompose_vietnamese_tones(true);
+                        if let Some(base) = iter.next() {
+                            iter.next().map(|tone| (base, tone))
+                        } else {
+                            None
+                        }
+                    })
+                    .collect();
+                assert!(!result.is_empty());
+                result
+            },
+        },
+    ]
+}
+
+fn function_under_bench(
+    canonical_composer: &CanonicalCompositionBorrowed,
+    composable_points: &[(char, char)],
+) {
+    for pair in composable_points.iter() {
+        canonical_composer.compose(pair.0, pair.1);
+    }
+}
+
+pub fn criterion_benchmark(criterion: &mut Criterion) {
+    let group_name = "canonical_composition";
+    let mut group = criterion.benchmark_group(group_name);
+
+    let composer = CanonicalCompositionBorrowed::new();
+
+    for bench_data_content in black_box(normalizer_bench_data()) {
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfc_{}", bench_data_content.file_name)),
+            |bencher| bencher.iter(|| function_under_bench(&composer, &bench_data_content.pairs)),
+        );
+    }
+
+    group.finish();
+}
+
+fn decompose_data(nfc: &str) -> Vec<(char, char)> {
+    let decomposer = CanonicalDecompositionBorrowed::new();
+    nfc.chars()
+        .map(|c| decomposer.decompose(c))
+        .filter_map(|decomposed| {
+            if let Decomposed::Expansion(a, b) = decomposed {
+                Some((a, b))
+            } else {
+                None
+            }
+        })
+        .collect()
+}

vendor/icu_normalizer/benches/canonical_decomposition.rs

@@ -0,0 +1,162 @@
+// This file is part of ICU4X. For terms of use, please see the file
+// called LICENSE at the top level of the ICU4X source tree
+// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+use criterion::{black_box, BenchmarkId, Criterion};
+
+use icu_normalizer::properties::CanonicalDecompositionBorrowed;
+use icu_normalizer::{ComposingNormalizerBorrowed, DecomposingNormalizerBorrowed};
+
+struct BenchDataContent {
+    pub file_name: String,
+    pub nfc: String,
+    pub nfd: String,
+    pub nfkc: String,
+    pub nfkd: String,
+}
+
+fn strip_headers(content: &str) -> String {
+    content
+        .lines()
+        .filter(|&s| !s.starts_with('#'))
+        .map(|s| s.to_owned())
+        .collect::<Vec<String>>()
+        .join("\n")
+}
+
+fn normalizer_bench_data() -> [BenchDataContent; 15] {
+    let nfc_normalizer = ComposingNormalizerBorrowed::new_nfc();
+    let nfd_normalizer = DecomposingNormalizerBorrowed::new_nfd();
+    let nfkc_normalizer = ComposingNormalizerBorrowed::new_nfkc();
+    let nfkd_normalizer = DecomposingNormalizerBorrowed::new_nfkd();
+
+    let content_latin: (&str, &str) = (
+        "TestNames_Latin",
+        &strip_headers(include_str!("./data/TestNames_Latin.txt")),
+    );
+    let content_jp_h: (&str, &str) = (
+        "TestNames_Japanese_h",
+        &strip_headers(include_str!("./data/TestNames_Japanese_h.txt")),
+    );
+    let content_jp_k: (&str, &str) = (
+        "TestNames_Japanese_k",
+        &strip_headers(include_str!("./data/TestNames_Japanese_k.txt")),
+    );
+    let content_korean: (&str, &str) = (
+        "TestNames_Korean",
+        &strip_headers(include_str!("./data/TestNames_Korean.txt")),
+    );
+    let content_random_words_ar: (&str, &str) = (
+        "TestRandomWordsUDHR_ar",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_ar.txt")),
+    );
+    let content_random_words_de: (&str, &str) = (
+        "TestRandomWordsUDHR_de",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_de.txt")),
+    );
+    let content_random_words_el: (&str, &str) = (
+        "TestRandomWordsUDHR_el",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_el.txt")),
+    );
+    let content_random_words_es: (&str, &str) = (
+        "TestRandomWordsUDHR_es",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_es.txt")),
+    );
+    let content_random_words_fr: (&str, &str) = (
+        "TestRandomWordsUDHR_fr",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_fr.txt")),
+    );
+    let content_random_words_he: (&str, &str) = (
+        "TestRandomWordsUDHR_he",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_he.txt")),
+    );
+    let content_random_words_pl: (&str, &str) = (
+        "TestRandomWordsUDHR_pl",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_pl.txt")),
+    );
+    let content_random_words_ru: (&str, &str) = (
+        "TestRandomWordsUDHR_ru",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_ru.txt")),
+    );
+    let content_random_words_th: (&str, &str) = (
+        "TestRandomWordsUDHR_th",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_th.txt")),
+    );
+    let content_random_words_tr: (&str, &str) = (
+        "TestRandomWordsUDHR_tr",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_tr.txt")),
+    );
+    let content_viet: (&str, &str) = ("udhr_vie", &strip_headers(include_str!("data/wotw.txt")));
+
+    [
+        content_latin,
+        content_viet,
+        content_jp_k,
+        content_jp_h,
+        content_korean,
+        content_random_words_ru,
+        content_random_words_ar,
+        content_random_words_el,
+        content_random_words_es,
+        content_random_words_fr,
+        content_random_words_tr,
+        content_random_words_th,
+        content_random_words_pl,
+        content_random_words_he,
+        content_random_words_de,
+    ]
+    .map(|(file_name, raw_content)| BenchDataContent {
+        file_name: file_name.to_owned(),
+        nfc: nfc_normalizer.normalize(raw_content).to_string(),
+        nfd: nfd_normalizer.normalize(raw_content).to_string(),
+        nfkc: nfkc_normalizer.normalize(raw_content).to_string(),
+        nfkd: nfkd_normalizer.normalize(raw_content).to_string(),
+    })
+}
+
+#[cfg(debug_assertions)]
+fn function_under_bench(
+    _canonical_decomposer: &CanonicalDecompositionBorrowed,
+    _decomposable_points: &str,
+) {
+    // using debug assertion fails some test.
+    // "cargo test --bench bench" will pass
+    // "cargo bench" will work as expected, because the profile doesn't include debug assertions.
+}
+
+#[cfg(not(debug_assertions))]
+fn function_under_bench(
+    canonical_decomposer: &CanonicalDecompositionBorrowed,
+    decomposable_points: &str,
+) {
+    decomposable_points.chars().for_each(|point| {
+        canonical_decomposer.decompose(point);
+    });
+}
+
+pub fn criterion_benchmark(criterion: &mut Criterion) {
+    let group_name = "canonical_decomposition";
+    let mut group = criterion.benchmark_group(group_name);
+
+    let decomposer = CanonicalDecompositionBorrowed::new();
+
+    for bench_data_content in black_box(normalizer_bench_data()) {
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfc_{}", bench_data_content.file_name)),
+            |bencher| bencher.iter(|| function_under_bench(&decomposer, &bench_data_content.nfc)),
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfd_{}", bench_data_content.file_name)),
+            |bencher| bencher.iter(|| function_under_bench(&decomposer, &bench_data_content.nfd)),
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkc_{}", bench_data_content.file_name)),
+            |bencher| bencher.iter(|| function_under_bench(&decomposer, &bench_data_content.nfkc)),
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkd_{}", bench_data_content.file_name)),
+            |bencher| bencher.iter(|| function_under_bench(&decomposer, &bench_data_content.nfkd)),
+        );
+    }
+    group.finish();
+}

vendor/icu_normalizer/benches/composing_normalizer_nfc.rs

@@ -0,0 +1,230 @@
+// This file is part of ICU4X. For terms of use, please see the file
+// called LICENSE at the top level of the ICU4X source tree
+// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+use criterion::{black_box, BenchmarkId, Criterion};
+
+use icu_normalizer::{ComposingNormalizerBorrowed, DecomposingNormalizerBorrowed};
+
+struct BenchDataContent {
+    pub file_name: String,
+    pub nfc: String,
+    pub nfd: String,
+    pub nfkc: String,
+    pub nfkd: String,
+    pub nfc_u16: Vec<u16>,
+    pub nfd_u16: Vec<u16>,
+    pub nfkc_u16: Vec<u16>,
+    pub nfkd_u16: Vec<u16>,
+}
+
+fn strip_headers(content: &str) -> String {
+    content
+        .lines()
+        .filter(|&s| !s.starts_with('#'))
+        .map(|s| s.to_owned())
+        .collect::<Vec<String>>()
+        .join("\n")
+}
+
+fn normalizer_bench_data() -> [BenchDataContent; 15] {
+    let nfc_normalizer = ComposingNormalizerBorrowed::new_nfc();
+    let nfd_normalizer = DecomposingNormalizerBorrowed::new_nfd();
+    let nfkc_normalizer = ComposingNormalizerBorrowed::new_nfkc();
+    let nfkd_normalizer = DecomposingNormalizerBorrowed::new_nfkd();
+
+    let content_latin: (&str, &str) = (
+        "TestNames_Latin",
+        &strip_headers(include_str!("./data/TestNames_Latin.txt")),
+    );
+    let content_jp_h: (&str, &str) = (
+        "TestNames_Japanese_h",
+        &strip_headers(include_str!("./data/TestNames_Japanese_h.txt")),
+    );
+    let content_jp_k: (&str, &str) = (
+        "TestNames_Japanese_k",
+        &strip_headers(include_str!("./data/TestNames_Japanese_k.txt")),
+    );
+    let content_korean: (&str, &str) = (
+        "TestNames_Korean",
+        &strip_headers(include_str!("./data/TestNames_Korean.txt")),
+    );
+    let content_random_words_ar: (&str, &str) = (
+        "TestRandomWordsUDHR_ar",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_ar.txt")),
+    );
+    let content_random_words_de: (&str, &str) = (
+        "TestRandomWordsUDHR_de",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_de.txt")),
+    );
+    let content_random_words_el: (&str, &str) = (
+        "TestRandomWordsUDHR_el",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_el.txt")),
+    );
+    let content_random_words_es: (&str, &str) = (
+        "TestRandomWordsUDHR_es",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_es.txt")),
+    );
+    let content_random_words_fr: (&str, &str) = (
+        "TestRandomWordsUDHR_fr",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_fr.txt")),
+    );
+    let content_random_words_he: (&str, &str) = (
+        "TestRandomWordsUDHR_he",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_he.txt")),
+    );
+    let content_random_words_pl: (&str, &str) = (
+        "TestRandomWordsUDHR_pl",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_pl.txt")),
+    );
+    let content_random_words_ru: (&str, &str) = (
+        "TestRandomWordsUDHR_ru",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_ru.txt")),
+    );
+    let content_random_words_th: (&str, &str) = (
+        "TestRandomWordsUDHR_th",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_th.txt")),
+    );
+    let content_random_words_tr: (&str, &str) = (
+        "TestRandomWordsUDHR_tr",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_tr.txt")),
+    );
+    let content_viet: (&str, &str) = ("wotw", &strip_headers(include_str!("./data/wotw.txt")));
+
+    [
+        content_latin,
+        content_viet,
+        content_jp_k,
+        content_jp_h,
+        content_korean,
+        content_random_words_ru,
+        content_random_words_ar,
+        content_random_words_el,
+        content_random_words_es,
+        content_random_words_fr,
+        content_random_words_tr,
+        content_random_words_th,
+        content_random_words_pl,
+        content_random_words_he,
+        content_random_words_de,
+    ]
+    .map(|(file_name, raw_content)| {
+        let nfc = &nfc_normalizer.normalize(raw_content);
+        let nfd = &nfd_normalizer.normalize(raw_content);
+        let nfkc = &nfkc_normalizer.normalize(raw_content);
+        let nfkd = &nfkd_normalizer.normalize(raw_content);
+        BenchDataContent {
+            file_name: file_name.to_owned(),
+            nfc: nfc.to_string(),
+            nfd: nfd.to_string(),
+            nfkc: nfkc.to_string(),
+            nfkd: nfkd.to_string(),
+            nfc_u16: nfc.encode_utf16().collect(),
+            nfd_u16: nfd.encode_utf16().collect(),
+            nfkc_u16: nfkc.encode_utf16().collect(),
+            nfkd_u16: nfkd.encode_utf16().collect(),
+        }
+    })
+}
+
+fn function_under_bench(normalizer: &ComposingNormalizerBorrowed, text: &str) {
+    normalizer.normalize(text);
+}
+
+fn function_under_bench_utf16(normalizer: &ComposingNormalizerBorrowed, text: &[u16]) {
+    normalizer.normalize_utf16(text);
+}
+
+pub fn criterion_benchmark(criterion: &mut Criterion) {
+    let group_name = "composing_normalizer_nfc";
+
+    let normalizer_under_bench = ComposingNormalizerBorrowed::new_nfc();
+
+    let mut group = criterion.benchmark_group(group_name);
+
+    for bench_data_content in black_box(normalizer_bench_data()) {
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfc_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher
+                    .iter(|| function_under_bench(&normalizer_under_bench, &bench_data_content.nfc))
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfd_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher
+                    .iter(|| function_under_bench(&normalizer_under_bench, &bench_data_content.nfd))
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkc_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench(&normalizer_under_bench, &bench_data_content.nfkc)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkd_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench(&normalizer_under_bench, &bench_data_content.nfkd)
+                })
+            },
+        );
+
+        // UTF_16
+        group.bench_function(
+            BenchmarkId::from_parameter(format!(
+                "from_nfc_{}_utf_16",
+                bench_data_content.file_name
+            )),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_utf16(&normalizer_under_bench, &bench_data_content.nfc_u16)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!(
+                "from_nfd_{}_utf_16",
+                bench_data_content.file_name
+            )),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_utf16(&normalizer_under_bench, &bench_data_content.nfd_u16)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!(
+                "from_nfkc_{}_utf_16",
+                bench_data_content.file_name
+            )),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_utf16(
+                        &normalizer_under_bench,
+                        &bench_data_content.nfkc_u16,
+                    )
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!(
+                "from_nfkd_{}_utf_16",
+                bench_data_content.file_name
+            )),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_utf16(
+                        &normalizer_under_bench,
+                        &bench_data_content.nfkd_u16,
+                    )
+                })
+            },
+        );
+    }
+    group.finish();
+}

vendor/icu_normalizer/benches/composing_normalizer_nfkc.rs

@@ -0,0 +1,211 @@
+// This file is part of ICU4X. For terms of use, please see the file
+// called LICENSE at the top level of the ICU4X source tree
+// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+use criterion::{black_box, BenchmarkId, Criterion};
+
+use icu_normalizer::{ComposingNormalizerBorrowed, DecomposingNormalizerBorrowed};
+
+struct BenchDataContent {
+    pub file_name: String,
+    pub nfc: String,
+    pub nfd: String,
+    pub nfkc: String,
+    pub nfkd: String,
+    pub nfc_u16: Vec<u16>,
+    pub nfd_u16: Vec<u16>,
+    pub nfkc_u16: Vec<u16>,
+    pub nfkd_u16: Vec<u16>,
+}
+
+fn strip_headers(content: &str) -> String {
+    content
+        .lines()
+        .filter(|&s| !s.starts_with('#'))
+        .map(|s| s.to_owned())
+        .collect::<Vec<String>>()
+        .join("\n")
+}
+
+fn normalizer_bench_data() -> [BenchDataContent; 15] {
+    let nfc_normalizer = ComposingNormalizerBorrowed::new_nfc();
+    let nfd_normalizer = DecomposingNormalizerBorrowed::new_nfd();
+    let nfkc_normalizer = ComposingNormalizerBorrowed::new_nfkc();
+    let nfkd_normalizer = DecomposingNormalizerBorrowed::new_nfkd();
+
+    let content_latin: (&str, &str) = (
+        "TestNames_Latin",
+        &strip_headers(include_str!("./data/TestNames_Latin.txt")),
+    );
+    let content_jp_h: (&str, &str) = (
+        "TestNames_Japanese_h",
+        &strip_headers(include_str!("./data/TestNames_Japanese_h.txt")),
+    );
+    let content_jp_k: (&str, &str) = (
+        "TestNames_Japanese_k",
+        &strip_headers(include_str!("./data/TestNames_Japanese_k.txt")),
+    );
+    let content_korean: (&str, &str) = (
+        "TestNames_Korean",
+        &strip_headers(include_str!("./data/TestNames_Korean.txt")),
+    );
+    let content_random_words_ar: (&str, &str) = (
+        "TestRandomWordsUDHR_ar",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_ar.txt")),
+    );
+    let content_random_words_de: (&str, &str) = (
+        "TestRandomWordsUDHR_de",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_de.txt")),
+    );
+    let content_random_words_el: (&str, &str) = (
+        "TestRandomWordsUDHR_el",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_el.txt")),
+    );
+    let content_random_words_es: (&str, &str) = (
+        "TestRandomWordsUDHR_es",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_es.txt")),
+    );
+    let content_random_words_fr: (&str, &str) = (
+        "TestRandomWordsUDHR_fr",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_fr.txt")),
+    );
+    let content_random_words_he: (&str, &str) = (
+        "TestRandomWordsUDHR_he",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_he.txt")),
+    );
+    let content_random_words_pl: (&str, &str) = (
+        "TestRandomWordsUDHR_pl",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_pl.txt")),
+    );
+    let content_random_words_ru: (&str, &str) = (
+        "TestRandomWordsUDHR_ru",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_ru.txt")),
+    );
+    let content_random_words_th: (&str, &str) = (
+        "TestRandomWordsUDHR_th",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_th.txt")),
+    );
+    let content_random_words_tr: (&str, &str) = (
+        "TestRandomWordsUDHR_tr",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_tr.txt")),
+    );
+    let content_viet: (&str, &str) = ("wotw", &strip_headers(include_str!("./data/wotw.txt")));
+
+    [
+        content_latin,
+        content_viet,
+        content_jp_k,
+        content_jp_h,
+        content_korean,
+        content_random_words_ru,
+        content_random_words_ar,
+        content_random_words_el,
+        content_random_words_es,
+        content_random_words_fr,
+        content_random_words_tr,
+        content_random_words_th,
+        content_random_words_pl,
+        content_random_words_he,
+        content_random_words_de,
+    ]
+    .map(|(file_name, raw_content)| {
+        let nfc = &nfc_normalizer.normalize(raw_content);
+        let nfd = &nfd_normalizer.normalize(raw_content);
+        let nfkc = &nfkc_normalizer.normalize(raw_content);
+        let nfkd = &nfkd_normalizer.normalize(raw_content);
+        BenchDataContent {
+            file_name: file_name.to_owned(),
+            nfc: nfc.to_string(),
+            nfd: nfd.to_string(),
+            nfkc: nfkc.to_string(),
+            nfkd: nfkd.to_string(),
+            nfc_u16: nfc.encode_utf16().collect(),
+            nfd_u16: nfd.encode_utf16().collect(),
+            nfkc_u16: nfkc.encode_utf16().collect(),
+            nfkd_u16: nfkd.encode_utf16().collect(),
+        }
+    })
+}
+
+fn function_under_bench(normalizer: &ComposingNormalizerBorrowed, text: &str) {
+    normalizer.normalize(text);
+}
+
+fn function_under_bench_u16(normalizer: &ComposingNormalizerBorrowed, text: &[u16]) {
+    normalizer.normalize_utf16(text);
+}
+
+pub fn criterion_benchmark(criterion: &mut Criterion) {
+    let group_name = "composing_normalizer_nfkc";
+
+    let normalizer_under_bench = ComposingNormalizerBorrowed::new_nfkc();
+
+    let mut group = criterion.benchmark_group(group_name);
+
+    for bench_data_content in black_box(normalizer_bench_data()) {
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfc_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher
+                    .iter(|| function_under_bench(&normalizer_under_bench, &bench_data_content.nfc))
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfd_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher
+                    .iter(|| function_under_bench(&normalizer_under_bench, &bench_data_content.nfd))
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkc_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench(&normalizer_under_bench, &bench_data_content.nfkc)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkd_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench(&normalizer_under_bench, &bench_data_content.nfkd)
+                })
+            },
+        );
+        // UTF 16
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfc_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfc_u16)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfd_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfd_u16)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkc_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfkc_u16)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkd_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfkd_u16)
+                })
+            },
+        );
+    }
+    group.finish();
+}

vendor/icu_normalizer/benches/data/README.md

@@ -0,0 +1,25 @@
+# Generating microbench data
+
+The full versions of these files are located 
+[in another part of the repository](https://github.com/unicode-org/icu/tree/main/icu4j/perf-tests/data).
+
+## Sanitizing the file
+
+```shell
+sed -i '/^#/d' ${filename}
+sed -i '/^$/d' ${filename}
+```
+
+## Shuffling the file
+
+```shell
+shuf -n 20 ${filename} -o ${filename}
+```
+
+## Add back the header (if you plan on submitting the files)
+
+```
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+```

vendor/icu_normalizer/benches/data/TestNames_Japanese_h.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+かげやま,みのる
+むらかみ,とおる
+つじさわ,けい
+やすい,たかゆき
+むらさき,としお
+はせがわ,ひであき
+うるしばら,よしひこ
+ままだ,ひろし
+おおぼら,えいじろう
+おおば,まさひで
+きたばたけ,たかひこ
+はまさき,あつし
+ほりい,つねお
+もり,だいいち
+いとう,しんいち
+くにもと,じゅんじ
+おか,のりひと
+たに,よしあき
+しらがき,ひろあき
+しらはま,たけひろ
+むらかみ,やすひろ
+うめはら,たかし
+いわた,ひろし
+すぎえ,かつとし
+てらにし,ひろみつ
+まつおか,だいすけ
+もろほし,すすむ
+いしはら,たかし
+おしま,ひろお
+なかお,ゆうじ
+いかり,はるお
+きまち,まさき
+ふるかわ,みちお
+かねこ,しゅうへい
+なかがわ,ともみ
+ささき,しんご
+うちだ,たくじ
+うめだ,さかえ
+しばた,いくこ
+まきした,けいこ
+まつもと,しんいちろう
+たかの,かずよし
+いしわた,なおひさ
+いうち,まこと
+いまい,りほ
+みずた,のりあき
+かくたに,まなぶ
+わだ,ほまれ
+わかまつ,かずき
+かわぐち,ひろき

vendor/icu_normalizer/benches/data/TestNames_Japanese_k.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+ホリモト,ユウジ
+ハナミ,ヤスヒデ
+イシザカ,タカユキ
+ゼンケ,トシオ
+ハトリ,ユウコ
+ナガオカ,トモユキ
+コウダ,ケンイチ
+イシダ,ヒロシ
+ミワ,シゲユキ
+イシカワ,ヒロシ
+スズキ,ユウスケ
+オクダ,ヨシノリ
+シムラ,サカエ
+エビシマ,ヤスユキ
+イブカ,ヨシテル
+タノ,マコト
+ドウゾノ,セイヤ
+ヤマナカ,サツミ
+トミイエ,ハヤト
+アザミ,ツトム
+タナカ,キョウコ
+コジマ,アツシ
+フミハラ,カオリ
+スズキ,マサユキ
+ナトリ,ケンヤ
+スズキ,ユウコ
+スズキ,ヒサエ
+ナカガワ,カツヨシ
+スズキ,マサフミ
+マツヤマ,トシオ
+ヨシナガ,チカエ
+キタムラ,リカコ
+アオキ,タクオ
+ヤマグチ,ヤスヒロ
+スギムラ,シゲオ
+ウエスギ,マサミ
+マツムラ,シンイチ
+クバ,タカシ
+スドウ,タカトシ
+フジモト,ヒロシ
+イトウ,シュウイチ
+コバヤシ,カズミ
+タナカ,ヒロカツ
+イシダ,ツカサ
+ヤマダ,マサコ
+カミヤ,トミエ
+タケモト,ユウジ
+スミノ,コウジ
+ヒロハタ,タクヤ
+ミヒラ,リョウヘイ

vendor/icu_normalizer/benches/data/TestNames_Korean.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+김명희
+홍차수
+허순재
+강영휘
+김운주
+이종환
+이은국
+강태호
+강일래
+김동현
+곽기자
+차재수
+표봉기
+문대원
+이형기
+최교표
+박식현
+홍종립
+서창수
+김쌍건
+서말도
+이병훈
+김희수
+박학태
+강태종
+조문란
+신범균
+백두진
+이철정
+김태중
+이성현
+김주조
+김강행
+이정길
+김완일
+권수자
+이춘철
+김판근
+김곡리
+이경형
+이운만
+손상철
+유기숙
+박정한
+조윤래
+유신호
+이두수
+김재률
+김성홍
+김혜경

vendor/icu_normalizer/benches/data/TestNames_Latin.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+González, Joan
+Reinders, Jim
+Applebroog, Ida
+Kidd, Joseph Bartholomew
+Gulácsy, Lajos
+Letendre, Rita
+Zuccaro, Federico
+Apt the Elder, Ulrich
+Drummond, Arthur
+Manley, Thomas
+Broc, Jean
+Ramunno, Tony
+Simone dei Crocifissi
+Lane, Theodore
+Symonds, William Robert
+Johnson, Frank Tenney
+Cox, Gardner
+Bunbury, Charles
+Pedro de la Cuadra
+Payne, William
+Lucas, John Seymour
+Holsman, Elizabeth T.
+de Vries, Auke
+Laszlo, Philip Alexius de
+Shigemasa
+Wolfe, Ruth Mitchell
+Buck, John
+Baselitz, Georg
+Hook, Walter
+Segall, Lasar
+Brush, George deForest
+Master of Jánosrét
+Sutherland, Elizabeth Leveson-Gower, Countess of
+Tuckerman, Jane
+Varley, F.H.
+Fosso, Samuel
+Gardner, Daniel
+Sadler, Walter Dendy
+Clausen, Franciska
+Coman, Charlotte Buell
+Wakelin, Roland
+Payne, Jon, CML
+Campagna, Girolamo
+Wiener, Phyllis
+Sallee, Charles
+Fitzgerald, John Anster
+Gribbroek, Robert
+Laporte, John
+Lévy-Dhurmer, Lucien
+Young, Stephen Scott

vendor/icu_normalizer/benches/data/TestNames_Thai.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+ณรงค์ โต๊ะเงิน
+กิตติ บุญวันต์
+สมหมาย ดาบทองดี
+ธวัชชัย อิสระนิมิตร
+วรรณา โสภณนรินทร์
+วินัย หมู่มิ่ง
+พัชรี ชูจิรวงศ์
+สมปอง จิวไพโรจน์กิจ
+บุญส่ง กวยรักษา
+นิพนธ์ นิ่มใหม่
+พัชรี สุวพรศิลป์
+เจริญ นววัฒนทรัพย์
+อรพินท์ แซ่เจี่ย
+ชัยพร สมใจนึก
+ประนอม โคศิลา
+ฉวีวรรณ ศรสังข์ทอง
+วัชรา เจริญรัตนพร
+สุภัท นกศิริ
+อู๋ มาลาเล็ก
+ประยูร ไชโย
+ละออ อยู่ยืนยง
+สมใจ วิวัฒน์วานิช
+จุมพล จันทรศรีเกษร
+พุฒ ดอกไม้จีน
+บุญชัย วรกิจพรสิน
+สมาน ธูปเทียน
+พงศ์ศักดิ์ แซ่แต้
+อำนาจ ไวจงเจริญ
+พรทิพย์ แซ่ลี้
+อุไรวรรณ สาครสินธุ์
+อำพล วีระตะนนท์
+สมจิตร ใจวังโลก
+สุเทพ ตันวินิจ
+สวาท ทรัพย์มาก
+สมศักดิ์ เจือจันทร์
+ดัสซันซิงห์ กุลาตี
+ธีร ศรแก้ว
+พรรณยุพา ฮ่อสกุล
+สำราญ จันทร์เอี่ยม
+พจน์ มั่นกันนาน
+สุธี บุณยเกียรติ
+บุญโชติ ทิพย์ประเสริฐสิน
+ประดิษฐ์ ทองพสิฐสมบัติ
+จำเนียร เพ็งเจริญ
+สมศักดิ์ อรุณรัตน์
+อนุชา จารุหิรัญสกุล
+พิกุล มโนภิญโญภิญญะ
+ผ่องศรี นกแก้ว
+อารี วิไลวรรณ
+ณรงค์วิทย์ วิทสัทธาวรกุล

vendor/icu_normalizer/benches/data/TestRandomWordsUDHR_ar.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+ممارسة مراعاة
+العنصرية
+حدود والشيخوخة
+بالحكم كهذا ينتفع
+البلاد
+تربية
+الغير التقدم والعدل
+نحو بالتعليم والحرية
+تأمين متساو
+للتعليم فيها
+آذت اعتداء للتعليم
+ليس المتأصلة
+والمساهمة الضروري تتناقض
+وتأسيس
+رضى
+شرعي الطبية
+لكيلا الجمعية والحرية
+للرجال التزوج
+بالكرامة
+حرية بين
+هذه العيش تنظر
+قيد
+يقررها والصداقة
+اعتُمد وينبغي اجتماعي
+حرمان
+للإدراك بأجر إنتاجه
+التربية القانون
+لإنصافه وتأسيس وسمعته
+أساسه للرجال
+كافة
+المجهود دولي أينما
+وإلى
+بنشاط تجري
+والأمم مثل لحقوق
+الإنسان بشروط بحماية
+شرفه
+كما الوظائف
+حياته ديسمبر
+ولما
+هذه
+غاية جديد إنسان
+حرية
+متهم الوطنية قدمًا
+التملك وضع
+شرعية ويعبر تأدية
+بنظام عمل والأخلاق
+التملك لشخصيته يلجأ
+بحال يضطر ولا
+الانضمام بالكرامة
+عضوا

vendor/icu_normalizer/benches/data/TestRandomWordsUDHR_de.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+Herrschaft Freiheiten Not
+Gewalt
+stets anderer begründet
+erhobenen innerstaatliche
+Heiratsfähige freie
+offenstehen Begrenzung grausamer
+Maßnahmen höchste
+unentbehrlich privat
+erniedrigender
+Verachtung freie
+innezuhaben innerstaatlichen
+kommen
+werden gleichgültig
+Würde überall höchste
+Schutzmaßnahmen den Pflichten
+Wille Bestimmung
+Leibeigenschaft einschließlich für
+gleiche bekräftigt Gewissens
+Wohles
+Generalversammlung
+Volkes
+Völkern gegenwärtig Zusammenarbeit
+Heiratsfähige sowie Jeder
+Stellung
+Lebensstandard
+seinem
+Rede strafbaren Sicherheit
+mit
+Kulthandlungen Grund
+ärztlicher
+Auflösung Anforderungen anzugehören
+Furcht
+keine Geburt
+Wohles Furcht genügen
+befriedigende Medien
+anzugehören Urlaub Vereinigungen
+hinzuwirken verboten Resolution
+kommen
+sozialer vor irgendein
+Bestimmung Bestimmung
+Fall natürliche kein
+Geschlecht Aufhetzung eigenen
+seinen
+über
+Unterlassung Berücksichtigung
+war
+Rufes stets
+Volkes anderer Beschränkungen
+Handlungen dessen
+Die

vendor/icu_normalizer/benches/data/TestRandomWordsUDHR_el.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+προάγει αλληλογραφία
+λογική έχει
+ιδρύει ζωή τεχνική
+δυνατότητες
+περιορισμό συνόλου
+ασκεί παραγνώριση συναφθεί
+αναγνωρίζουν ποινικής εκδηλώνει
+κοινότητας διακυβέρνηση στα
+απέναντι υψηλή
+περιστάσεων αξιόποινη
+σεβασμό
+συντήρησής κατά εξασφαλίσουν
+παραβιάζουν συμπληρώνεται νόμο
+άμεσα
+σημαίνει καθεστώς
+ΑΝΘΡΩΠΙΝΑ θέλησης ανθρωπίνων
+ΔΙΑΚΗΡΥΞΗ αθλιότητα ασφάλιση
+μέσο
+ίση Εχει
+ειρήνης Κάθε
+μέλη μορφή
+όσο
+κρατείται Στο Διακηρύσσει
+οικονομικών έκφρασης εξασφαλίζεται
+κάθε
+περίπτωση απολαμβάνουν
+ποινικό γεροντική
+είναι μαζί δικαστήρια
+μαζί προοπτική
+δική
+βαρβαρότητας
+οικονομικών εξασφαλίσει
+υποχρεώσεις οδήγησαν
+Οικουμενική Διακήρυξης γονείς
+στις μυστική αντιπροσώπους
+Διακήρυξης άδειες βιοτικό
+αναπηρία ομάδα
+πραγματικό
+καλύτερες
+ανάπαυση
+δίκαιες ένα δικαίου
+μετέχει στους
+θρησκευτικών ποινικής
+Κανείς ίσα
+πεποιθήσεις
+πολιτικές ανάλογα δουλεία
+πολιτικές ιατρική ωσότου
+ηθικής χωρίς
+ανδρών ικανό
+καθώς

vendor/icu_normalizer/benches/data/TestRandomWordsUDHR_es.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+duración común
+delito reconocimiento alimentación
+inalienables
+entre seguridad escogidos
+comportarse dignidad
+autónomo gobierno tiempo
+omisiones
+comisión
+Derechos territorios
+debe
+han
+regresar inalienables
+regresar
+desempleo científico
+arbitrariamente proclamada
+están contraerse esposos
+cualesquiera
+salir carácter desarrollo
+solamente justas
+personalidad una
+cuanto
+garantice resolución
+concepción
+tomar impondrá
+cualquier reconocimiento
+obligatoria obligatoria satisfactoria
+acusación sin
+artísticas penal culturales
+pagadas examen
+Además Organización dignidad
+opresión esposos ejercidos
+barbarie están mientras
+por
+idioma
+recursos pagadas
+materia Nada ella
+con injerencias
+inspirándose
+organización
+gozar jurisdicción
+que
+asegurar
+humana libertad
+nadie equivalente
+escoger remuneración
+torturas
+individuos poder
+disfruten seres Preámbulo
+desempleo
+liberados

vendor/icu_normalizer/benches/data/TestRandomWordsUDHR_fr.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+conforme êtres fonctions
+non tout généralisé
+premier lui
+faire hommes d’égalité
+peuple volonté bénéficier
+générale nationales
+cruels plus
+d’encourager opinions
+genre l’esprit
+d’origine effectif
+exigences auront
+résultent situation recevoir
+peuples Chacun
+sont d’égalité
+jouissent
+auront l’esprit
+pays telle
+publiquement
+mariage foi
+travail démocratique religieux
+rémunération
+omissions telles
+L’éducation
+raison complétée donner
+invoqué auront arbitraires
+l’amitié suffisant affaires
+travaille l’accomplissement l’intermédiaire
+race
+opinions celles
+assurer par privée
+valeur
+violant traite premier
+inhérente
+bienfaits l’avènement
+Unies s’il actions
+inquiété l’esclavage
+inquiété
+esclaves lieu
+salaire
+par
+toute
+innocente procédure membres
+arts l’idéal envers
+suffrage territoires inhumains
+d’immixtions l’organisation progrès
+comme égalité Unies
+maternité
+violerait suprême sécurité
+impliquant eux loisirs
+nationalité

vendor/icu_normalizer/benches/data/TestRandomWordsUDHR_he.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+זקנה משפילים
+ינתן חברתי עניניו
+הפוב
+ולהיות זכויות הישגים
+יאסרו מטעמי וללא
+ספרותית השלם
+למנוחה חינם
+וההתאגדות
+לטפח
+באלה במלואן
+יהנו
+ולרווחתם לגבר האדם
+בכבודו שבארצות כבוד
+ובינלאומיים
+בכך לתנאי אישי
+שאינן
+שרירותי
+במשפט
+ולעקרונותיהן מטעם
+שרירותית האשמה יהיה
+החינוך ולבטחון
+סובלנות אשמתו במגילה
+המאוחדות חיוני
+חשוב במקרה
+כלתי העולם
+שמקורה כציבור
+לשויון
+לתקנה
+תלוי ההתאספות
+הדיבור שהוא
+והבלתי והבסיסית
+ולעקרונותיהן יהא וישאף
+ביתנ הבינלאומי
+והזלזול להקנות
+בגלל כולם שיושלם
+לחיים
+בדבר
+לשירות
+זכויות
+לפני
+אדם ולא מזזמנות
+קנינו שהיה ההתאספות
+בינלאומי חיוניות לבקש
+תהיינה
+ובזכות בכורה מהגנה
+מתוך
+ובמצפון מזומנות לאגד
+והחמריים סוציאלי
+אנושיים ובהצבעה
+פראיים

vendor/icu_normalizer/benches/data/TestRandomWordsUDHR_pl.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+uciskowi posiadania prawo
+społecznego największych skazany
+czy
+potrzeby samodzielnie przystępowania
+Krzewi też dokonania
+pełną prawo
+buntu
+moralności
+zapewnienia znaczenie
+nieludzki wypadek Nikt
+zasadności jakikolwiek Każdy
+samowolnie krajem
+międzynarodowego
+członek wielu
+rozwój wynikających obalenia
+rasy
+grudnia która
+jedynie urlopu ani
+małżeńskie stanowi ustaniu
+człowieka postępowych
+prześladowania
+politycznej które zawarcia
+Deklaracja
+ingerować wyłącznie
+studia Nikt
+innego uprawianie zrozumienie
+wybranych swobodę wyznania
+wolni osobowości
+ograniczenie Nie
+równej społecznego uciekać
+będącą POWSZECHNA
+niezdolności poszukiwania międzynarodowej
+konieczne potrzeby posiada
+opinii wychowywania 1948
+międzynarodowej zatrzymać
+przedstawicieli
+przeciw
+wynikających organy pracę
+człowiek grupami
+niezbędnych
+wolności podstawowym
+opinii małżonków wolność
+postępować zdecydowanie komórką
+odniesieniu
+pokoju azyl
+zawodowych powrócić człowiek
+konstytucję
+takiej postaciach powszechnego
+wygnać wygnać
+wspólny poszanowania

vendor/icu_normalizer/benches/data/TestRandomWordsUDHR_ru.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+областях
+будут должен
+обеспечиваются нежели
+котором Уставе
+социального моральных
+совершеннолетия предоставление
+том независимо
+существование
+вмешательства какому ограниченной
+распространять
+находить помощь
+искусством
+унижающим положения искать
+изгнанию член совершеннолетия
+обществом имуществом государственной
+идеи братства
+наслаждаться значение социальной
+осуществления юрисдикцией наказанию
+достойное свою III
+жизнь расторжения инвалидности
+терпимости этого
+целях равны
+обеспечиваются законным
+принуждаем правосубъектности
+пыткам доступа неприкосновенность
+Брак против
+прибегать независимой
+человека человеческой
+быть независимо религии
+публичным
+членам против
+разумом результатом семью
+Принята участие
+беспристрастным тем
+частным основной
+правового
+страной обслуживание
+было свободу полное
+рабочего свободны
+состоянии помощь религиозными
+полное
+владеть власти морали
+меньшей
+братства социальному убежища
+государств
+равны который дети
+терпимости
+получать бесплатным полного
+богослужении
+отдельным

vendor/icu_normalizer/benches/data/TestRandomWordsUDHR_th.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+คิด ใตัอำ เคลื่อนไหว
+บังคับ บาก
+สิ่ง สิ้น
+วัตถุ
+ชาย อาศัย เท่านั้น
+สิน
+เกา
+ดูแล พิธีกรรม
+ภายใน
+เพศ
+หนัก ประสงค์
+เหตุ
+งาน รักษา
+เพศ ภาษา
+นี้
+คู่ สัญชาติ ต้องการ
+วิธี ระหว่าง ตกลง
+ทำนอง
+สืบ กับ ศิลปกรรม
+เหนือ วรรณกรรม
+คิด การก หน้าที่
+ชาติ ศิลปกรรม แต่
+สามัญ สอด
+เหยียด วิธี จุด
+หน้า ถ้า เบื้อง
+ประชุม
+ศิลปกรรม
+เสรีภาพ โหด ก่อ
+เกียรติศักดิ์ ป่วย เอกราช
+ประหัต มโนธรรม การ
+แทน
+ขัดขืน เวลา เสียง
+กฎบัตร พยายาม
+สิน หน้า
+จำเป็น
+ประชาธิปไตย หน่วย
+กรณี จริงจัง
+ทำนอง
+ทาษ
+เพิ่ม
+บรรดา ขวาง
+กักขัง
+มนุษย์
+ชาย ประกัน มนุษยธรรม
+จะบัน มูลฐาน เถื่อน
+พฤติ
+มิได้
+หญิง คู่
+สมา ปฏิบัติ อนึ่ง
+สิ่ง ทาษ

vendor/icu_normalizer/benches/data/TestRandomWordsUDHR_tr.txt

@@ -0,0 +1,54 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+mecburidir ilim
+isnadın sınırları suç
+tutuklanamaz diğer
+memleket korunmasi kullanılamaz
+İnsanlık ilerlemeyi
+bir mülk menfaatlerinin
+usul zümreye herhangi
+mahkeme vicdana ilerleyişe
+zulüm zalimane
+ilim öncelikle çocuk
+mevzubahis ancak
+muamelesi dinlenmeye
+eşitlikle prensiplerine ülkenin
+öğretim bulunmalarına yardım
+memleketler amacıyla
+birbirlerine
+olmalıdır
+bırakılamaz serbestisine
+hürriyetin iyi
+hükmü işbu zalimane
+evlenme memleketi tedbirlerle
+evlenmek ahalisi işini
+hürriyetler
+belirlenmiş kere
+elde cürüme
+tanınan dünyaca yüksek
+müddetinin ailesine
+vicdan kırıcı itibariyle
+geniş inanma
+kendi görevleri Teşkilatı
+yaymak
+öğretim vesayet
+renk kişiliğinin
+tamamlanan
+haklara bulunma
+hükmü uygulanabilecek
+etmiş geliştirilmesini hoşgörü
+sahiptir temel
+giyim
+Bundan temeli
+icaplarını
+mülk karışma tekmil
+vicdana hürriyetine işini
+Herkesin vahşiliklere
+dolaşma dünyanın
+davasının Uluslararasında idamesi
+eşittir
+haklardan hakkı
+kovuşturmalar hürriyetlerden gözönünde
+Evrensel fiilli beyannamesi

vendor/icu_normalizer/benches/data/wotw.txt

@@ -0,0 +1,58 @@
+# This file is part of ICU4X. For terms of use, please see the file
+# called LICENSE at the top level of the ICU4X source tree
+# (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+# The contents of this file have been translated by "Google Translate".
+
+Vào những năm cuối của thế kỷ 19, không ai có thể tin rằng thế giới này
+đang được theo dõi một cách sâu sắc và chặt chẽ bởi những trí thông minh
+lớn hơn con người nhưng cũng nguy hiểm như chính con người; rằng khi con
+người bận rộn với những mối quan tâm khác nhau của họ, họ bị xem xét và
+nghiên cứu kỹ lưỡng, có lẽ gần như một người đàn ông với kính hiển vi có thể
+xem xét kỹ lưỡng những sinh vật nhất thời tụ tập và sinh sôi nảy nở trong
+một giọt nước. Với sự tự mãn vô hạn, con người đi đi lại lại khắp thế giới
+này chỉ vì những công việc nhỏ nhặt của họ, thanh thản với niềm tin chắc
+chắn về đế chế của họ đối với vật chất. Có thể là infusoria dưới kính hiển
+vi cũng làm như vậy. Không ai coi các thế giới cũ hơn trong không gian là
+nguồn gây nguy hiểm cho con người, hoặc nghĩ về chúng chỉ để bác bỏ ý
+tưởng về sự sống đối với chúng là không thể hoặc không thể xảy ra.
+Thật tò mò khi nhớ lại một số thói quen tinh thần của những ngày đã
+qua. Hầu hết những người trên trái đất đều tưởng tượng rằng có thể có
+những người khác trên sao Hỏa, có lẽ thấp kém hơn họ và sẵn sàng chào
+đón một doanh nghiệp truyền giáo. Tuy nhiên, bên kia vịnh không gian,
+những bộ óc đối với tâm trí của chúng ta cũng như tâm trí của chúng ta đối
+với những con thú bị diệt vong, những bộ óc rộng lớn, lạnh lùng và vô cảm,
+nhìn trái đất này với con mắt ghen tị, và dần dần và chắc chắn vạch ra
+những kế hoạch chống lại chúng ta. Và đầu thế kỷ 20 đã xảy ra sự vỡ mộng
+lớn. Hành tinh sao Hỏa, tôi không cần nhắc độc giả, quay xung quanh mặt
+trời ở khoảng cách trung bình 140.000.000 dặm, và ánh sáng và nhiệt mà
+nó nhận được từ mặt trời chỉ bằng một nửa so với thế giới này nhận được.
+Nếu giả thuyết về tinh vân có bất kỳ sự thật nào, nó phải tồn tại lâu
+đời hơn thế giới của chúng ta; và rất lâu trước khi trái đất này ngừng
+nóng chảy, sự sống trên bề mặt của nó hẳn đã bắt đầu quá trình của nó.
+Thực tế là nó chỉ chiếm một phần bảy thể tích của trái đất đã làm tăng
+tốc độ nguội đi của nó đến nhiệt độ mà sự sống có thể bắt đầu. Nó có
+không khí và nước và tất cả những gì cần thiết để hỗ trợ sự tồn tại
+sinh động. Tuy nhiên, con người quá hão huyền và bị mù quáng bởi sự phù
+phiếm của mình, đến nỗi cho đến tận cuối thế kỷ 19, không có nhà văn nào
+bày tỏ bất kỳ ý tưởng nào rằng sự sống thông minh có thể đã phát triển ở đó xa,
+hoặc thực sự là ở tất cả, vượt ra ngoài mức độ trần gian của nó. Người ta
+cũng không hiểu một cách tổng quát rằng vì sao Hỏa già hơn trái đất của chúng
+ta, chỉ bằng một phần tư diện tích bề mặt và ở xa mặt trời hơn, nên điều tất
+yếu dẫn đến là nó không chỉ xa hơn so với thời điểm bắt đầu mà còn gần ngày kết
+thúc hơn. Sự nguội lạnh thế tục mà một ngày nào đó phải vượt qua hành tinh của chúng
+ta đã thực sự đi xa với người hàng xóm của chúng ta. Tình trạng vật lý của nó phần lớn
+vẫn còn là một bí ẩn, nhưng giờ đây chúng ta biết rằng ngay cả ở vùng xích đạo của nó,
+nhiệt độ giữa trưa hầu như không bằng nhiệt độ của mùa đông lạnh nhất của chúng ta.
+Không khí của nó loãng hơn nhiều so với không khí của chúng ta, các đại dương của nó đã
+thu hẹp lại cho đến khi chỉ bao phủ một phần ba bề mặt của nó, và khi các mùa chậm chạp 
+của nó thay đổi, các chỏm tuyết khổng lồ tụ lại và tan chảy ở hai cực và định kỳ làm ngập các vùng ôn đới của nó. 
+Giai đoạn cuối cùng của sự kiệt sức, mà đối với chúng ta vẫn còn quá xa vời, đã trở thành 
+một vấn đề ngày nay đối với các cư dân trên sao Hỏa. Áp lực trước mắt của sự cần 
+thiết đã làm sáng tỏ trí tuệ của họ, mở rộng sức mạnh của họ và làm chai đá trái
+tim họ. Và nhìn xuyên qua không gian với các công cụ, và trí thông minh như chúng 
+ta hiếm khi mơ tới, họ thấy, ở khoảng cách gần nhất chỉ cách họ 35.000.000 dặm 
+về phía mặt trời, một ngôi sao buổi sáng của hy vọng, hành tinh ấm áp hơn của chúng 
+ta, màu xanh lục của thảm thực vật và màu xám của nước , với bầu không khí nhiều 
+mây hùng hồn của sự màu mỡ, với những cái nhìn thoáng qua qua những đám mây 
+trôi dạt của nó là những dải đất rộng lớn đông dân và những vùng biển chật hẹp đông đúc hải quân.

vendor/icu_normalizer/benches/decomposing_normalizer_nfd.rs

@@ -0,0 +1,213 @@
+// This file is part of ICU4X. For terms of use, please see the file
+// called LICENSE at the top level of the ICU4X source tree
+// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+use criterion::{black_box, BenchmarkId, Criterion};
+
+use icu_normalizer::{ComposingNormalizerBorrowed, DecomposingNormalizerBorrowed};
+
+struct BenchDataContent {
+    pub file_name: String,
+    pub nfc: String,
+    pub nfd: String,
+    pub nfkc: String,
+    pub nfkd: String,
+    pub nfc_u16: Vec<u16>,
+    pub nfd_u16: Vec<u16>,
+    pub nfkc_u16: Vec<u16>,
+    pub nfkd_u16: Vec<u16>,
+}
+
+fn strip_headers(content: &str) -> String {
+    content
+        .lines()
+        .filter(|&s| !s.starts_with('#'))
+        .map(|s| s.to_owned())
+        .collect::<Vec<String>>()
+        .join("\n")
+}
+
+fn normalizer_bench_data() -> [BenchDataContent; 15] {
+    let nfc_normalizer = ComposingNormalizerBorrowed::new_nfc();
+    let nfd_normalizer = DecomposingNormalizerBorrowed::new_nfd();
+    let nfkc_normalizer = ComposingNormalizerBorrowed::new_nfkc();
+    let nfkd_normalizer = DecomposingNormalizerBorrowed::new_nfkd();
+
+    let content_latin: (&str, &str) = (
+        "TestNames_Latin",
+        &strip_headers(include_str!("./data/TestNames_Latin.txt")),
+    );
+    let content_jp_h: (&str, &str) = (
+        "TestNames_Japanese_h",
+        &strip_headers(include_str!("./data/TestNames_Japanese_h.txt")),
+    );
+    let content_jp_k: (&str, &str) = (
+        "TestNames_Japanese_k",
+        &strip_headers(include_str!("./data/TestNames_Japanese_k.txt")),
+    );
+    let content_korean: (&str, &str) = (
+        "TestNames_Korean",
+        &strip_headers(include_str!("./data/TestNames_Korean.txt")),
+    );
+    let content_random_words_ar: (&str, &str) = (
+        "TestRandomWordsUDHR_ar",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_ar.txt")),
+    );
+    let content_random_words_de: (&str, &str) = (
+        "TestRandomWordsUDHR_de",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_de.txt")),
+    );
+    let content_random_words_el: (&str, &str) = (
+        "TestRandomWordsUDHR_el",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_el.txt")),
+    );
+    let content_random_words_es: (&str, &str) = (
+        "TestRandomWordsUDHR_es",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_es.txt")),
+    );
+    let content_random_words_fr: (&str, &str) = (
+        "TestRandomWordsUDHR_fr",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_fr.txt")),
+    );
+    let content_random_words_he: (&str, &str) = (
+        "TestRandomWordsUDHR_he",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_he.txt")),
+    );
+    let content_random_words_pl: (&str, &str) = (
+        "TestRandomWordsUDHR_pl",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_pl.txt")),
+    );
+    let content_random_words_ru: (&str, &str) = (
+        "TestRandomWordsUDHR_ru",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_ru.txt")),
+    );
+    let content_random_words_th: (&str, &str) = (
+        "TestRandomWordsUDHR_th",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_th.txt")),
+    );
+    let content_random_words_tr: (&str, &str) = (
+        "TestRandomWordsUDHR_tr",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_tr.txt")),
+    );
+    let content_viet: (&str, &str) = ("wotw", &strip_headers(include_str!("./data/wotw.txt")));
+
+    [
+        content_latin,
+        content_viet,
+        content_jp_k,
+        content_jp_h,
+        content_korean,
+        content_random_words_ru,
+        content_random_words_ar,
+        content_random_words_el,
+        content_random_words_es,
+        content_random_words_fr,
+        content_random_words_tr,
+        content_random_words_th,
+        content_random_words_pl,
+        content_random_words_he,
+        content_random_words_de,
+    ]
+    .map(|(file_name, raw_content)| {
+        let nfc = &nfc_normalizer.normalize(raw_content);
+        let nfd = &nfd_normalizer.normalize(raw_content);
+        let nfkc = &nfkc_normalizer.normalize(raw_content);
+        let nfkd = &nfkd_normalizer.normalize(raw_content);
+        BenchDataContent {
+            file_name: file_name.to_owned(),
+            nfc: nfc.to_string(),
+            nfd: nfd.to_string(),
+            nfkc: nfkc.to_string(),
+            nfkd: nfkd.to_string(),
+            nfc_u16: nfc.encode_utf16().collect(),
+            nfd_u16: nfd.encode_utf16().collect(),
+            nfkc_u16: nfkc.encode_utf16().collect(),
+            nfkd_u16: nfkd.encode_utf16().collect(),
+        }
+    })
+}
+
+fn function_under_bench(normalizer: &DecomposingNormalizerBorrowed, text: &str) {
+    normalizer.normalize(text);
+}
+
+fn function_under_bench_u16(normalizer: &DecomposingNormalizerBorrowed, text: &[u16]) {
+    normalizer.normalize_utf16(text);
+}
+
+pub fn criterion_benchmark(criterion: &mut Criterion) {
+    let group_name = "decomposing_normalizer_nfd";
+
+    let normalizer_under_bench = DecomposingNormalizerBorrowed::new_nfd();
+
+    let mut group = criterion.benchmark_group(group_name);
+
+    for bench_data_content in black_box(normalizer_bench_data()) {
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfc_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher
+                    .iter(|| function_under_bench(&normalizer_under_bench, &bench_data_content.nfc))
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfd_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher
+                    .iter(|| function_under_bench(&normalizer_under_bench, &bench_data_content.nfd))
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkc_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench(&normalizer_under_bench, &bench_data_content.nfkc)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkd_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench(&normalizer_under_bench, &bench_data_content.nfkd)
+                })
+            },
+        );
+
+        // UTF 16
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfc_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfc_u16)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfd_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfd_u16)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkc_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfkc_u16)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkd_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfkd_u16)
+                })
+            },
+        );
+    }
+
+    group.finish();
+}

vendor/icu_normalizer/benches/decomposing_normalizer_nfkd.rs

@@ -0,0 +1,211 @@
+// This file is part of ICU4X. For terms of use, please see the file
+// called LICENSE at the top level of the ICU4X source tree
+// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+use criterion::{black_box, BenchmarkId, Criterion};
+
+use icu_normalizer::{ComposingNormalizerBorrowed, DecomposingNormalizerBorrowed};
+
+struct BenchDataContent {
+    pub file_name: String,
+    pub nfc: String,
+    pub nfd: String,
+    pub nfkc: String,
+    pub nfkd: String,
+    pub nfc_u16: Vec<u16>,
+    pub nfd_u16: Vec<u16>,
+    pub nfkc_u16: Vec<u16>,
+    pub nfkd_u16: Vec<u16>,
+}
+
+fn strip_headers(content: &str) -> String {
+    content
+        .lines()
+        .filter(|&s| !s.starts_with('#'))
+        .map(|s| s.to_owned())
+        .collect::<Vec<String>>()
+        .join("\n")
+}
+
+fn normalizer_bench_data() -> [BenchDataContent; 15] {
+    let nfc_normalizer = ComposingNormalizerBorrowed::new_nfc();
+    let nfd_normalizer = DecomposingNormalizerBorrowed::new_nfd();
+    let nfkc_normalizer = ComposingNormalizerBorrowed::new_nfkc();
+    let nfkd_normalizer = DecomposingNormalizerBorrowed::new_nfkd();
+
+    let content_latin: (&str, &str) = (
+        "TestNames_Latin",
+        &strip_headers(include_str!("./data/TestNames_Latin.txt")),
+    );
+    let content_jp_h: (&str, &str) = (
+        "TestNames_Japanese_h",
+        &strip_headers(include_str!("./data/TestNames_Japanese_h.txt")),
+    );
+    let content_jp_k: (&str, &str) = (
+        "TestNames_Japanese_k",
+        &strip_headers(include_str!("./data/TestNames_Japanese_k.txt")),
+    );
+    let content_korean: (&str, &str) = (
+        "TestNames_Korean",
+        &strip_headers(include_str!("./data/TestNames_Korean.txt")),
+    );
+    let content_random_words_ar: (&str, &str) = (
+        "TestRandomWordsUDHR_ar",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_ar.txt")),
+    );
+    let content_random_words_de: (&str, &str) = (
+        "TestRandomWordsUDHR_de",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_de.txt")),
+    );
+    let content_random_words_el: (&str, &str) = (
+        "TestRandomWordsUDHR_el",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_el.txt")),
+    );
+    let content_random_words_es: (&str, &str) = (
+        "TestRandomWordsUDHR_es",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_es.txt")),
+    );
+    let content_random_words_fr: (&str, &str) = (
+        "TestRandomWordsUDHR_fr",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_fr.txt")),
+    );
+    let content_random_words_he: (&str, &str) = (
+        "TestRandomWordsUDHR_he",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_he.txt")),
+    );
+    let content_random_words_pl: (&str, &str) = (
+        "TestRandomWordsUDHR_pl",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_pl.txt")),
+    );
+    let content_random_words_ru: (&str, &str) = (
+        "TestRandomWordsUDHR_ru",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_ru.txt")),
+    );
+    let content_random_words_th: (&str, &str) = (
+        "TestRandomWordsUDHR_th",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_th.txt")),
+    );
+    let content_random_words_tr: (&str, &str) = (
+        "TestRandomWordsUDHR_tr",
+        &strip_headers(include_str!("./data/TestRandomWordsUDHR_tr.txt")),
+    );
+    let content_viet: (&str, &str) = ("wotw", &strip_headers(include_str!("./data/wotw.txt")));
+
+    [
+        content_latin,
+        content_viet,
+        content_jp_k,
+        content_jp_h,
+        content_korean,
+        content_random_words_ru,
+        content_random_words_ar,
+        content_random_words_el,
+        content_random_words_es,
+        content_random_words_fr,
+        content_random_words_tr,
+        content_random_words_th,
+        content_random_words_pl,
+        content_random_words_he,
+        content_random_words_de,
+    ]
+    .map(|(file_name, raw_content)| {
+        let nfc = &nfc_normalizer.normalize(raw_content);
+        let nfd = &nfd_normalizer.normalize(raw_content);
+        let nfkc = &nfkc_normalizer.normalize(raw_content);
+        let nfkd = &nfkd_normalizer.normalize(raw_content);
+        BenchDataContent {
+            file_name: file_name.to_owned(),
+            nfc: nfc.to_string(),
+            nfd: nfd.to_string(),
+            nfkc: nfkc.to_string(),
+            nfkd: nfkd.to_string(),
+            nfc_u16: nfc.encode_utf16().collect(),
+            nfd_u16: nfd.encode_utf16().collect(),
+            nfkc_u16: nfkc.encode_utf16().collect(),
+            nfkd_u16: nfkd.encode_utf16().collect(),
+        }
+    })
+}
+
+fn function_under_bench(normalizer: &DecomposingNormalizerBorrowed, text: &str) {
+    normalizer.normalize(text);
+}
+
+fn function_under_bench_u16(normalizer: &DecomposingNormalizerBorrowed, text: &[u16]) {
+    normalizer.normalize_utf16(text);
+}
+
+pub fn criterion_benchmark(criterion: &mut Criterion) {
+    let group_name = "decomposing_normalizer_nfkd";
+
+    let normalizer_under_bench = DecomposingNormalizerBorrowed::new_nfkd();
+
+    let mut group = criterion.benchmark_group(group_name);
+    for bench_data_content in black_box(normalizer_bench_data()) {
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfc_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher
+                    .iter(|| function_under_bench(&normalizer_under_bench, &bench_data_content.nfc))
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfd_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher
+                    .iter(|| function_under_bench(&normalizer_under_bench, &bench_data_content.nfd))
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkc_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench(&normalizer_under_bench, &bench_data_content.nfkc)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkd_{}", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench(&normalizer_under_bench, &bench_data_content.nfkd)
+                })
+            },
+        );
+
+        // UTF 16
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfc_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfc_u16)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfd_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfd_u16)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkc_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfkc_u16)
+                })
+            },
+        );
+        group.bench_function(
+            BenchmarkId::from_parameter(format!("from_nfkd_{}_u16", bench_data_content.file_name)),
+            |bencher| {
+                bencher.iter(|| {
+                    function_under_bench_u16(&normalizer_under_bench, &bench_data_content.nfkd_u16)
+                })
+            },
+        );
+    }
+    group.finish();
+}

vendor/icu_normalizer/benches/utf16_throughput.rs

@@ -0,0 +1,170 @@
+// This file is part of ICU4X. For terms of use, please see the file
+// called LICENSE at the top level of the ICU4X source tree
+// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+use icu_normalizer::{ComposingNormalizerBorrowed, DecomposingNormalizerBorrowed};
+
+use criterion::{black_box, Criterion, Throughput};
+
+use smallvec::SmallVec;
+
+//use detone::IterDecomposeVietnamese;
+
+// 2048 times size of u16 fits on one 4KB memory page, which maximizes
+// the run to take average over without introducing cross-page effects.
+const INPUT_SIZE: usize = 2048;
+
+fn generate_bmp_input_nfc(s: &str) -> Vec<u16> {
+    ComposingNormalizerBorrowed::new_nfc()
+        .normalize_iter(s.chars().cycle())
+        .take(INPUT_SIZE)
+        .map(|c| {
+            if c <= '\u{FFFF}' {
+                c as u16
+            } else {
+                unreachable!("Data should stay on the BMP!")
+            }
+        })
+        .collect()
+}
+
+fn generate_bmp_input_nfd(s: &str) -> Vec<u16> {
+    DecomposingNormalizerBorrowed::new_nfd()
+        .normalize_iter(s.chars().cycle())
+        .take(INPUT_SIZE)
+        .map(|c| {
+            if c <= '\u{FFFF}' {
+                c as u16
+            } else {
+                unreachable!("Data should stay on the BMP!")
+            }
+        })
+        .collect()
+}
+
+/// Removes headers and replaces line feed with space.
+/// Do not use for languages that don't use spaces!
+fn prepare_file_contents(content: &str) -> String {
+    content
+        .lines()
+        .filter(|&s| !s.starts_with('#'))
+        .map(|s| s.to_owned())
+        .collect::<Vec<String>>()
+        .join(" ")
+}
+
+fn slice_as_slice(s: &[u16]) -> &[u16] {
+    black_box(s)
+}
+
+fn bench_lang(name: &str, data: &str, c: &mut Criterion) {
+    let input_nfc = generate_bmp_input_nfc(data);
+    let input_nfd = generate_bmp_input_nfd(data);
+    let nfc = ComposingNormalizerBorrowed::new_nfc();
+    let nfd = DecomposingNormalizerBorrowed::new_nfd();
+
+    // Appending to this output is infallible (does not return `Err`) and
+    // this is sized to be large enough not to actually take the the heap
+    // allocation path.
+    let mut output: SmallVec<[u16; INPUT_SIZE * 2]> = SmallVec::new();
+    {
+        let mut group_name = "utf16_throughput_nfc_".to_string();
+        group_name.push_str(name);
+
+        let mut group = c.benchmark_group(&group_name);
+        group.throughput(Throughput::Elements(input_nfc.len() as u64));
+
+        group.bench_function("read", |b| {
+            b.iter(|| {
+                let _ = black_box(
+                    nfc.split_normalized_utf16(slice_as_slice(&input_nfc))
+                        .0
+                        .len(),
+                );
+            })
+        });
+
+        group.bench_function("writing_to_nfc", |b| {
+            b.iter(|| {
+                output.clear(); // Should be trivial and OK to do from within here.
+                let _ = black_box(
+                    nfc.normalize_utf16_to(slice_as_slice(&input_nfc), black_box(&mut output)),
+                );
+            })
+        });
+        group.bench_function("writing_to_nfd", |b| {
+            b.iter(|| {
+                output.clear(); // Should be trivial and OK to do from within here.
+                let _ = black_box(
+                    nfd.normalize_utf16_to(slice_as_slice(&input_nfc), black_box(&mut output)),
+                );
+            })
+        });
+        group.finish();
+    }
+    {
+        let mut group_name = "utf16_throughput_nfd_".to_string();
+        group_name.push_str(name);
+
+        let mut group = c.benchmark_group(&group_name);
+        group.throughput(Throughput::Elements(input_nfd.len() as u64));
+
+        group.bench_function("read", |b| {
+            b.iter(|| {
+                let _ = black_box(
+                    nfd.split_normalized_utf16(slice_as_slice(&input_nfd))
+                        .0
+                        .len(),
+                );
+            })
+        });
+
+        group.bench_function("writing_to_nfd", |b| {
+            b.iter(|| {
+                output.clear(); // Should be trivial and OK to do from within here.
+                let _ = black_box(
+                    nfd.normalize_utf16_to(slice_as_slice(&input_nfd), black_box(&mut output)),
+                );
+            })
+        });
+        group.bench_function("writing_to_nfc", |b| {
+            b.iter(|| {
+                output.clear(); // Should be trivial and OK to do from within here.
+                let _ = black_box(
+                    nfc.normalize_utf16_to(slice_as_slice(&input_nfd), black_box(&mut output)),
+                );
+            })
+        });
+        group.finish();
+    }
+}
+
+static EL: &str = include_str!("./data/TestRandomWordsUDHR_el.txt");
+static EN: &str = "The ICU4X normalizer is an implementation of Unicode Normalization Forms. ";
+static FR: &str = include_str!("./data/TestRandomWordsUDHR_fr.txt");
+static VI: &str = include_str!("./data/wotw.txt");
+static ZH: &str = "單父人呂公善沛令，辟仇，從之客，因家焉。沛中豪傑吏聞令有重客，皆往賀。";
+// zh text from https://www.gutenberg.org/cache/epub/23841/pg23841.txt
+// metadata at https://www.gutenberg.org/ebooks/23841
+// If you replace this text, be sure not to include ASCII spaces and be sure
+// to include punctuation using code points actually used for punctuation in
+// Chinese.
+
+// TODO: Add:
+// * Japanese with realistic proportion of kana voicing marks
+// * Korean, since Hangul is special-cased in the normalizer
+// * Kannada or some other non-Korean BMP language that uses
+//   backward-combining starters (with realistic proportion of such
+//   characters).
+// * Chakma or some other living non-BMP language.
+// * Vietnamese in the orthographic form (i.e. as produced by
+//   the official non-IME keyboard layout that's less common
+//   than the NFC-producing IME.)
+
+pub fn criterion_benchmark(c: &mut Criterion) {
+    bench_lang("el", prepare_file_contents(EL).as_str(), c);
+    bench_lang("en", EN, c);
+    bench_lang("fr", prepare_file_contents(FR).as_str(), c);
+    bench_lang("vi", prepare_file_contents(VI).as_str(), c);
+    bench_lang("zh", ZH, c);
+}

vendor/icu_normalizer/src/properties.rs

@@ -0,0 +1,663 @@
+// This file is part of ICU4X. For terms of use, please see the file
+// called LICENSE at the top level of the ICU4X source tree
+// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+//! Access to the Unicode properties or property-based operations that
+//! are required for NFC and NFD.
+//!
+//! Applications should generally use the full normalizers that are
+//! provided at the top level of this crate. However, the APIs in this
+//! module are provided for callers such as HarfBuzz that specifically
+//! want access to the raw canonical composition operation e.g. for use in a
+//! glyph-availability-guided custom normalizer.
+
+use crate::char_from_u16;
+use crate::char_from_u32;
+use crate::in_inclusive_range;
+use crate::provider::CanonicalCompositions;
+use crate::provider::DecompositionData;
+use crate::provider::DecompositionTables;
+use crate::provider::NonRecursiveDecompositionSupplement;
+use crate::provider::NormalizerNfcV1;
+use crate::provider::NormalizerNfdDataV1;
+use crate::provider::NormalizerNfdSupplementV1;
+use crate::provider::NormalizerNfdTablesV1;
+use crate::trie_value_has_ccc;
+use crate::CanonicalCombiningClass;
+use crate::BACKWARD_COMBINING_MARKER;
+use crate::FDFA_MARKER;
+use crate::HANGUL_L_BASE;
+use crate::HANGUL_N_COUNT;
+use crate::HANGUL_S_BASE;
+use crate::HANGUL_S_COUNT;
+use crate::HANGUL_T_BASE;
+use crate::HANGUL_T_COUNT;
+use crate::HANGUL_V_BASE;
+use crate::HIGH_ZEROS_MASK;
+use crate::LOW_ZEROS_MASK;
+use crate::NON_ROUND_TRIP_MARKER;
+use icu_provider::prelude::*;
+
+/// Borrowed version of the raw canonical composition operation.
+///
+/// Callers should generally use `ComposingNormalizer` instead of this API.
+/// However, this API is provided for callers such as HarfBuzz that specifically
+/// want access to the raw canonical composition operation e.g. for use in a
+/// glyph-availability-guided custom normalizer.
+#[derive(Debug, Copy, Clone)]
+pub struct CanonicalCompositionBorrowed<'a> {
+    canonical_compositions: &'a CanonicalCompositions<'a>,
+}
+
+#[cfg(feature = "compiled_data")]
+impl Default for CanonicalCompositionBorrowed<'static> {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+impl CanonicalCompositionBorrowed<'static> {
+    /// Cheaply converts a [`CanonicalCompositionBorrowed<'static>`] into a [`CanonicalComposition`].
+    ///
+    /// Note: Due to branching and indirection, using [`CanonicalComposition`] might inhibit some
+    /// compile-time optimizations that are possible with [`CanonicalCompositionBorrowed`].
+    pub const fn static_to_owned(self) -> CanonicalComposition {
+        CanonicalComposition {
+            canonical_compositions: DataPayload::from_static_ref(self.canonical_compositions),
+        }
+    }
+
+    /// Constructs a new `CanonicalComposition` using compiled data.
+    ///
+    /// ✨ *Enabled with the `compiled_data` Cargo feature.*
+    ///
+    /// [📚 Help choosing a constructor](icu_provider::constructors)
+    #[cfg(feature = "compiled_data")]
+    pub const fn new() -> Self {
+        Self {
+            canonical_compositions: crate::provider::Baked::SINGLETON_NORMALIZER_NFC_V1,
+        }
+    }
+}
+
+impl CanonicalCompositionBorrowed<'_> {
+    /// Performs canonical composition (including Hangul) on a pair of
+    /// characters or returns `None` if these characters don't compose.
+    /// Composition exclusions are taken into account.
+    ///
+    /// # Examples
+    ///
+    /// ```
+    /// let comp = icu::normalizer::properties::CanonicalCompositionBorrowed::new();
+    ///
+    /// assert_eq!(comp.compose('a', 'b'), None); // Just two non-composing starters
+    /// assert_eq!(comp.compose('a', '\u{0308}'), Some('ä'));
+    /// assert_eq!(comp.compose('ẹ', '\u{0302}'), Some('ệ'));
+    /// assert_eq!(comp.compose('𝅗', '𝅥'), None); // Composition exclusion
+    /// assert_eq!(comp.compose('ে', 'া'), Some('ো')); // Second is starter
+    /// assert_eq!(comp.compose('ᄀ', 'ᅡ'), Some('가')); // Hangul LV
+    /// assert_eq!(comp.compose('가', 'ᆨ'), Some('각')); // Hangul LVT
+    /// ```
+    #[inline(always)]
+    pub fn compose(self, starter: char, second: char) -> Option<char> {
+        crate::compose(
+            self.canonical_compositions.canonical_compositions.iter(),
+            starter,
+            second,
+        )
+    }
+}
+
+/// The raw canonical composition operation.
+///
+/// Callers should generally use `ComposingNormalizer` instead of this API.
+/// However, this API is provided for callers such as HarfBuzz that specifically
+/// want access to the raw canonical composition operation e.g. for use in a
+/// glyph-availability-guided custom normalizer.
+#[derive(Debug)]
+pub struct CanonicalComposition {
+    canonical_compositions: DataPayload<NormalizerNfcV1>,
+}
+
+#[cfg(feature = "compiled_data")]
+impl Default for CanonicalComposition {
+    fn default() -> Self {
+        Self::new().static_to_owned()
+    }
+}
+
+impl CanonicalComposition {
+    /// Constructs a borrowed version of this type for more efficient querying.
+    pub fn as_borrowed(&self) -> CanonicalCompositionBorrowed<'_> {
+        CanonicalCompositionBorrowed {
+            canonical_compositions: self.canonical_compositions.get(),
+        }
+    }
+
+    /// Constructs a new `CanonicalCompositionBorrowed` using compiled data.
+    ///
+    /// ✨ *Enabled with the `compiled_data` Cargo feature.*
+    ///
+    /// [📚 Help choosing a constructor](icu_provider::constructors)
+    #[cfg(feature = "compiled_data")]
+    #[expect(clippy::new_ret_no_self)]
+    pub const fn new() -> CanonicalCompositionBorrowed<'static> {
+        CanonicalCompositionBorrowed::new()
+    }
+
+    icu_provider::gen_buffer_data_constructors!(() -> error: DataError,
+        functions: [
+            new: skip,
+            try_new_with_buffer_provider,
+            try_new_unstable,
+            Self,
+        ]
+    );
+
+    #[doc = icu_provider::gen_buffer_unstable_docs!(UNSTABLE, Self::new)]
+    pub fn try_new_unstable<D>(provider: &D) -> Result<Self, DataError>
+    where
+        D: DataProvider<NormalizerNfcV1> + ?Sized,
+    {
+        let canonical_compositions: DataPayload<NormalizerNfcV1> =
+            provider.load(Default::default())?.payload;
+        Ok(CanonicalComposition {
+            canonical_compositions,
+        })
+    }
+}
+
+/// The outcome of non-recursive canonical decomposition of a character.
+#[allow(clippy::exhaustive_enums)]
+#[derive(Debug, PartialEq, Eq)]
+pub enum Decomposed {
+    /// The character is its own canonical decomposition.
+    Default,
+    /// The character decomposes to a single different character.
+    Singleton(char),
+    /// The character decomposes to two characters.
+    Expansion(char, char),
+}
+
+/// Borrowed version of the raw (non-recursive) canonical decomposition operation.
+///
+/// Callers should generally use `DecomposingNormalizer` instead of this API.
+/// However, this API is provided for callers such as HarfBuzz that specifically
+/// want access to non-recursive canonical decomposition e.g. for use in a
+/// glyph-availability-guided custom normalizer.
+#[derive(Debug)]
+pub struct CanonicalDecompositionBorrowed<'a> {
+    decompositions: &'a DecompositionData<'a>,
+    tables: &'a DecompositionTables<'a>,
+    non_recursive: &'a NonRecursiveDecompositionSupplement<'a>,
+}
+
+#[cfg(feature = "compiled_data")]
+impl Default for CanonicalDecompositionBorrowed<'static> {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+impl CanonicalDecompositionBorrowed<'static> {
+    /// Cheaply converts a [`CanonicalDecompositionBorrowed<'static>`] into a [`CanonicalDecomposition`].
+    ///
+    /// Note: Due to branching and indirection, using [`CanonicalDecomposition`] might inhibit some
+    /// compile-time optimizations that are possible with [`CanonicalDecompositionBorrowed`].
+    pub const fn static_to_owned(self) -> CanonicalDecomposition {
+        CanonicalDecomposition {
+            decompositions: DataPayload::from_static_ref(self.decompositions),
+            tables: DataPayload::from_static_ref(self.tables),
+            non_recursive: DataPayload::from_static_ref(self.non_recursive),
+        }
+    }
+
+    /// Construct from compiled data.
+    ///
+    /// ✨ *Enabled with the `compiled_data` Cargo feature.*
+    ///
+    /// [📚 Help choosing a constructor](icu_provider::constructors)
+    #[cfg(feature = "compiled_data")]
+    pub const fn new() -> Self {
+        const _: () = assert!(
+            crate::provider::Baked::SINGLETON_NORMALIZER_NFD_TABLES_V1
+                .scalars16
+                .const_len()
+                + crate::provider::Baked::SINGLETON_NORMALIZER_NFD_TABLES_V1
+                    .scalars24
+                    .const_len()
+                <= 0xFFF,
+            "future extension"
+        );
+
+        Self {
+            decompositions: crate::provider::Baked::SINGLETON_NORMALIZER_NFD_DATA_V1,
+            tables: crate::provider::Baked::SINGLETON_NORMALIZER_NFD_TABLES_V1,
+            non_recursive: crate::provider::Baked::SINGLETON_NORMALIZER_NFD_SUPPLEMENT_V1,
+        }
+    }
+}
+
+impl CanonicalDecompositionBorrowed<'_> {
+    /// Performs non-recursive canonical decomposition (including for Hangul).
+    ///
+    /// ```
+    ///     use icu::normalizer::properties::Decomposed;
+    ///     let decomp = icu::normalizer::properties::CanonicalDecompositionBorrowed::new();
+    ///
+    ///     assert_eq!(decomp.decompose('e'), Decomposed::Default);
+    ///     assert_eq!(
+    ///         decomp.decompose('ệ'),
+    ///         Decomposed::Expansion('ẹ', '\u{0302}')
+    ///     );
+    ///     assert_eq!(decomp.decompose('각'), Decomposed::Expansion('가', 'ᆨ'));
+    ///     assert_eq!(decomp.decompose('\u{212B}'), Decomposed::Singleton('Å')); // ANGSTROM SIGN
+    ///     assert_eq!(decomp.decompose('\u{2126}'), Decomposed::Singleton('Ω')); // OHM SIGN
+    ///     assert_eq!(decomp.decompose('\u{1F71}'), Decomposed::Singleton('ά')); // oxia
+    /// ```
+    #[inline]
+    pub fn decompose(&self, c: char) -> Decomposed {
+        let lvt = u32::from(c).wrapping_sub(HANGUL_S_BASE);
+        if lvt >= HANGUL_S_COUNT {
+            return self.decompose_non_hangul(c);
+        }
+        // Invariant: lvt ≤ HANGUL_S_COUNT = 1172
+        let t = lvt % HANGUL_T_COUNT;
+        // Invariant: t ≤ (1172 / HANGUL_T_COUNT = 1172 / 28 = 41)
+        if t == 0 {
+            let l = lvt / HANGUL_N_COUNT;
+            // Invariant: v ≤ (1172 / HANGUL_N_COUNT = 1172 / 588 ≈ 2)
+            let v = (lvt % HANGUL_N_COUNT) / HANGUL_T_COUNT;
+            // Invariant: v < (HANGUL_N_COUNT / HANGUL_T_COUNT = 588 / 28 = 21)
+            return Decomposed::Expansion(
+                // Safety: HANGUL_*_BASE are 0x1nnn, addding numbers that are 21 and 41
+                // max will keep it in range, less than 0xD800
+                unsafe { char::from_u32_unchecked(HANGUL_L_BASE + l) },
+                unsafe { char::from_u32_unchecked(HANGUL_V_BASE + v) },
+            );
+        }
+        let lv = lvt - t;
+        // Invariant: lvt < 1172
+        // Safe because values known to be in range
+        Decomposed::Expansion(
+            // Safety: HANGUL_*_BASE are 0x1nnn, addding numbers that are 1172 and 41
+            // max will keep it in range, less than 0xD800
+            unsafe { char::from_u32_unchecked(HANGUL_S_BASE + lv) },
+            unsafe { char::from_u32_unchecked(HANGUL_T_BASE + t) },
+        )
+    }
+
+    /// Performs non-recursive canonical decomposition except Hangul syllables
+    /// are reported as `Decomposed::Default`.
+    #[inline(always)]
+    fn decompose_non_hangul(&self, c: char) -> Decomposed {
+        let decomposition = self.decompositions.trie.get(c);
+        // The REPLACEMENT CHARACTER has `NON_ROUND_TRIP_MARKER` set,
+        // and that flag needs to be ignored here.
+        if (decomposition & !(BACKWARD_COMBINING_MARKER | NON_ROUND_TRIP_MARKER)) == 0 {
+            return Decomposed::Default;
+        }
+        // The loop is only broken out of as goto forward
+        #[expect(clippy::never_loop)]
+        loop {
+            let high_zeros = (decomposition & HIGH_ZEROS_MASK) == 0;
+            let low_zeros = (decomposition & LOW_ZEROS_MASK) == 0;
+            if !high_zeros && !low_zeros {
+                // Decomposition into two BMP characters: starter and non-starter
+                if in_inclusive_range(c, '\u{1F71}', '\u{1FFB}') {
+                    // Look in the other trie due to oxia singleton
+                    // mappings to corresponding character with tonos.
+                    break;
+                }
+                let starter = char_from_u32(decomposition & 0x7FFF);
+                let combining = char_from_u32((decomposition >> 15) & 0x7FFF);
+                return Decomposed::Expansion(starter, combining);
+            }
+            if high_zeros {
+                // Decomposition into one BMP character or non-starter
+                if trie_value_has_ccc(decomposition) {
+                    // Non-starter
+                    if !in_inclusive_range(c, '\u{0340}', '\u{0F81}') {
+                        return Decomposed::Default;
+                    }
+                    return match c {
+                        '\u{0340}' => {
+                            // COMBINING GRAVE TONE MARK
+                            Decomposed::Singleton('\u{0300}')
+                        }
+                        '\u{0341}' => {
+                            // COMBINING ACUTE TONE MARK
+                            Decomposed::Singleton('\u{0301}')
+                        }
+                        '\u{0343}' => {
+                            // COMBINING GREEK KORONIS
+                            Decomposed::Singleton('\u{0313}')
+                        }
+                        '\u{0344}' => {
+                            // COMBINING GREEK DIALYTIKA TONOS
+                            Decomposed::Expansion('\u{0308}', '\u{0301}')
+                        }
+                        '\u{0F73}' => {
+                            // TIBETAN VOWEL SIGN II
+                            Decomposed::Expansion('\u{0F71}', '\u{0F72}')
+                        }
+                        '\u{0F75}' => {
+                            // TIBETAN VOWEL SIGN UU
+                            Decomposed::Expansion('\u{0F71}', '\u{0F74}')
+                        }
+                        '\u{0F81}' => {
+                            // TIBETAN VOWEL SIGN REVERSED II
+                            Decomposed::Expansion('\u{0F71}', '\u{0F80}')
+                        }
+                        _ => Decomposed::Default,
+                    };
+                }
+                let singleton = decomposition as u16;
+                debug_assert_ne!(
+                    singleton, FDFA_MARKER,
+                    "How come we got the U+FDFA NFKD marker here?"
+                );
+                return Decomposed::Singleton(char_from_u16(singleton));
+            }
+            if c == '\u{212B}' {
+                // ANGSTROM SIGN
+                return Decomposed::Singleton('\u{00C5}');
+            }
+            // Only 12 of 14 bits used as of Unicode 16.
+            let offset = (((decomposition & !(0b11 << 30)) >> 16) as usize) - 1;
+            // Only 3 of 4 bits used as of Unicode 16.
+            let len_bits = decomposition & 0b1111;
+            let tables = self.tables;
+            if offset < tables.scalars16.len() {
+                if len_bits != 0 {
+                    // i.e. logical len isn't 2
+                    break;
+                }
+                if let Some(first) = tables.scalars16.get(offset) {
+                    if let Some(second) = tables.scalars16.get(offset + 1) {
+                        // Two BMP starters
+                        return Decomposed::Expansion(char_from_u16(first), char_from_u16(second));
+                    }
+                }
+                // GIGO case
+                debug_assert!(false);
+                return Decomposed::Default;
+            }
+            let len = len_bits + 1;
+            if len > 2 {
+                break;
+            }
+            let offset24 = offset - tables.scalars16.len();
+            if let Some(first_c) = tables.scalars24.get(offset24) {
+                if len == 1 {
+                    return Decomposed::Singleton(first_c);
+                }
+                if let Some(second_c) = tables.scalars24.get(offset24 + 1) {
+                    return Decomposed::Expansion(first_c, second_c);
+                }
+            }
+            // GIGO case
+            debug_assert!(false);
+            return Decomposed::Default;
+        }
+        let non_recursive = self.non_recursive;
+        let non_recursive_decomposition = non_recursive.trie.get(c);
+        if non_recursive_decomposition == 0 {
+            // GIGO case
+            debug_assert!(false);
+            return Decomposed::Default;
+        }
+        let trail_or_complex = (non_recursive_decomposition >> 16) as u16;
+        let lead = non_recursive_decomposition as u16;
+        if lead != 0 && trail_or_complex != 0 {
+            // Decomposition into two BMP characters
+            return Decomposed::Expansion(char_from_u16(lead), char_from_u16(trail_or_complex));
+        }
+        if lead != 0 {
+            // Decomposition into one BMP character
+            return Decomposed::Singleton(char_from_u16(lead));
+        }
+        // Decomposition into two non-BMP characters
+        // Low is offset into a table plus one to keep it non-zero.
+        let offset = usize::from(trail_or_complex - 1);
+        if let Some(first) = non_recursive.scalars24.get(offset) {
+            if let Some(second) = non_recursive.scalars24.get(offset + 1) {
+                return Decomposed::Expansion(first, second);
+            }
+        }
+        // GIGO case
+        debug_assert!(false);
+        Decomposed::Default
+    }
+}
+
+/// The raw (non-recursive) canonical decomposition operation.
+///
+/// Callers should generally use `DecomposingNormalizer` instead of this API.
+/// However, this API is provided for callers such as HarfBuzz that specifically
+/// want access to non-recursive canonical decomposition e.g. for use in a
+/// glyph-availability-guided custom normalizer.
+#[derive(Debug)]
+pub struct CanonicalDecomposition {
+    decompositions: DataPayload<NormalizerNfdDataV1>,
+    tables: DataPayload<NormalizerNfdTablesV1>,
+    non_recursive: DataPayload<NormalizerNfdSupplementV1>,
+}
+
+#[cfg(feature = "compiled_data")]
+impl Default for CanonicalDecomposition {
+    fn default() -> Self {
+        Self::new().static_to_owned()
+    }
+}
+
+impl CanonicalDecomposition {
+    /// Constructs a borrowed version of this type for more efficient querying.
+    pub fn as_borrowed(&self) -> CanonicalDecompositionBorrowed<'_> {
+        CanonicalDecompositionBorrowed {
+            decompositions: self.decompositions.get(),
+            tables: self.tables.get(),
+            non_recursive: self.non_recursive.get(),
+        }
+    }
+
+    /// Construct from compiled data.
+    ///
+    /// ✨ *Enabled with the `compiled_data` Cargo feature.*
+    ///
+    /// [📚 Help choosing a constructor](icu_provider::constructors)
+    #[cfg(feature = "compiled_data")]
+    #[expect(clippy::new_ret_no_self)]
+    pub const fn new() -> CanonicalDecompositionBorrowed<'static> {
+        CanonicalDecompositionBorrowed::new()
+    }
+
+    icu_provider::gen_buffer_data_constructors!(() -> error: DataError,
+        functions: [
+            new: skip,
+            try_new_with_buffer_provider,
+            try_new_unstable,
+            Self,
+        ]
+    );
+
+    #[doc = icu_provider::gen_buffer_unstable_docs!(UNSTABLE, Self::new)]
+    pub fn try_new_unstable<D>(provider: &D) -> Result<Self, DataError>
+    where
+        D: DataProvider<NormalizerNfdDataV1>
+            + DataProvider<NormalizerNfdTablesV1>
+            + DataProvider<NormalizerNfdSupplementV1>
+            + ?Sized,
+    {
+        let decompositions: DataPayload<NormalizerNfdDataV1> =
+            provider.load(Default::default())?.payload;
+        let tables: DataPayload<NormalizerNfdTablesV1> = provider.load(Default::default())?.payload;
+
+        if tables.get().scalars16.len() + tables.get().scalars24.len() > 0xFFF {
+            // The data is from a future where there exists a normalization flavor whose
+            // complex decompositions take more than 0xFFF but fewer than 0x1FFF code points
+            // of space. If a good use case from such a decomposition flavor arises, we can
+            // dynamically change the bit masks so that the length mask becomes 0x1FFF instead
+            // of 0xFFF and the all-non-starters mask becomes 0 instead of 0x1000. However,
+            // since for now the masks are hard-coded, error out.
+            return Err(DataError::custom("future extension"));
+        }
+
+        let non_recursive: DataPayload<NormalizerNfdSupplementV1> =
+            provider.load(Default::default())?.payload;
+
+        Ok(CanonicalDecomposition {
+            decompositions,
+            tables,
+            non_recursive,
+        })
+    }
+}
+
+/// Borrowed version of lookup of the Canonical_Combining_Class Unicode property.
+///
+/// # Example
+///
+/// ```
+/// use icu::properties::props::CanonicalCombiningClass;
+/// use icu::normalizer::properties::CanonicalCombiningClassMapBorrowed;
+///
+/// let map = CanonicalCombiningClassMapBorrowed::new();
+/// assert_eq!(map.get('a'), CanonicalCombiningClass::NotReordered); // U+0061: LATIN SMALL LETTER A
+/// assert_eq!(map.get32(0x0301), CanonicalCombiningClass::Above); // U+0301: COMBINING ACUTE ACCENT
+/// ```
+#[derive(Debug)]
+pub struct CanonicalCombiningClassMapBorrowed<'a> {
+    /// The data trie
+    decompositions: &'a DecompositionData<'a>,
+}
+
+#[cfg(feature = "compiled_data")]
+impl Default for CanonicalCombiningClassMapBorrowed<'static> {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+impl CanonicalCombiningClassMapBorrowed<'static> {
+    /// Cheaply converts a [`CanonicalCombiningClassMapBorrowed<'static>`] into a [`CanonicalCombiningClassMap`].
+    ///
+    /// Note: Due to branching and indirection, using [`CanonicalCombiningClassMap`] might inhibit some
+    /// compile-time optimizations that are possible with [`CanonicalCombiningClassMapBorrowed`].
+    pub const fn static_to_owned(self) -> CanonicalCombiningClassMap {
+        CanonicalCombiningClassMap {
+            decompositions: DataPayload::from_static_ref(self.decompositions),
+        }
+    }
+
+    /// Construct from compiled data.
+    ///
+    /// ✨ *Enabled with the `compiled_data` Cargo feature.*
+    ///
+    /// [📚 Help choosing a constructor](icu_provider::constructors)
+    #[cfg(feature = "compiled_data")]
+    pub const fn new() -> Self {
+        CanonicalCombiningClassMapBorrowed {
+            decompositions: crate::provider::Baked::SINGLETON_NORMALIZER_NFD_DATA_V1,
+        }
+    }
+}
+
+impl CanonicalCombiningClassMapBorrowed<'_> {
+    /// Look up the canonical combining class for a scalar value.
+    ///
+    /// The return value is a u8 representing the canonical combining class,
+    /// you may enable the `"icu_properties"` feature if you would like to use a typed
+    /// `CanonicalCombiningClass`.
+    #[inline(always)]
+    pub fn get_u8(&self, c: char) -> u8 {
+        self.get32_u8(u32::from(c))
+    }
+
+    /// Look up the canonical combining class for a scalar value
+    /// represented as `u32`. If the argument is outside the scalar
+    /// value range, `Not_Reordered` is returned.
+    ///
+    /// The return value is a u8 representing the canonical combining class,
+    /// you may enable the `"icu_properties"` feature if you would like to use a typed
+    /// `CanonicalCombiningClass`.
+    pub fn get32_u8(&self, c: u32) -> u8 {
+        let trie_value = self.decompositions.trie.get32(c);
+        if trie_value_has_ccc(trie_value) {
+            trie_value as u8
+        } else {
+            ccc!(NotReordered, 0).to_icu4c_value()
+        }
+    }
+
+    /// Look up the canonical combining class for a scalar value
+    ///
+    /// ✨ *Enabled with the `icu_properties` Cargo feature.*
+    #[inline(always)]
+    #[cfg(feature = "icu_properties")]
+    pub fn get(&self, c: char) -> CanonicalCombiningClass {
+        CanonicalCombiningClass::from_icu4c_value(self.get_u8(c))
+    }
+
+    /// Look up the canonical combining class for a scalar value
+    /// represented as `u32`. If the argument is outside the scalar
+    /// value range, `CanonicalCombiningClass::NotReordered` is returned.
+    ///
+    /// ✨ *Enabled with the `icu_properties` Cargo feature.*
+    #[cfg(feature = "icu_properties")]
+    pub fn get32(&self, c: u32) -> CanonicalCombiningClass {
+        CanonicalCombiningClass::from_icu4c_value(self.get32_u8(c))
+    }
+}
+
+/// Lookup of the Canonical_Combining_Class Unicode property.
+#[derive(Debug)]
+pub struct CanonicalCombiningClassMap {
+    /// The data trie
+    decompositions: DataPayload<NormalizerNfdDataV1>,
+}
+
+#[cfg(feature = "compiled_data")]
+impl Default for CanonicalCombiningClassMap {
+    fn default() -> Self {
+        Self::new().static_to_owned()
+    }
+}
+
+impl CanonicalCombiningClassMap {
+    /// Constructs a borrowed version of this type for more efficient querying.
+    pub fn as_borrowed(&self) -> CanonicalCombiningClassMapBorrowed<'_> {
+        CanonicalCombiningClassMapBorrowed {
+            decompositions: self.decompositions.get(),
+        }
+    }
+
+    /// Construct from compiled data.
+    ///
+    /// ✨ *Enabled with the `compiled_data` Cargo feature.*
+    ///
+    /// [📚 Help choosing a constructor](icu_provider::constructors)
+    #[cfg(feature = "compiled_data")]
+    #[expect(clippy::new_ret_no_self)]
+    pub const fn new() -> CanonicalCombiningClassMapBorrowed<'static> {
+        CanonicalCombiningClassMapBorrowed::new()
+    }
+
+    icu_provider::gen_buffer_data_constructors!(() -> error: DataError,
+        functions: [
+            new: skip,
+            try_new_with_buffer_provider,
+            try_new_unstable,
+            Self,
+    ]);
+
+    #[doc = icu_provider::gen_buffer_unstable_docs!(UNSTABLE, Self::new)]
+    pub fn try_new_unstable<D>(provider: &D) -> Result<Self, DataError>
+    where
+        D: DataProvider<NormalizerNfdDataV1> + ?Sized,
+    {
+        let decompositions: DataPayload<NormalizerNfdDataV1> =
+            provider.load(Default::default())?.payload;
+        Ok(CanonicalCombiningClassMap { decompositions })
+    }
+}

vendor/icu_normalizer/src/provider.rs

@@ -0,0 +1,216 @@
+// This file is part of ICU4X. For terms of use, please see the file
+// called LICENSE at the top level of the ICU4X source tree
+// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+//! 🚧 \[Unstable\] Data provider struct definitions for this ICU4X component.
+//!
+//! <div class="stab unstable">
+//! 🚧 This code is considered unstable; it may change at any time, in breaking or non-breaking ways,
+//! including in SemVer minor releases. While the serde representation of data structs is guaranteed
+//! to be stable, their Rust representation might not be. Use with caution.
+//! </div>
+//!
+//! Read more about data providers: [`icu_provider`]
+
+// Provider structs must be stable
+#![allow(clippy::exhaustive_structs, clippy::exhaustive_enums)]
+
+use icu_collections::char16trie::Char16Trie;
+use icu_collections::codepointtrie::CodePointTrie;
+use icu_provider::prelude::*;
+use zerovec::ZeroVec;
+
+#[cfg(feature = "compiled_data")]
+#[derive(Debug)]
+/// Baked data
+///
+/// <div class="stab unstable">
+/// 🚧 This code is considered unstable; it may change at any time, in breaking or non-breaking ways,
+/// including in SemVer minor releases. In particular, the `DataProvider` implementations are only
+/// guaranteed to match with this version's `*_unstable` providers. Use with caution.
+/// </div>
+pub struct Baked;
+
+#[cfg(feature = "compiled_data")]
+#[allow(unused_imports)]
+const _: () = {
+    use icu_normalizer_data::*;
+    pub mod icu {
+        pub use crate as normalizer;
+        pub use icu_collections as collections;
+    }
+    make_provider!(Baked);
+    impl_normalizer_nfc_v1!(Baked);
+    impl_normalizer_nfd_data_v1!(Baked);
+    impl_normalizer_nfd_supplement_v1!(Baked);
+    impl_normalizer_nfd_tables_v1!(Baked);
+    impl_normalizer_nfkd_data_v1!(Baked);
+    impl_normalizer_nfkd_tables_v1!(Baked);
+    impl_normalizer_uts46_data_v1!(Baked);
+};
+
+icu_provider::data_marker!(
+    /// Marker for data for canonical decomposition.
+    NormalizerNfdDataV1,
+    "normalizer/nfd/data/v1",
+    DecompositionData<'static>,
+    is_singleton = true
+);
+icu_provider::data_marker!(
+    /// Marker for additional data for canonical decomposition.
+    NormalizerNfdTablesV1,
+    "normalizer/nfd/tables/v1",
+    DecompositionTables<'static>,
+    is_singleton = true
+);
+icu_provider::data_marker!(
+    /// Marker for data for compatibility decomposition.
+    NormalizerNfkdDataV1,
+    "normalizer/nfkd/data/v1",
+    DecompositionData<'static>,
+    is_singleton = true
+);
+icu_provider::data_marker!(
+    /// Marker for additional data for compatibility decomposition.
+    NormalizerNfkdTablesV1,
+    "normalizer/nfkd/tables/v1",
+    DecompositionTables<'static>,
+    is_singleton = true
+);
+icu_provider::data_marker!(
+    /// Marker for data for UTS-46 decomposition.
+    NormalizerUts46DataV1,
+    "normalizer/uts46/data/v1",
+    DecompositionData<'static>,
+    is_singleton = true
+);
+icu_provider::data_marker!(
+    /// Marker for data for composition.
+    NormalizerNfcV1,
+    "normalizer/nfc/v1",
+    CanonicalCompositions<'static>,
+    is_singleton = true
+);
+icu_provider::data_marker!(
+    /// Marker for additional data for non-recusrsive composition.
+    NormalizerNfdSupplementV1,
+    "normalizer/nfd/supplement/v1",
+    NonRecursiveDecompositionSupplement<'static>,
+    is_singleton = true
+);
+
+#[cfg(feature = "datagen")]
+/// The latest minimum set of markers required by this component.
+pub const MARKERS: &[DataMarkerInfo] = &[
+    NormalizerNfcV1::INFO,
+    NormalizerNfdDataV1::INFO,
+    NormalizerNfdTablesV1::INFO,
+    NormalizerNfkdDataV1::INFO,
+    NormalizerNfkdTablesV1::INFO,
+    NormalizerNfdSupplementV1::INFO,
+    NormalizerUts46DataV1::INFO,
+];
+
+/// Decomposition data
+///
+/// <div class="stab unstable">
+/// 🚧 This code is considered unstable; it may change at any time, in breaking or non-breaking ways,
+/// including in SemVer minor releases. While the serde representation of data structs is guaranteed
+/// to be stable, their Rust representation might not be. Use with caution.
+/// </div>
+#[derive(Debug, PartialEq, Clone, yoke::Yokeable, zerofrom::ZeroFrom)]
+#[cfg_attr(feature = "datagen", derive(serde::Serialize, databake::Bake))]
+#[cfg_attr(feature = "datagen", databake(path = icu_normalizer::provider))]
+#[cfg_attr(feature = "serde", derive(serde::Deserialize))]
+pub struct DecompositionData<'data> {
+    /// Trie for decomposition.
+    #[cfg_attr(feature = "serde", serde(borrow))]
+    pub trie: CodePointTrie<'data, u32>,
+    /// The passthrough bounds of NFD/NFC are lowered to this
+    /// maximum instead. (16-bit, because cannot be higher
+    /// than 0x0300, which is the bound for NFC.)
+    pub passthrough_cap: u16,
+}
+
+icu_provider::data_struct!(
+    DecompositionData<'_>,
+    #[cfg(feature = "datagen")]
+);
+
+/// The expansion tables for cases where the decomposition isn't
+/// contained in the trie value
+///
+/// <div class="stab unstable">
+/// 🚧 This code is considered unstable; it may change at any time, in breaking or non-breaking ways,
+/// including in SemVer minor releases. While the serde representation of data structs is guaranteed
+/// to be stable, their Rust representation might not be. Use with caution.
+/// </div>
+#[derive(Debug, PartialEq, Clone, yoke::Yokeable, zerofrom::ZeroFrom)]
+#[cfg_attr(feature = "datagen", derive(serde::Serialize, databake::Bake))]
+#[cfg_attr(feature = "datagen", databake(path = icu_normalizer::provider))]
+#[cfg_attr(feature = "serde", derive(serde::Deserialize))]
+pub struct DecompositionTables<'data> {
+    /// Decompositions that are fully within the BMP
+    #[cfg_attr(feature = "serde", serde(borrow))]
+    pub scalars16: ZeroVec<'data, u16>,
+    /// Decompositions with at least one character outside
+    /// the BMP
+    #[cfg_attr(feature = "serde", serde(borrow))]
+    pub scalars24: ZeroVec<'data, char>,
+}
+
+icu_provider::data_struct!(
+    DecompositionTables<'_>,
+    #[cfg(feature = "datagen")]
+);
+
+/// Non-Hangul canonical compositions
+///
+/// <div class="stab unstable">
+/// 🚧 This code is considered unstable; it may change at any time, in breaking or non-breaking ways,
+/// including in SemVer minor releases. While the serde representation of data structs is guaranteed
+/// to be stable, their Rust representation might not be. Use with caution.
+/// </div>
+#[derive(Debug, PartialEq, Clone, yoke::Yokeable, zerofrom::ZeroFrom)]
+#[cfg_attr(feature = "datagen", derive(serde::Serialize, databake::Bake))]
+#[cfg_attr(feature = "datagen", databake(path = icu_normalizer::provider))]
+#[cfg_attr(feature = "serde", derive(serde::Deserialize))]
+pub struct CanonicalCompositions<'data> {
+    /// Trie keys are two-`char` strings with the second
+    /// character coming first. The value, if any, is the
+    /// (non-Hangul) canonical composition.
+    #[cfg_attr(feature = "serde", serde(borrow))]
+    pub canonical_compositions: Char16Trie<'data>,
+}
+
+icu_provider::data_struct!(
+    CanonicalCompositions<'_>,
+    #[cfg(feature = "datagen")]
+);
+
+/// Non-recursive canonical decompositions that differ from
+/// `DecompositionData`.
+///
+/// <div class="stab unstable">
+/// 🚧 This code is considered unstable; it may change at any time, in breaking or non-breaking ways,
+/// including in SemVer minor releases. While the serde representation of data structs is guaranteed
+/// to be stable, their Rust representation might not be. Use with caution.
+/// </div>
+#[derive(Debug, PartialEq, Clone, yoke::Yokeable, zerofrom::ZeroFrom)]
+#[cfg_attr(feature = "datagen", derive(serde::Serialize, databake::Bake))]
+#[cfg_attr(feature = "datagen", databake(path = icu_normalizer::provider))]
+#[cfg_attr(feature = "serde", derive(serde::Deserialize))]
+pub struct NonRecursiveDecompositionSupplement<'data> {
+    /// Trie for the supplementary non-recursive decompositions
+    #[cfg_attr(feature = "serde", serde(borrow))]
+    pub trie: CodePointTrie<'data, u32>,
+    /// Decompositions with at least one character outside
+    /// the BMP
+    #[cfg_attr(feature = "serde", serde(borrow))]
+    pub scalars24: ZeroVec<'data, char>,
+}
+
+icu_provider::data_struct!(
+    NonRecursiveDecompositionSupplement<'_>,
+    #[cfg(feature = "datagen")]
+);

vendor/icu_normalizer/src/uts46.rs

@@ -0,0 +1,177 @@
+// This file is part of ICU4X. For terms of use, please see the file
+// called LICENSE at the top level of the ICU4X source tree
+// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ).
+
+//! Bundles the part of UTS 46 that makes sense to implement as a
+//! normalization.
+//!
+//! This is meant to be used as a building block of an UTS 46
+//! implementation, such as the `idna` crate.
+
+use crate::ComposingNormalizer;
+use crate::ComposingNormalizerBorrowed;
+use crate::NormalizerNfcV1;
+use crate::NormalizerNfdTablesV1;
+use crate::NormalizerNfkdTablesV1;
+use crate::NormalizerUts46DataV1;
+use icu_provider::DataError;
+use icu_provider::DataProvider;
+
+// Implementation note: Despite merely wrapping a `ComposingNormalizer`,
+// having a `Uts46Mapper` serves two purposes:
+//
+// 1. Denying public access to parts of the `ComposingNormalizer` API
+//    that don't work when the data contains markers for ignorables.
+// 2. Providing a place where additional iterator pre-processing or
+//    post-processing can take place if needed in the future. (When
+//    writing this, it looked like such processing was needed but
+//    now isn't needed after all.)
+
+/// A borrowed version of a mapper that knows how to performs the
+/// subsets of UTS 46 processing documented on the methods.
+#[derive(Debug)]
+pub struct Uts46MapperBorrowed<'a> {
+    normalizer: ComposingNormalizerBorrowed<'a>,
+}
+
+#[cfg(feature = "compiled_data")]
+impl Default for Uts46MapperBorrowed<'static> {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+impl Uts46MapperBorrowed<'static> {
+    /// Cheaply converts a [`Uts46MapperBorrowed<'static>`] into a [`Uts46Mapper`].
+    ///
+    /// Note: Due to branching and indirection, using [`Uts46Mapper`] might inhibit some
+    /// compile-time optimizations that are possible with [`Uts46MapperBorrowed`].
+    pub const fn static_to_owned(self) -> Uts46Mapper {
+        Uts46Mapper {
+            normalizer: self.normalizer.static_to_owned(),
+        }
+    }
+
+    /// Construct with compiled data.
+    #[cfg(feature = "compiled_data")]
+    pub const fn new() -> Self {
+        Uts46MapperBorrowed {
+            normalizer: ComposingNormalizerBorrowed::new_uts46(),
+        }
+    }
+}
+
+impl Uts46MapperBorrowed<'_> {
+    /// Returns an iterator adaptor that turns an `Iterator` over `char`
+    /// into an iterator yielding a `char` sequence that gets the following
+    /// operations from the "Map" and "Normalize" steps of the "Processing"
+    /// section of UTS 46 lazily applied to it:
+    ///
+    /// 1. The _ignored_ characters are ignored.
+    /// 2. The _mapped_ characters are mapped.
+    /// 3. The _disallowed_ characters are replaced with U+FFFD,
+    ///    which itself is a disallowed character.
+    /// 4. The _deviation_ characters are treated as _mapped_ or _valid_
+    ///    as appropriate.
+    /// 5. The _disallowed_STD3_valid_ characters are treated as allowed.
+    /// 6. The _disallowed_STD3_mapped_ characters are treated as
+    ///    _mapped_.
+    /// 7. The result is normalized to NFC.
+    ///
+    /// Notably:
+    ///
+    /// * The STD3 or WHATWG ASCII deny list should be implemented as a
+    ///   post-processing step.
+    /// * Transitional processing is not performed. Transitional mapping
+    ///   would be a pre-processing step, but transitional processing is
+    ///   deprecated, and none of Firefox, Safari, or Chrome use it.
+    pub fn map_normalize<'delegate, I: Iterator<Item = char> + 'delegate>(
+        &'delegate self,
+        iter: I,
+    ) -> impl Iterator<Item = char> + 'delegate {
+        self.normalizer
+            .normalize_iter_private(iter, crate::IgnorableBehavior::Ignored)
+    }
+
+    /// Returns an iterator adaptor that turns an `Iterator` over `char`
+    /// into an iterator yielding a `char` sequence that gets the following
+    /// operations from the NFC check and statucs steps of the "Validity
+    /// Criteria" section of UTS 46 lazily applied to it:
+    ///
+    /// 1. The _ignored_ characters are treated as _disallowed_.
+    /// 2. The _mapped_ characters are mapped.
+    /// 3. The _disallowed_ characters are replaced with U+FFFD,
+    ///    which itself is a disallowed character.
+    /// 4. The _deviation_ characters are treated as _mapped_ or _valid_
+    ///    as appropriate.
+    /// 5. The _disallowed_STD3_valid_ characters are treated as allowed.
+    /// 6. The _disallowed_STD3_mapped_ characters are treated as
+    ///    _mapped_.
+    /// 7. The result is normalized to NFC.
+    ///
+    /// Notably:
+    ///
+    /// * The STD3 or WHATWG ASCII deny list should be implemented as a
+    ///   post-processing step.
+    /// * Transitional processing is not performed. Transitional mapping
+    ///   would be a pre-processing step, but transitional processing is
+    ///   deprecated, and none of Firefox, Safari, or Chrome use it.
+    /// * The output needs to be compared with input to see if anything
+    ///   changed. This check catches failures to adhere to the normalization
+    ///   and status requirements. In particular, this comparison results
+    ///   in _mapped_ characters resulting in error like "Validity Criteria"
+    ///   requires.
+    pub fn normalize_validate<'delegate, I: Iterator<Item = char> + 'delegate>(
+        &'delegate self,
+        iter: I,
+    ) -> impl Iterator<Item = char> + 'delegate {
+        self.normalizer
+            .normalize_iter_private(iter, crate::IgnorableBehavior::ReplacementCharacter)
+    }
+}
+
+/// A mapper that knows how to performs the subsets of UTS 46 processing
+/// documented on the methods.
+#[derive(Debug)]
+pub struct Uts46Mapper {
+    normalizer: ComposingNormalizer,
+}
+
+#[cfg(feature = "compiled_data")]
+impl Default for Uts46Mapper {
+    fn default() -> Self {
+        Self::new().static_to_owned()
+    }
+}
+
+impl Uts46Mapper {
+    /// Constructs a borrowed version of this type for more efficient querying.
+    pub fn as_borrowed(&self) -> Uts46MapperBorrowed<'_> {
+        Uts46MapperBorrowed {
+            normalizer: self.normalizer.as_borrowed(),
+        }
+    }
+
+    /// Construct with compiled data.
+    #[cfg(feature = "compiled_data")]
+    #[expect(clippy::new_ret_no_self)]
+    pub const fn new() -> Uts46MapperBorrowed<'static> {
+        Uts46MapperBorrowed::new()
+    }
+
+    /// Construct with provider.
+    #[doc = icu_provider::gen_buffer_unstable_docs!(UNSTABLE, Self::new)]
+    pub fn try_new<D>(provider: &D) -> Result<Self, DataError>
+    where
+        D: DataProvider<NormalizerUts46DataV1>
+            + DataProvider<NormalizerNfdTablesV1>
+            + DataProvider<NormalizerNfkdTablesV1>
+            // UTS 46 tables merged into NormalizerNfkdTablesV1
+            + DataProvider<NormalizerNfcV1>
+            + ?Sized,
+    {
+        let normalizer = ComposingNormalizer::try_new_uts46_unstable(provider)?;
+
+        Ok(Uts46Mapper { normalizer })
+    }
+}

vendor/icu_normalizer/tests/data/NormalizationTest.txt

@@ -0,0 +1,4 @@
+# This is a placeholder in the interest of keeping the repository size smaller.
+# Replace this file with the contents of
+# https://www.unicode.org/Public/UCD/latest/ucd/NormalizationTest.txt to actually
+# run the conformance test.

vendor/icu_normalizer/tests/data/README.md

@@ -0,0 +1,2 @@
+The test data comes from
+https://www.unicode.org/Public/UCD/latest/ucd/NormalizationTest.txt