rust
S
#rustc 1.98.1 (48a229cea 2026-09-01)СћЪЦЭRL ё.хSлAњwjL -df1bddb45adbe94bС Бђ<х%Z>И64ЫV` -2d2ab9fb0fe1def5С ћ
is_xid_startС is_xid_continueС tablesС Align8С
ѓ Align64С
ѓ UNICODE_VERSIONС ASCII_STARTС ASCII_CONTINUEС CHUNKС
ZERO_STARTС
ZERO_CONTINUEС
TRIE_STARTС
TRIE_CONTINUEС LEAFС " T ѕ ѕ ќх5 ѕѕѕ d + _С ўџџўџџ !0x7fffffe07fffffe0000000000000000С џўџџўџџф !0x7fffffe87fffffe03ff000000000000С @ 64С д 17С д 17С џ д 511С д 1793С д 7968Смї + ј ѓ iKц;Ѓ `н d н k r фИ н ( ѓ Џ 6GК6ymн a н h o
ћ З|ч\ пt] Э\Ё] є,Ў] в$Е] " КlЛ] ЅTЪ]
lж] Tх] ќ Ѕe ќ І Ѓ [![github]](https://github.com/dtolnay/unicode-ident) [![crates-io]](https://crates.io/crates/unicode-ident) [![docs-rs]](https://docs.rs/unicode-ident)С Ї ќЋj g [github]: https://img.shields.io/badge/github-8da0cb?style=for-the-badge&labelColor=555555&logo=githubС ќn k [crates-io]: https://img.shields.io/badge/crates.io-fc8d62?style=for-the-badge&labelColor=555555&logo=rustС ќ
m j [docs-rs]: https://img.shields.io/badge/docs.rs-66c2a5?style=for-the-badge&labelColor=555555&logo=docs.rsС ѓ Dї
С ќN K Implementation of [Unicode Standard Annex #31][tr31] for determining whichС ќг@ = `char` values are valid in programming language identifiers.С ќ1 . [tr31]: https://www.unicode.org/reports/tr31/С Ъ ќЮN K This crate is a better optimized implementation of the older `unicode-xid`С ќL I crate. This crate uses less static storage, and is able to classify bothС ќъP M ASCII and non-ASCII codepoints with better performance, 6× faster thanС Л `unicode-xid`.С Ю DвЦ л ќп ## Comparison of performanceС ќJ G The following table shows a comparison between five Unicode identifierС ЄЯ implementations.С ф ќш$ ! - `unicode-ident` is this crate;С ќ I F - [`unicode-xid`] is a widely used crate run by the "unicode-rs" org;С ќз C @ - `ucd-trie` and `fst` are two data structures supported by theС ф
[`ucd-generate`] tool;С ќИ
= : - [`roaring`] is a Rust implementation of Roaring bitmap.С і
ќњ
P M The *static storage* column shows the total size of `static` tables that theС ќЫ = : crate bakes into your binary, measured in 1000s of bytes.С ќ J G The remaining columns show the **cost per call** to evaluate whether aС ќи G D single `char` has the XID\_Start or XID\_Continue Unicode property,С ќ
M J comparing across different ratios of ASCII to non-ASCII codepoints in theС |ю
input data.С ў
ќ> ; [`unicode-xid`]: https://github.com/unicode-rs/unicode-xidС ќС@ = [`ucd-generate`]: https://github.com/BurntSushi/ucd-generateС ќ< 9 [`roaring`]: https://github.com/RoaringBitmap/roaring-rsС П ќУA > | | static storage | 0% nonascii | 1% | 10% | 100% nonascii |С ь
|---|---|---|---|---|---|С ќЃL I | **`unicode-ident`** | 10.3 K | 0.36 ns | 0.37 ns | 0.37 ns | 0.43 ns |С ќ№J G | **`unicode-xid`** | 12.2 K | 1.63 ns | 1.70 ns | 1.82 ns | 4.56 ns |С ќЛG D | **`ucd-trie`** | 10.8 K | 1.01 ns | 0.73 ns | 0.97 ns | 1.09 ns |С ќA > | **`fst`** | 149 K | 22.0 ns | 21.9 ns | 20.9 ns | 10.5 ns |С ќХF C | **`roaring`** | 66.1 K | 1.91 ns | 1.90 ns | 1.94 ns | 2.67 ns |С ќN K Source code for the benchmark is provided in the *bench* directory of thisС ќп: 7 repo and may be repeated by running `cargo criterion`.С DЦ Ї ќЋ$ ! ## Comparison of data structuresС а Єд #### unicode-xidС щ ќэO L They use a sorted array of character ranges, and do a binary search to lookС ќНB ? up whether a given character lands inside one of those ranges.С \ ```rustС ќ" # const _: &str = stringify! {С ќГ6 3 static XID_Continue_table: [(char, char); 763] = [С ќъ% " ('\u{30}', '\u{39}'), // 0-9С ќ% " ('\u{41}', '\u{5a}'), // A-ZС <Ж # "С \О тІС <Ъ& ќв# ('\u{e0100}', '\u{e01ef}'),С 4і ];С D§ # };С < ```С ќL I The static storage used by this data structure scales with the number ofС ќпL I contiguous ranges of identifier codepoints in Unicode. Every table entryС ќЌL I consumes 8 bytes, because it consists of a pair of 32-bit `char` values.С љ ќ§I F In some ranges of the Unicode codepoint space, this is quite a sparseС ќЧK H representation – there are some ranges where tens of thousands ofС ќM J adjacent codepoints are all valid identifier characters. In other places,С ќсL I the representation is quite inefficient. A characater like `ТЕ` (U+00B5)С ќЎ L I which is surrounded by non-identifier codepoints consumes 64 bits in theС ќћ : 7 table, while it would be just 1 bit in a dense bitmap.С Ж ќК N K On a system with 64-byte cache lines, binary searching the table touches 7С ќ F C cache lines on average. Each cache line fits only 8 table entries.С ќа N K Additionally, the branching performed during the binary search is probablyС ќ 1 . mostly unpredictable to the branch predictor.С б ќе M J Overall, the crate ends up being about 6× slower on non-ASCII inputС ќЃ " compared to the fastest crate.С Ц ќЪ N K A potential improvement would be to pack the table entries more compactly.С ќ!O L Rust's `char` type is a 21-bit integer padded to 32 bits, which means everyС ќщ!L I table entry is holding 22 bits of wasted space, adding up to 3.9 K. TheyС ќЖ"M J could instead fit every table entry into 6 bytes, leaving out some of theС ќ#P M padding, for a 25% improvement in space used. With some cleverness it may beС ќе#K H possible to fit in 5 bytes or even 4 bytes by storing a low char and anС ќЁ$N K extent, instead of low char and high char. I don't expect that performanceС ќ№$P M would improve much but this could be the most efficient for space across allС ќС%3 0 the libraries, needing only about 7 K to store.С ѕ% љ% #### ucd-trieС & ќ&K H Their data structure is a compressed trie set specifically tailored forС ќл&@ = Unicode codepoints. The design is credited to Raph Levien inС м' [rust-lang/rust#33098].С И' ќМ'H E [rust-lang/rust#33098]: https://github.com/rust-lang/rust/pull/33098С
( \(А$ Ф( pub struct TrieSet {С ќЎ() & tree1_level1: &'static [u64; 32],С ќи() & tree2_level1: &'static [u8; 992],С ќ)% " tree2_level2: &'static [u64],С ќЈ)) & tree3_level1: &'static [u8; 256],С ќв)$ ! tree3_level2: &'static [u8],С ќї)% " tree3_level3: &'static [u64],С ,* }С <Ѓ*' Ћ* ќЏ*P M It represents codepoint sets using a trie to achieve prefix compression. TheС ќ+K H final states of the trie are embedded in leaves or "chunks", where eachС ќЬ+N K chunk is a 64-bit integer. Each bit position of the integer corresponds toС ќ,M J whether a particular codepoint is in the set or not. These chunks are notС ќщ,O L just a compact representation of the final states of the trie, but are alsoС ќЙ-I F a form of suffix compression. In particular, if multiple ranges of 64С ќ.P M contiguous codepoints have the same Unicode properties, then they all map toС ќд.2 / the same chunk in the final level of the trie.С / ќ/N K Being tailored for Unicode codepoints, this trie is partitioned into threeС ќк/O L disjoint sets: tree1, tree2, tree3. The first set corresponds to codepointsС ќЊ0F C \[0, 0x800), the second \[0x800, 0x10000) and the third \[0x10000,С ќё0N K 0x110000). These partitions conveniently correspond to the space of 1 or 2С ќР1M J byte UTF-8 encoded codepoints, 3 byte UTF-8 encoded codepoints and 4 byteС ќ2+ ( UTF-8 encoded codepoints, respectively.С К2 ќО2O L Lookups in this data structure are significantly more efficient than binaryС ќ3P M search. A lookup touches either 1, 2, or 3 cache lines based on which of theС ќп3& # trie partitions is being accessed.С 4 ќ4P M One possible performance improvement would be for this crate to expose a wayС ќл4L I to query based on a UTF-8 encoded string, returning the Unicode propertyС ќЈ5P M corresponding to the first character in the string. Without such an API, theС ќљ5N K caller is required to tokenize their UTF-8 encoded input data into `char`,С ќШ6M J hand the `char` into `ucd-trie`, only for `ucd-trie` to undo that work byС ќ7O L converting back into the variable-length representation for trie traversal.С ц7 dъ7 #### fstС ї7 ќћ7L I Uses a [finite state transducer][fst]. This representation is built intoС ќШ8J G [ucd-generate] but I am not aware of any advantage over the `ucd-trie`С ќ9M J representation. In particular `ucd-trie` is optimized for storing UnicodeС ќс9" properties while `fst` is not.С : ќ:, ) [fst]: https://github.com/BurntSushi/fstС ќЕ:> ; [ucd-generate]: https://github.com/BurntSushi/ucd-generateС є: ќј:M J As far as I can tell, the main thing that causes `fst` to have large sizeС ќЦ;M J and slow lookups for this use case relative to `ucd-trie` is that it doesС ќD A structure designed for storing sets of 32-bit unsigned integers.С §> ќ?6 3 [Roaring Bitmap]: https://roaringbitmap.org/about/С И? ќМ?P M Roaring bitmaps are compressed bitmaps which tend to outperform conventionalС ќ@P M compressed bitmaps such as WAH, EWAH or Concise. In some instances, they canС ќо@I F be hundreds of times faster and they often offer significantly betterС ЈA
compression.С ЙA ќНAI F In this use case the performance was reasonably competitive but stillС ќBI F substantially slower than the Unicode-optimized crates. Meanwhile theС ќбBO L compression was significantly worse, requiring 6× as much storage forС МЁC the data structure.С ЙC ќНCP M I also benchmarked the [`croaring`] crate which is an FFI wrapper around theС ќDM J C reference implementation of Roaring Bitmap. This crate was consistentlyС ќмDP M about 15% slower than pure-Rust `roaring`, which could just be FFI overhead.С ќE" I did not investigate further.С аE ќдE3 0 [`croaring`]: https://crates.io/crates/croaringС F ДF #### unicode-identС ЃF ќЇFO L This crate is most similar to the `ucd-trie` library, in that it's based onС ќїFO L bitmaps stored in the leafs of a trie representation, achieving both prefixС ќЧG' $ compression and suffix compression.С яG фѓG The key differences are:С H ќHP M - Uses a single 2-level trie, rather than 3 disjoint partitions of differentС хH depth each.С ќїHE B - Uses significantly larger chunks: 512 bits rather than 64 bits.С ќНIE B - Compresses the XID\_Start and XID\_Continue properties togetherС ќJO L simultaneously, rather than duplicating identical trie leaf chunks acrossС tгJ the two.С тJ ќцJO L The following diagram show the XID\_Start and XID\_Continue Unicode booleanС ќЖK8 5 properties in uncompressed form, in row-major order:С яK \ѓK С ќџK4 1 | XID_Start | XID_Continue |
|---|
С DДL С ќНL  | С ќоMЃ  | С LO
С dO
С O ќOP M Uncompressed, these would take 140 K to store, which is beyond what would beС ќюOM J reasonable. However, as you can see there is a large degree of similarityС ќМPD A between the two bitmaps and across the rows, which lends well toС QЂW Q ќQN K This crate stores one 512-bit "row" of the above bitmaps in the leaf levelС ќхQO L of a trie, and a single additional level to index into the leafs. There areС ќЕR5 2 134 unique 512-bit chunks across the two bitmaps.С ыR ќяRO L The chunk size of 512 bits is selected as the size that minimizes the totalС ќПSL I size of the data structure. A smaller chunk, like 256 or 128 bits, wouldС ќTK H achieve better deduplication but require a larger index. A larger chunkС ќиTI F would increase redundancy in the leaf bitmaps. 512 bit chunks are theС ќЂU: 7 optimum for total size of the index plus leaf bitmaps.С нU ќсUQ N The chunk data is compressed using the KuhnтMunkres algorithm for bipartiteС ќГVO L matching to eliminate redundancies between the second half of any chunk andС ќWE B the first half of any other chunk. This achieves an additional 9%С ќЩWP M compression of the leaf level, leaving 122 chunks that can be indexed at theС ќXL I half-chunk level using an 8-bit index. Note that this is not the same asС ќчXI F using chunks which are half the size, because it does not necessitateС ќБY/ , raising the size of the trie's first level.С сY ќхYO L In contrast to binary search or the `ucd-trie` crate, performing lookups inС ќЕZI F this data structure is straight-line code with no need for branching.С ќ[? $https://docs.rs/unicode-ident/1.0.26Сl[
ћO З_ пp Э є вЉ " КО Ѕг
ш ћ OOOєи\_lќ\ppЉЉООггшшћћќД^% ќѕ]> ; Whether the character has the Unicode property XID\_Start.СdЛ^ Bџ
chСШ^ќb( ќХaA > Whether the character has the Unicode property XID\_Continue.С|b Bџ
иxbTУ\у
skipСlД\ В\ 4Ч\ "N 4 ўЦн iKц;Ѓ ` нBџNR нBџ
н`b d нЮ <Ъ Д{ в ў Џ 6GК6ym нўBџЮI нўBџ
н`b dе нќх' |я ќ" \ фќф% tѕ фьА ,С дќ
" T дќЎ% lП дќл/ Tэ ј ѓ iKц;Ѓ `ѕд џ
!%+/26:>ЎC G K ЊЊЊЊЊЊЊЊЊЊЊЊыЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊORVZЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊP ъ^bdhlp0tvz~ЊЋЂЄ лЊЊЉЊЊЊЊЊЊAЊА з ЊГЧЗЊЊЊЊЊЊЊЊЊЊЊЊЊЊЛЊ) Ф&Н Р ЩЫ Бвжнсх щ ЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊэЊЊЊЊЊЊЊЊяЊЊЊЊЊЊЊЊЊЊёЊЊЊЊЊЊЊЊЊЊЊЊЊЊѓ Њи ЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊѕ ЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊї џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ
дќ 3 lЇ ј ѓ iKц;Ѓ `ѕд #(-/48<@ЎE I M ЊЊЊЊЊЊЊЊЊЊЊЊыЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊOTX\ЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊP ъ`bfjnr0tx| ЊЋЂЄ лЊЊЊЊЊЊЊЊAЊА а ЊЕЇЙЊЊЊЊЊЊЊЊЊЊЊЊЊЊЛЊ) Ф&Н У ІСОЦЩЭ Я Бдкпуч щ Ѕ ЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊэЊЊЊЊЊЊЊЊяЊЊЊЊЊЊЊЊЊЊёЊЊЊЊЊЊЊЊЊЊЊЊЊЊѓ Њи ЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊѕ ЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊЊї џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ$Т дќЃw+ $Еw ѓ Џ 6GК6ymѕд Р> џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ џџџџџџџџџџџџџџџџџџџџџџџџџџУџ P пИ@зџџћџџџџџџџџџПџџџџџџџџџџџџџџџџџџџџџџџџџУџ P џџџџџџџџџџџџџџпИРзџџћџџџџџџџџџПџџџџџџџџџџџџџџџџџќџџџџџџџџџџџџџџџџџџџџўџџџџџџџџ џџџ џџџџџџџџџџџџџџџџћќџџџџџџџџџџџџџџџџџџџџўџџџџџџџџўџџџџПЖџџџ џџџџџ Рўџџџџџџџџџџџ/ `Р §џџџ рџџџџџџџџџџ? ќџџџ0 џџџџџџџџџџУџџџџџџџџџџџџяџ§џ џџџџџџџчџџџџџџџџџџџџ џџџџџџ?$џџ? џџџџџџџў џџџџџ №џџџџџџ# џ ўџсљџџ§Х# @ А џџџџџ? џџџџџџџўџџџџџџџџџћџџџџџџџџџџџџџџџЯџўџяљџџ§ХѓyАЯџPрљџџ§m ^ рПћџџ§э# рљџџ§э# А шЧ=жЧџ юљџџ§mг9^Рџ? юПћџџ§эѓП; Яџ ўюљџџ§эѓ9јАЯџ ьЧ=жЧџУЧ= Рџ рп§џџ§џ# 7 сп§џџ§я# p №п§џџџџ' @p ќрџќџџћ/ џп§џџ§џѓп=`7Яџ яп§џџ§яѓп=`pЯџ џп§џџџџџп}№Яџ ќюџќџџћ/_џРџ ўџџџџџ жїџџЏџ _ № џўџџџ ўџџџџџџџџ жїџџЏџџ?_џѓ џ Тџўџџџ ўџпџџўџџџ @ џџџџџ ? ўџџџџџџџџџрўџџџџџџџџџџїрџџџџџўџџџџџџџџџџ џџџџ џџр ўџ> ўџџџџџџџџџцўџџџџџџџџџџџрџџџџџўџџџџџџџџџџ џџџџ џџџџџџџџџџџџџџџџџџџ џџџџџ?џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ џџјџџџџџџ џџ џџџџџ џџџ?џџџџџџџџџџ џќџџџџџџџџџџџџљџџџџџџџџџ? ўџџ џџџ џџџџџџ№Пџџџџџџџџџџџџџџ џќџџџџџџџџџџџџљџџџџџџџџџ? ўџЛїџџ џџџџџџ ќџџџџџ ќh ќџџ? џџ џџџ №џџџџџ пџ |џџџџџ џџџџџџ џџџџџџџџ? џџџџшџџџџџ?џџџџ џџџ џџџџџџџџџџџџџџџџџ ї џџФџџџџџџb> 8џ ~~~ џџџџџїџ3џџџџџџџџџџџџџџ џџџџџџ џ?џџџќџџџџџџџџ 8џџ| ~~~ џџџџџїџ3џџџџџџџџџџџџџџџ7џџџџџџџџџџџџџџ?џџџџџџџџџџџџџ ј џ§_лџџџџџџџџџџџџџ јџџџџџџџџџџџџџџџџџџ?џџџџџџџџџџџџџ јрџ§_лџџџџџџџџџџџџџ јџџџџџџџџџџџџџџџџ?№џџџџџџџџџџџџџџџџџџџџџџџџџџ? џџџџџџџџќџџџџџџ џ Њџџџџџџџџџџџџџџџ ўџџўџџРџџџџџџ?џџџќќќ џџ џџ р Њџџџџџџџџџџџџџџџ џўџџўџџрџџџџџџџџџџќќќ џяџџџџЗџ?џ? џџџџџџџџџџџџџџџ џџџџџџ џяџџџџЗџ?џ? џџџџџџџџџџџџџџџ џџџџџџ џџџ џџџџџџ џџџџ рџџџџџџџ? џџџ?џџџџџ> џџџ џџџџџџ џџџџ рџџџџџџџџџџџ?џџџџџ> џџџџџџџџџџџџџџџџџџџ? џџџџџџџџџџџџџ џџџџџџ џїџїЗџћџћџџџџџџ џџџџџџџџџџџџџџџџџџџ?џџџџџџџџџџџџџџ џџџџџџ џїџїЗџћџћџџџџџџ ?§џџџџПџџ? џџ џџџ џџ7 џџ? џџџџџџ џџџџџџџР яўџџ? џџџ џџџ џўџџ џџџџџџ? џџ? џџ џџ o№яўџџ? џџџ џџџ џўџџ џџџџџџ? џџ? џџ џџ џџџџџџџџџ џџџџџџ џџџџџџ џџџџ ќџџ?џџ? џџџџџџџџџ џџџџџџ џџџџџџ џџџџџ џџџџџ?Оџџ? џџџџџ ќ ўџ џџџ џџ? џџ џџ џџ џџџџџ ќј ўџџџџџџ џџџџ џџ? џџ џџ јџџџџџџ & јџџџџџ џџџ јџџџ џџџџG јџџџџџ џџџџџџџџ Рџ?џџџџџџџ џџџџџџџџџџпџ№ џџџџO џџџџџџџџ оџ џџћџџ НџПџџџџџџ рљџџ§э# р џKџџџџП
џџћџџџџР НџПџџџџџџџџџяљџџ§эћ9рЯ џKџџџџПџЅї џџџџџџ џџџџџџ А џџџџџ џџџџџџџџџџУ џџџџџџџџП џ џџџџџџ?џ ? џџџџџџ џџџџџ џџџ џџџџџџџџ џ џџџџџџџџџџ џџџчџџ џџџџџ џџџџџџџџ ђoџџџ џќџџџџ
џџџџџџџ џџџџџџџџџ ђoџџџПљ џ џќџџџџџќ јџџџџ №џџџџџ џџџџџџџџџ џџџџ џџџџџџџ џџџџџџџџџ# џџџџџџџџџ џ џџџџ џџ§џџџ ќџџџ ћџџџџ @ П§џџџ џџџџџ џ§џџџџџ џ ќџџџќџџў ћџџџџДџ џП§џџџћџџџџџџџ џџ єџ§џџџ џџ џџ§џџџџЧ џ џџџџџџџџџџџџџџрџџџџџџџџџџџџџџџџџџџџџџџџџ џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ џ џџџџџ џ џџџџџџ ~ џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ џџџџџџ џџ? џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ??џџџџ??џЊџџџ?џџџџџџп_м Яџ м џџџџџџџџ џџџџџџџџџџџџџџџџ рџџџџџџџџџџџџџџџџ џџџџџџџџџ џџџ? џџџџџџ јџџрџџ џџџџџџџџџџџџџџџџџџџџџџџџ? џџџџџџ џјџџрџџ џџџџџџџџ џџџљџџ џџџџџџџџџ јџ | џџџџџџџџ џџџљџџ џџџџџџџџџџџџџџџџџџ џџџџџџџџџџџџџџџџџџџџџџџџџџџ џџџџ џџџџџџџџџџџџџџ џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ рујч < џџџџџџџџџџџџџџ џџ џџџџџ?џџ џџџџџџџџџџџџџџ џџc яo џџџџџ џџџџџџџџџџпџџџџџџџџпdоџыяџџџџџџџПчппџџџ{_ќ§џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ§џџїџџџїџџпџџџпџџџџџџџџ§џџџ§џџї џџџџџџџџџџџџџџџџџџџџџџџ§џџїџџџїџџпџџџпџџџџџџџџ§џџџ§џџїЯџџџџџџџџџџџџјџџџџџ јўџ џџџџџџџ џџџџџџџ? џџџџџ ? @ џџљлџџџџџџџ? џџџџџ џ?џC џџџ? џџџџџ џџџ? џџџ џџџџџџџ џџџџџџџџџџџџ џџџ џџџ? џџџџџ џџџџџ џџџЗ? Р oџ џџџџџ?Р oџџџџџџџџџџџџџџџџџџџџџџџџџ џџџџџџџџ џџџџџџџџџџџџџџџџџџџџџџџџ џџџџџџџџџџ яџџџўї
ъЊїї^џћџюћџ џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ џџџџџџџџџџџџџџџџџџџџџџџџџџџџ џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ?џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ? џџџџџџџџџџџџџџџ џџџџџџџџ ~џџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ џџџџџџџџР>$Шw дK unicode-xidС )https://github.com/unicode-rs/unicode-xidС ucd-generateС !https://crates.io/crates/croaringС ]https://img.shields.io/badge/crates.io-fc8d62?style=for-the-badge&labelColor=555555&logo=rustС (https://github.com/dtolnay/unicode-identС crates-ioС &https://crates.io/crates/unicode-identС ^https://img.shields.io/badge/docs.rs-66c2a5?style=for-the-badge&labelColor=555555&logo=docs.rsС rust-lang/rust#33098С ,https://github.com/rust-lang/rust/pull/33098С finite state transducerС %https://www.unicode.org/reports/tr31/С Ул гл croaringС githubС н Чн м Им +https://github.com/RoaringBitmap/roaring-rsС уо ќо Unicode Standard Annex #31С Щп Ул https://roaringbitmap.org/about/С џп м р н roaringС Џр уо https://docs.rs/unicode-identС ыр Щп Roaring BitmapС с џп р *https://github.com/BurntSushi/ucd-generateС гс Џр \https://img.shields.io/badge/github-8da0cb?style=for-the-badge&labelColor=555555&logo=githubС docs-rsС ыс ыр !https://github.com/BurntSushi/fstС т с м От гс о љт ен ку ыс п ђу т гл м От Им о Чн ен ку љт ќо п ђу `Иgх9иv_П I&иђЭnNб"iFтОЮЕцpYу!$Ѓй ЂЅћЪЦ=їћ_sJ*b ъ%<ЕббЎlаяЙvн@мНaѕќЂ:в!
иВ1RPT г)2њЦЁаы$IМѓpФђйP9G ВјLхиГAВ fZ
ншкчВ J$Гъh4RъС;ъЙyHі4н?+ЏюНщQіWмѕ~д ѓnDЎяѕ=)Эвфя@ћlWіёЛ)YКв;kд,*7ьAТCмрє{ЬшЅ.Х_O.
aТшЃ #Pіх1 чv
k ає у
: ѓ Ѓ ;Ї ;Ў ;Б ;З ;Н ;У ;Щ ;Я ;е ;л ;у;e<ч< = А =Х = W >n > > >Ћ >Р >е >ъ > Ў A Jр:
= ]= > ' ' п: C<Ш<=== В=Ы=&> Y>r>>>Џ>Ф>й>? ЦA 2J ; ;Ѓ;Њ;Ў;Д;К;Р;Ц;Ь;в;и;о;`<у<=f==Ќ=Т=
>F>S>i>}>>Ї>Л>а>х>AЉAJJВm ; Ї; Б;З;Н;У;Щ;Я;е;л;+<А<џ<=i=Ё=А=Ч=>H>W>n>>>Ћ>Р>е>ъ> ЎA J P<е< 1==Љ=М=с=9>P>c>y>>Ѓ>З>Ь>с>
?ЄAЮAJ:JМm J<Я< ==Ѓ=Д=Ю=1>J>[>s>>>Б>Ц>л>?AШA
J4JЖm R<з< C== Р=ъ=?> g>{>>Ѕ>Й>Ю>у> ?ІAаAJ D<Щ< ^== >)> /<Д< k= > ю> ВA J Ћ= R> L Ь н QЩоѓ 5 Ы H P = => IfЂжы + B MjІкя / F V<л< J? B zJн: §<= Х= S= њ= E= ь= Ђ І Њ Ў В Ж К О б ч є !%,3I[p|ЁВЖЪЮй V= §=Сm ЈA J Рm
йwѕд еwѕд љW ѕ ћnѕ љnѕ " %м `Иgх9ЋXСђWЂ
rustfmt::skipС љ< [ `Иgх9#Ћ9vЭк e{щzэz{3{ўz#{U{№ODHT $ @ Зо ,*7ьAТC Z
нш кчВ J$Г ћЪЦ=їћ_ nNб" G ВјLх њЦЁаы$ %<Ебб ѕ=)Эвфя@ ќЂ:в!
Ѕ.Х_O.
a k ає # $Ѓй ЂЅ щQіWмѕ~ @мНaѕ иГAВ f иv_П pYу! sJ*b ъ ТшЃ #P! `Иgх9 іх1 чv
" T г)2 4н?+ЏюН YКв;kд д ѓnDЎя I&иђЭ иВ1RP
IМѓpФ ъh4R ћlWіёЛ) ЎlаяЙvн
ђйP9 ъС;ъЙyHі мрє{Ьш iFтОЮЕц џ!џџџџџџџџtџY/r
h 0џџџRџџ?zџ3GEu(џ@џ
џџ^KџwџџџC()џџn{Cџsџџ!џџџџџџџџtџY/r ]/root/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/unicode-ident-1.0.26/src/tables.rsСO/root/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/unicode-ident-1.0.26С]/root/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/unicode-ident-1.0.26/src/tables.rsС чЖКщ
Ў|Ое5ГsZ чќчќЁQ@3+,6HK$0),;ddddddddddddddddddddddddddddddd^?dddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddd
8dddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddd яШЯ<Й3)u8$шНIХ Z/root/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/unicode-ident-1.0.26/src/lib.rsСO/root/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/unicode-ident-1.0.26СZ/root/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/unicode-ident-1.0.26/src/lib.rsС !ђxКUќьИ
Р§ ІeІeЇkon OA2OMQ !K%JD >Q>KHN?A=B MKHBGO; %PC #7&& $ MMMJLNMM;OGO2N#OPMNQLOQ4LA I **&*%&QLONPJQ3OPGON,PQ'QMQONP
MKN#-?NNLO'IE7QQJJJPQNQ#4PP( QFFPP9 5 ЁЄ
QNEOP6PMLJ;RPFQMJ0PJ @
(d?(5 &/GQB+8 &5GQЦЁ VЄJеЇЕВpЉ8T ы~м x86_64-unknown-linux-gnuСЄђYћГЃКќYЙЛО
unicode_identС -de61ca3b509bd8dfС`Иgх9 R Іф $ # i0$?#$$F#H F $ $H$HHB#, F . $2 H 2 ,$ $H H - # F * *$* H * $ H Ха rust-end-file