| 1 |
// Copyright Mozilla Foundation. See the COPYRIGHT |
| 2 |
// file at the top-level directory of this distribution. |
| 3 |
// |
| 4 |
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or |
| 5 |
// https://www.apache.org/licenses/LICENSE-2.0> or the MIT license |
| 6 |
// <LICENSE-MIT or https://opensource.org/licenses/MIT>, at your |
| 7 |
// option. This file may not be copied, modified, or distributed |
| 8 |
// except according to those terms. |
| 9 |
|
| 10 |
#![cfg_attr( |
| 11 |
feature = "cargo-clippy", |
| 12 |
allow(doc_markdown, inline_always, new_ret_no_self) |
| 13 |
)] |
| 14 |
|
| 15 |
//! encoding_rs is a Gecko-oriented Free Software / Open Source implementation |
| 16 |
//! of the [Encoding Standard](https://encoding.spec.whatwg.org/) in Rust. |
| 17 |
//! Gecko-oriented means that converting to and from UTF-16 is supported in |
| 18 |
//! addition to converting to and from UTF-8, that the performance and |
| 19 |
//! streamability goals are browser-oriented, and that FFI-friendliness is a |
| 20 |
//! goal. |
| 21 |
//! |
| 22 |
//! Additionally, the `mem` module provides functions that are useful for |
| 23 |
//! applications that need to be able to deal with legacy in-memory |
| 24 |
//! representations of Unicode. |
| 25 |
//! |
| 26 |
//! For expectation setting, please be sure to read the sections |
| 27 |
//! [_UTF-16LE, UTF-16BE and Unicode Encoding Schemes_](#utf-16le-utf-16be-and-unicode-encoding-schemes), |
| 28 |
//! [_ISO-8859-1_](#iso-8859-1) and [_Web / Browser Focus_](#web--browser-focus) below. |
| 29 |
//! |
| 30 |
//! There is a [long-form write-up](https://hsivonen.fi/encoding_rs/) about the |
| 31 |
//! design and internals of the crate. |
| 32 |
//! |
| 33 |
//! # Availability |
| 34 |
//! |
| 35 |
//! The code is available under the |
| 36 |
//! [Apache license, Version 2.0](https://www.apache.org/licenses/LICENSE-2.0) |
| 37 |
//! or the [MIT license](https://opensource.org/licenses/MIT), at your option. |
| 38 |
//! See the |
| 39 |
//! [`COPYRIGHT`](https://github.com/hsivonen/encoding_rs/blob/master/COPYRIGHT) |
| 40 |
//! file for details. |
| 41 |
//! The [repository is on GitHub](https://github.com/hsivonen/encoding_rs). The |
| 42 |
//! [crate is available on crates.io](https://crates.io/crates/encoding_rs). |
| 43 |
//! |
| 44 |
//! # Integration with `std::io` |
| 45 |
//! |
| 46 |
//! This crate doesn't implement traits from `std::io`. However, for the case of |
| 47 |
//! wrapping a `std::io::Read` in a decoder that implements `std::io::Read` and |
| 48 |
//! presents the data from the wrapped `std::io::Read` as UTF-8 is addressed by |
| 49 |
//! the [`encoding_rs_io`](https://docs.rs/encoding_rs_io/) crate. |
| 50 |
//! |
| 51 |
//! # Examples |
| 52 |
//! |
| 53 |
//! Example programs: |
| 54 |
//! |
| 55 |
//! * [Rust](https://github.com/hsivonen/recode_rs) |
| 56 |
//! * [C](https://github.com/hsivonen/recode_c) |
| 57 |
//! * [C++](https://github.com/hsivonen/recode_cpp) |
| 58 |
//! |
| 59 |
//! Decode using the non-streaming API: |
| 60 |
//! |
| 61 |
//! ``` |
| 62 |
//! #[cfg(feature = "alloc")] { |
| 63 |
//! use encoding_rs::*; |
| 64 |
//! |
| 65 |
//! let expectation = "\u{30CF}\u{30ED}\u{30FC}\u{30FB}\u{30EF}\u{30FC}\u{30EB}\u{30C9}"; |
| 66 |
//! let bytes = b"\x83n\x83\x8D\x81[\x81E\x83\x8F\x81[\x83\x8B\x83h"; |
| 67 |
//! |
| 68 |
//! let (cow, encoding_used, had_errors) = SHIFT_JIS.decode(bytes); |
| 69 |
//! assert_eq!(&cow[..], expectation); |
| 70 |
//! assert_eq!(encoding_used, SHIFT_JIS); |
| 71 |
//! assert!(!had_errors); |
| 72 |
//! } |
| 73 |
//! ``` |
| 74 |
//! |
| 75 |
//! Decode using the streaming API with minimal `unsafe`: |
| 76 |
//! |
| 77 |
//! ``` |
| 78 |
//! use encoding_rs::*; |
| 79 |
//! |
| 80 |
//! let expectation = "\u{30CF}\u{30ED}\u{30FC}\u{30FB}\u{30EF}\u{30FC}\u{30EB}\u{30C9}"; |
| 81 |
//! |
| 82 |
//! // Use an array of byte slices to demonstrate content arriving piece by |
| 83 |
//! // piece from the network. |
| 84 |
//! let bytes: [&'static [u8]; 4] = [b"\x83", |
| 85 |
//! b"n\x83\x8D\x81", |
| 86 |
//! b"[\x81E\x83\x8F\x81[\x83", |
| 87 |
//! b"\x8B\x83h"]; |
| 88 |
//! |
| 89 |
//! // Very short output buffer to demonstrate the output buffer getting full. |
| 90 |
//! // Normally, you'd use something like `[0u8; 2048]`. |
| 91 |
//! let mut buffer_bytes = [0u8; 8]; |
| 92 |
//! let mut buffer: &mut str = std::str::from_utf8_mut(&mut buffer_bytes[..]).unwrap(); |
| 93 |
//! |
| 94 |
//! // How many bytes in the buffer currently hold significant data. |
| 95 |
//! let mut bytes_in_buffer = 0usize; |
| 96 |
//! |
| 97 |
//! // Collect the output to a string for demonstration purposes. |
| 98 |
//! let mut output = String::new(); |
| 99 |
//! |
| 100 |
//! // The `Decoder` |
| 101 |
//! let mut decoder = SHIFT_JIS.new_decoder(); |
| 102 |
//! |
| 103 |
//! // Track whether we see errors. |
| 104 |
//! let mut total_had_errors = false; |
| 105 |
//! |
| 106 |
//! // Decode using a fixed-size intermediate buffer (for demonstrating the |
| 107 |
//! // use of a fixed-size buffer; normally when the output of an incremental |
| 108 |
//! // decode goes to a `String` one would use `Decoder.decode_to_string()` to |
| 109 |
//! // avoid the intermediate buffer). |
| 110 |
//! for input in &bytes[..] { |
| 111 |
//! // The number of bytes already read from current `input` in total. |
| 112 |
//! let mut total_read_from_current_input = 0usize; |
| 113 |
//! |
| 114 |
//! loop { |
| 115 |
//! let (result, read, written, had_errors) = |
| 116 |
//! decoder.decode_to_str(&input[total_read_from_current_input..], |
| 117 |
//! &mut buffer[bytes_in_buffer..], |
| 118 |
//! false); |
| 119 |
//! total_read_from_current_input += read; |
| 120 |
//! bytes_in_buffer += written; |
| 121 |
//! total_had_errors |= had_errors; |
| 122 |
//! match result { |
| 123 |
//! CoderResult::InputEmpty => { |
| 124 |
//! // We have consumed the current input buffer. Break out of |
| 125 |
//! // the inner loop to get the next input buffer from the |
| 126 |
//! // outer loop. |
| 127 |
//! break; |
| 128 |
//! }, |
| 129 |
//! CoderResult::OutputFull => { |
| 130 |
//! // Write the current buffer out and consider the buffer |
| 131 |
//! // empty. |
| 132 |
//! output.push_str(&buffer[..bytes_in_buffer]); |
| 133 |
//! bytes_in_buffer = 0usize; |
| 134 |
//! continue; |
| 135 |
//! } |
| 136 |
//! } |
| 137 |
//! } |
| 138 |
//! } |
| 139 |
//! |
| 140 |
//! // Process EOF |
| 141 |
//! loop { |
| 142 |
//! let (result, _, written, had_errors) = |
| 143 |
//! decoder.decode_to_str(b"", |
| 144 |
//! &mut buffer[bytes_in_buffer..], |
| 145 |
//! true); |
| 146 |
//! bytes_in_buffer += written; |
| 147 |
//! total_had_errors |= had_errors; |
| 148 |
//! // Write the current buffer out and consider the buffer empty. |
| 149 |
//! // Need to do this here for both `match` arms, because we exit the |
| 150 |
//! // loop on `CoderResult::InputEmpty`. |
| 151 |
//! output.push_str(&buffer[..bytes_in_buffer]); |
| 152 |
//! bytes_in_buffer = 0usize; |
| 153 |
//! match result { |
| 154 |
//! CoderResult::InputEmpty => { |
| 155 |
//! // Done! |
| 156 |
//! break; |
| 157 |
//! }, |
| 158 |
//! CoderResult::OutputFull => { |
| 159 |
//! continue; |
| 160 |
//! } |
| 161 |
//! } |
| 162 |
//! } |
| 163 |
//! |
| 164 |
//! assert_eq!(&output[..], expectation); |
| 165 |
//! assert!(!total_had_errors); |
| 166 |
//! ``` |
| 167 |
//! |
| 168 |
//! ## UTF-16LE, UTF-16BE and Unicode Encoding Schemes |
| 169 |
//! |
| 170 |
//! The Encoding Standard doesn't specify encoders for UTF-16LE and UTF-16BE, |
| 171 |
//! __so this crate does not provide encoders for those encodings__! |
| 172 |
//! Along with the replacement encoding, their _output encoding_ (i.e. the |
| 173 |
//! encoding used for form submission and error handling in the query string |
| 174 |
//! of URLs) is UTF-8, so you get an UTF-8 encoder if you request an encoder |
| 175 |
//! for them. |
| 176 |
//! |
| 177 |
//! Additionally, the Encoding Standard factors BOM handling into wrapper |
| 178 |
//! algorithms so that BOM handling isn't part of the definition of the |
| 179 |
//! encodings themselves. The Unicode _encoding schemes_ in the Unicode |
| 180 |
//! Standard define BOM handling or lack thereof as part of the encoding |
| 181 |
//! scheme. |
| 182 |
//! |
| 183 |
//! When used with the `_without_bom_handling` entry points, the UTF-16LE |
| 184 |
//! and UTF-16BE _encodings_ match the same-named _encoding schemes_ from |
| 185 |
//! the Unicode Standard. |
| 186 |
//! |
| 187 |
//! When used with the `_with_bom_removal` entry points, the UTF-8 |
| 188 |
//! _encoding_ matches the UTF-8 _encoding scheme_ from the Unicode |
| 189 |
//! Standard. |
| 190 |
//! |
| 191 |
//! This crate does not provide a mode that matches the UTF-16 _encoding |
| 192 |
//! scheme_ from the Unicode Stardard. The UTF-16BE encoding used with |
| 193 |
//! the entry points without `_bom_` qualifiers is the closest match, |
| 194 |
//! but in that case, the UTF-8 BOM triggers UTF-8 decoding, which is |
| 195 |
//! not part of the behavior of the UTF-16 _encoding scheme_ per the |
| 196 |
//! Unicode Standard. |
| 197 |
//! |
| 198 |
//! The UTF-32 family of Unicode encoding schemes is not supported |
| 199 |
//! by this crate. The Encoding Standard doesn't define any UTF-32 |
| 200 |
//! family encodings, since they aren't necessary for consuming Web |
| 201 |
//! content. |
| 202 |
//! |
| 203 |
//! While gb18030 is capable of representing U+FEFF, the Encoding |
| 204 |
//! Standard does not treat the gb18030 byte representation of U+FEFF |
| 205 |
//! as a BOM, so neither does this crate. |
| 206 |
//! |
| 207 |
//! ## ISO-8859-1 |
| 208 |
//! |
| 209 |
//! ISO-8859-1 does not exist as a distinct encoding from windows-1252 in |
| 210 |
//! the Encoding Standard. Therefore, an encoding that maps the unsigned |
| 211 |
//! byte value to the same Unicode scalar value is not available via |
| 212 |
//! `Encoding` in this crate. |
| 213 |
//! |
| 214 |
//! However, the functions whose name starts with `convert` and contains |
| 215 |
//! `latin1` in the `mem` module support such conversions, which are known as |
| 216 |
//! [_isomorphic decode_](https://infra.spec.whatwg.org/#isomorphic-decode) |
| 217 |
//! and [_isomorphic encode_](https://infra.spec.whatwg.org/#isomorphic-encode) |
| 218 |
//! in the [Infra Standard](https://infra.spec.whatwg.org/). |
| 219 |
//! |
| 220 |
//! ## Web / Browser Focus |
| 221 |
//! |
| 222 |
//! Both in terms of scope and performance, the focus is on the Web. For scope, |
| 223 |
//! this means that encoding_rs implements the Encoding Standard fully and |
| 224 |
//! doesn't implement encodings that are not specified in the Encoding |
| 225 |
//! Standard. For performance, this means that decoding performance is |
| 226 |
//! important as well as performance for encoding into UTF-8 or encoding the |
| 227 |
//! Basic Latin range (ASCII) into legacy encodings. Non-Basic Latin needs to |
| 228 |
//! be encoded into legacy encodings in only two places in the Web platform: in |
| 229 |
//! the query part of URLs, in which case it's a matter of relatively rare |
| 230 |
//! error handling, and in form submission, in which case the user action and |
| 231 |
//! networking tend to hide the performance of the encoder. |
| 232 |
//! |
| 233 |
//! Deemphasizing performance of encoding non-Basic Latin text into legacy |
| 234 |
//! encodings enables smaller code size thanks to the encoder side using the |
| 235 |
//! decode-optimized data tables without having encode-optimized data tables at |
| 236 |
//! all. Even in decoders, smaller lookup table size is preferred over avoiding |
| 237 |
//! multiplication operations. |
| 238 |
//! |
| 239 |
//! Additionally, performance is a non-goal for the ASCII-incompatible |
| 240 |
//! ISO-2022-JP encoding, which are rarely used on the Web. Instead of |
| 241 |
//! performance, the decoder for ISO-2022-JP optimizes for ease/clarity |
| 242 |
//! of implementation. |
| 243 |
//! |
| 244 |
//! Despite the browser focus, the hope is that non-browser applications |
| 245 |
//! that wish to consume Web content or submit Web forms in a Web-compatible |
| 246 |
//! way will find encoding_rs useful. While encoding_rs does not try to match |
| 247 |
//! Windows behavior, many of the encodings are close enough to legacy |
| 248 |
//! encodings implemented by Windows that applications that need to consume |
| 249 |
//! data in legacy Windows encodins may find encoding_rs useful. The |
| 250 |
//! [codepage](https://crates.io/crates/codepage) crate maps from Windows |
| 251 |
//! code page identifiers onto encoding_rs `Encoding`s and vice versa. |
| 252 |
//! |
| 253 |
//! For decoding email, UTF-7 support is needed (unfortunately) in additition |
| 254 |
//! to the encodings defined in the Encoding Standard. The |
| 255 |
//! [charset](https://crates.io/crates/charset) wraps encoding_rs and adds |
| 256 |
//! UTF-7 decoding for email purposes. |
| 257 |
//! |
| 258 |
//! For single-byte DOS encodings beyond the ones supported by the Encoding |
| 259 |
//! Standard, there is the [`oem_cp`](https://crates.io/crates/oem_cp) crate. |
| 260 |
//! |
| 261 |
//! # Preparing Text for the Encoders |
| 262 |
//! |
| 263 |
//! Normalizing text into Unicode Normalization Form C prior to encoding text |
| 264 |
//! into a legacy encoding minimizes unmappable characters. Text can be |
| 265 |
//! normalized to Unicode Normalization Form C using the |
| 266 |
//! [`icu_normalizer`](https://crates.io/crates/icu_normalizer) crate, which |
| 267 |
//! is part of [ICU4X](https://icu4x.unicode.org/). |
| 268 |
//! |
| 269 |
//! The exception is windows-1258, which after normalizing to Unicode |
| 270 |
//! Normalization Form C requires tone marks to be decomposed in order to |
| 271 |
//! minimize unmappable characters. Vietnamese tone marks can be decomposed |
| 272 |
//! using the [`detone`](https://crates.io/crates/detone) crate. |
| 273 |
//! |
| 274 |
//! # Streaming & Non-Streaming; Rust & C/C++ |
| 275 |
//! |
| 276 |
//! The API in Rust has two modes of operation: streaming and non-streaming. |
| 277 |
//! The streaming API is the foundation of the implementation and should be |
| 278 |
//! used when processing data that arrives piecemeal from an i/o stream. The |
| 279 |
//! streaming API has an FFI wrapper (as a [separate crate][1]) that exposes it |
| 280 |
//! to C callers. The non-streaming part of the API is for Rust callers only and |
| 281 |
//! is smart about borrowing instead of copying when possible. When |
| 282 |
//! streamability is not needed, the non-streaming API should be preferrer in |
| 283 |
//! order to avoid copying data when a borrow suffices. |
| 284 |
//! |
| 285 |
//! There is no analogous C API exposed via FFI, mainly because C doesn't have |
| 286 |
//! standard types for growable byte buffers and Unicode strings that know |
| 287 |
//! their length. |
| 288 |
//! |
| 289 |
//! The C API (header file generated at `target/include/encoding_rs.h` when |
| 290 |
//! building encoding_rs) can, in turn, be wrapped for use from C++. Such a |
| 291 |
//! C++ wrapper can re-create the non-streaming API in C++ for C++ callers. |
| 292 |
//! The C binding comes with a [C++17 wrapper][2] that uses standard library + |
| 293 |
//! [GSL][3] types and that recreates the non-streaming API in C++ on top of |
| 294 |
//! the streaming API. A C++ wrapper with XPCOM/MFBT types is available as |
| 295 |
//! [`mozilla::Encoding`][4]. |
| 296 |
//! |
| 297 |
//! The `Encoding` type is common to both the streaming and non-streaming |
| 298 |
//! modes. In the streaming mode, decoding operations are performed with a |
| 299 |
//! `Decoder` and encoding operations with an `Encoder` object obtained via |
| 300 |
//! `Encoding`. In the non-streaming mode, decoding and encoding operations are |
| 301 |
//! performed using methods on `Encoding` objects themselves, so the `Decoder` |
| 302 |
//! and `Encoder` objects are not used at all. |
| 303 |
//! |
| 304 |
//! [1]: https://github.com/hsivonen/encoding_c |
| 305 |
//! [2]: https://github.com/hsivonen/encoding_c/blob/master/include/encoding_rs_cpp.h |
| 306 |
//! [3]: https://github.com/Microsoft/GSL/ |
| 307 |
//! [4]: https://searchfox.org/mozilla-central/source/intl/Encoding.h |
| 308 |
//! |
| 309 |
//! # Memory management |
| 310 |
//! |
| 311 |
//! The non-streaming mode never performs heap allocations (even the methods |
| 312 |
//! that write into a `Vec<u8>` or a `String` by taking them as arguments do |
| 313 |
//! not reallocate the backing buffer of the `Vec<u8>` or the `String`). That |
| 314 |
//! is, the non-streaming mode uses caller-allocated buffers exclusively. |
| 315 |
//! |
| 316 |
//! The methods of the streaming mode that return a `Vec<u8>` or a `String` |
| 317 |
//! perform heap allocations but only to allocate the backing buffer of the |
| 318 |
//! `Vec<u8>` or the `String`. |
| 319 |
//! |
| 320 |
//! `Encoding` is always statically allocated. `Decoder` and `Encoder` need no |
| 321 |
//! `Drop` cleanup. |
| 322 |
//! |
| 323 |
//! # Buffer reading and writing behavior |
| 324 |
//! |
| 325 |
//! Based on experience gained with the `java.nio.charset` encoding converter |
| 326 |
//! API and with the Gecko uconv encoding converter API, the buffer reading |
| 327 |
//! and writing behaviors of encoding_rs are asymmetric: input buffers are |
| 328 |
//! fully drained but output buffers are not always fully filled. |
| 329 |
//! |
| 330 |
//! When reading from an input buffer, encoding_rs always consumes all input |
| 331 |
//! up to the next error or to the end of the buffer. In particular, when |
| 332 |
//! decoding, even if the input buffer ends in the middle of a byte sequence |
| 333 |
//! for a character, the decoder consumes all input. This has the benefit that |
| 334 |
//! the caller of the API can always fill the next buffer from the start from |
| 335 |
//! whatever source the bytes come from and never has to first copy the last |
| 336 |
//! bytes of the previous buffer to the start of the next buffer. However, when |
| 337 |
//! encoding, the UTF-8 input buffers have to end at a character boundary, which |
| 338 |
//! is a requirement for the Rust `str` type anyway, and UTF-16 input buffer |
| 339 |
//! boundaries falling in the middle of a surrogate pair result in both |
| 340 |
//! suggorates being treated individually as unpaired surrogates. |
| 341 |
//! |
| 342 |
//! Additionally, decoders guarantee that they can be fed even one byte at a |
| 343 |
//! time and encoders guarantee that they can be fed even one code point at a |
| 344 |
//! time. This has the benefit of not placing restrictions on the size of |
| 345 |
//! chunks the content arrives e.g. from network. |
| 346 |
//! |
| 347 |
//! When writing into an output buffer, encoding_rs makes sure that the code |
| 348 |
//! unit sequence for a character is never split across output buffer |
| 349 |
//! boundaries. This may result in wasted space at the end of an output buffer, |
| 350 |
//! but the advantages are that the output side of both decoders and encoders |
| 351 |
//! is greatly simplified compared to designs that attempt to fill output |
| 352 |
//! buffers exactly even when that entails splitting a code unit sequence and |
| 353 |
//! when encoding_rs methods return to the caller, the output produces thus |
| 354 |
//! far is always valid taken as whole. (In the case of encoding to ISO-2022-JP, |
| 355 |
//! the output needs to be considered as a whole, because the latest output |
| 356 |
//! buffer taken alone might not be valid taken alone if the transition away |
| 357 |
//! from the ASCII state occurred in an earlier output buffer. However, since |
| 358 |
//! the ISO-2022-JP decoder doesn't treat streams that don't end in the ASCII |
| 359 |
//! state as being in error despite the encoder generating a transition to the |
| 360 |
//! ASCII state at the end, the claim about the partial output taken as a whole |
| 361 |
//! being valid is true even for ISO-2022-JP.) |
| 362 |
//! |
| 363 |
//! # Error Reporting |
| 364 |
//! |
| 365 |
//! Based on experience gained with the `java.nio.charset` encoding converter |
| 366 |
//! API and with the Gecko uconv encoding converter API, the error reporting |
| 367 |
//! behaviors of encoding_rs are asymmetric: decoder errors include offsets |
| 368 |
//! that leave it up to the caller to extract the erroneous bytes from the |
| 369 |
//! input stream if the caller wishes to do so but encoder errors provide the |
| 370 |
//! code point associated with the error without requiring the caller to |
| 371 |
//! extract it from the input on its own. |
| 372 |
//! |
| 373 |
//! On the encoder side, an error is always triggered by the most recently |
| 374 |
//! pushed Unicode scalar, which makes it simple to pass the `char` to the |
| 375 |
//! caller. Also, it's very typical for the caller to wish to do something with |
| 376 |
//! this data: generate a numeric escape for the character. Additionally, the |
| 377 |
//! ISO-2022-JP encoder reports U+FFFD instead of the actual input character in |
| 378 |
//! certain cases, so requiring the caller to extract the character from the |
| 379 |
//! input buffer would require the caller to handle ISO-2022-JP details. |
| 380 |
//! Furthermore, requiring the caller to extract the character from the input |
| 381 |
//! buffer would require the caller to implement UTF-8 or UTF-16 math, which is |
| 382 |
//! the job of an encoding conversion library. |
| 383 |
//! |
| 384 |
//! On the decoder side, errors are triggered in more complex ways. For |
| 385 |
//! example, when decoding the sequence ESC, '$', _buffer boundary_, 'A' as |
| 386 |
//! ISO-2022-JP, the ESC byte is in error, but this is discovered only after |
| 387 |
//! the buffer boundary when processing 'A'. Thus, the bytes in error might not |
| 388 |
//! be the ones most recently pushed to the decoder and the error might not even |
| 389 |
//! be in the current buffer. |
| 390 |
//! |
| 391 |
//! Some encoding conversion APIs address the problem by not acknowledging |
| 392 |
//! trailing bytes of an input buffer as consumed if it's still possible for |
| 393 |
//! future bytes to cause the trailing bytes to be in error. This way, error |
| 394 |
//! reporting can always refer to the most recently pushed buffer. This has the |
| 395 |
//! problem that the caller of the API has to copy the unconsumed trailing |
| 396 |
//! bytes to the start of the next buffer before being able to fill the rest |
| 397 |
//! of the next buffer. This is annoying, error-prone and inefficient. |
| 398 |
//! |
| 399 |
//! A possible solution would be making the decoder remember recently consumed |
| 400 |
//! bytes in order to be able to include a copy of the erroneous bytes when |
| 401 |
//! reporting an error. This has two problem: First, callers a rarely |
| 402 |
//! interested in the erroneous bytes, so attempts to identify them are most |
| 403 |
//! often just overhead anyway. Second, the rare applications that are |
| 404 |
//! interested typically care about the location of the error in the input |
| 405 |
//! stream. |
| 406 |
//! |
| 407 |
//! To keep the API convenient for common uses and the overhead low while making |
| 408 |
//! it possible to develop applications, such as HTML validators, that care |
| 409 |
//! about which bytes were in error, encoding_rs reports the length of the |
| 410 |
//! erroneous sequence and the number of bytes consumed after the erroneous |
| 411 |
//! sequence. As long as the caller doesn't discard the 6 most recent bytes, |
| 412 |
//! this makes it possible for callers that care about the erroneous bytes to |
| 413 |
//! locate them. |
| 414 |
//! |
| 415 |
//! # No Convenience API for Custom Replacements |
| 416 |
//! |
| 417 |
//! The Web Platform and, therefore, the Encoding Standard supports only one |
| 418 |
//! error recovery mode for decoders and only one error recovery mode for |
| 419 |
//! encoders. The supported error recovery mode for decoders is emitting the |
| 420 |
//! REPLACEMENT CHARACTER on error. The supported error recovery mode for |
| 421 |
//! encoders is emitting an HTML decimal numeric character reference for |
| 422 |
//! unmappable characters. |
| 423 |
//! |
| 424 |
//! Since encoding_rs is Web-focused, these are the only error recovery modes |
| 425 |
//! for which convenient support is provided. Moreover, on the decoder side, |
| 426 |
//! there aren't really good alternatives for emitting the REPLACEMENT CHARACTER |
| 427 |
//! on error (other than treating errors as fatal). In particular, simply |
| 428 |
//! ignoring errors is a |
| 429 |
//! [security problem](http://www.unicode.org/reports/tr36/#Substituting_for_Ill_Formed_Subsequences), |
| 430 |
//! so it would be a bad idea for encoding_rs to provide a mode that encouraged |
| 431 |
//! callers to ignore errors. |
| 432 |
//! |
| 433 |
//! On the encoder side, there are plausible alternatives for HTML decimal |
| 434 |
//! numeric character references. For example, when outputting CSS, CSS-style |
| 435 |
//! escapes would seem to make sense. However, instead of facilitating the |
| 436 |
//! output of CSS, JS, etc. in non-UTF-8 encodings, encoding_rs takes the design |
| 437 |
//! position that you shouldn't generate output in encodings other than UTF-8, |
| 438 |
//! except where backward compatibility with interacting with the legacy Web |
| 439 |
//! requires it. The legacy Web requires it only when parsing the query strings |
| 440 |
//! of URLs and when submitting forms, and those two both use HTML decimal |
| 441 |
//! numeric character references. |
| 442 |
//! |
| 443 |
//! While encoding_rs doesn't make encoder replacements other than HTML decimal |
| 444 |
//! numeric character references easy, it does make them _possible_. |
| 445 |
//! `encode_from_utf8()`, which emits HTML decimal numeric character references |
| 446 |
//! for unmappable characters, is implemented on top of |
| 447 |
//! `encode_from_utf8_without_replacement()`. Applications that really, really |
| 448 |
//! want other replacement schemes for unmappable characters can likewise |
| 449 |
//! implement them on top of `encode_from_utf8_without_replacement()`. |
| 450 |
//! |
| 451 |
//! # No Extensibility by Design |
| 452 |
//! |
| 453 |
//! The set of encodings supported by encoding_rs is not extensible by design. |
| 454 |
//! That is, `Encoding`, `Decoder` and `Encoder` are intentionally `struct`s |
| 455 |
//! rather than `trait`s. encoding_rs takes the design position that all future |
| 456 |
//! text interchange should be done using UTF-8, which can represent all of |
| 457 |
//! Unicode. (It is, in fact, the only encoding supported by the Encoding |
| 458 |
//! Standard and encoding_rs that can represent all of Unicode and that has |
| 459 |
//! encoder support. UTF-16LE and UTF-16BE don't have encoder support, and |
| 460 |
//! gb18030 cannot encode U+E5E5.) The other encodings are supported merely for |
| 461 |
//! legacy compatibility and not due to non-UTF-8 encodings having benefits |
| 462 |
//! other than being able to consume legacy content. |
| 463 |
//! |
| 464 |
//! Considering that UTF-8 can represent all of Unicode and is already supported |
| 465 |
//! by all Web browsers, introducing a new encoding wouldn't add to the |
| 466 |
//! expressiveness but would add to compatibility problems. In that sense, |
| 467 |
//! adding new encodings to the Web Platform doesn't make sense, and, in fact, |
| 468 |
//! post-UTF-8 attempts at encodings, such as BOCU-1, have been rejected from |
| 469 |
//! the Web Platform. On the other hand, the set of legacy encodings that must |
| 470 |
//! be supported for a Web browser to be able to be successful is not going to |
| 471 |
//! expand. Empirically, the set of encodings specified in the Encoding Standard |
| 472 |
//! is already sufficient and the set of legacy encodings won't grow |
| 473 |
//! retroactively. |
| 474 |
//! |
| 475 |
//! Since extensibility doesn't make sense considering the Web focus of |
| 476 |
//! encoding_rs and adding encodings to Web clients would be actively harmful, |
| 477 |
//! it makes sense to make the set of encodings that encoding_rs supports |
| 478 |
//! non-extensible and to take the (admittedly small) benefits arising from |
| 479 |
//! that, such as the size of `Decoder` and `Encoder` objects being known ahead |
| 480 |
//! of time, which enables stack allocation thereof. |
| 481 |
//! |
| 482 |
//! This does have downsides for applications that might want to put encoding_rs |
| 483 |
//! to non-Web uses if those non-Web uses involve legacy encodings that aren't |
| 484 |
//! needed for Web uses. The needs of such applications should not complicate |
| 485 |
//! encoding_rs itself, though. It is up to those applications to provide a |
| 486 |
//! framework that delegates the operations with encodings that encoding_rs |
| 487 |
//! supports to encoding_rs and operations with other encodings to something |
| 488 |
//! else (as opposed to encoding_rs itself providing an extensibility |
| 489 |
//! framework). |
| 490 |
//! |
| 491 |
//! # Panics |
| 492 |
//! |
| 493 |
//! Methods in encoding_rs can panic if the API is used against the requirements |
| 494 |
//! stated in the documentation, if a state that's supposed to be impossible |
| 495 |
//! is reached due to an internal bug or on integer overflow. When used |
| 496 |
//! according to documentation with buffer sizes that stay below integer |
| 497 |
//! overflow, in the absence of internal bugs, encoding_rs does not panic. |
| 498 |
//! |
| 499 |
//! Panics arising from API misuse aren't documented beyond this on individual |
| 500 |
//! methods. |
| 501 |
//! |
| 502 |
//! # At-Risk Parts of the API |
| 503 |
//! |
| 504 |
//! The foreseeable source of partially backward-incompatible API change is the |
| 505 |
//! way the instances of `Encoding` are made available. |
| 506 |
//! |
| 507 |
//! If Rust changes to allow the entries of `[&'static Encoding; N]` to be |
| 508 |
//! initialized with `static`s of type `&'static Encoding`, the non-reference |
| 509 |
//! `FOO_INIT` public `Encoding` instances will be removed from the public API. |
| 510 |
//! |
| 511 |
//! If Rust changes to make the referent of `pub const FOO: &'static Encoding` |
| 512 |
//! unique when the constant is used in different crates, the reference-typed |
| 513 |
//! `static`s for the encoding instances will be changed from `static` to |
| 514 |
//! `const` and the non-reference-typed `_INIT` instances will be removed. |
| 515 |
//! |
| 516 |
//! # Mapping Spec Concepts onto the API |
| 517 |
//! |
| 518 |
//! <table> |
| 519 |
//! <thead> |
| 520 |
//! <tr><th>Spec Concept</th><th>Streaming</th><th>Non-Streaming</th></tr> |
| 521 |
//! </thead> |
| 522 |
//! <tbody> |
| 523 |
//! <tr><td><a href="https://encoding.spec.whatwg.org/#encoding">encoding</a></td><td><code>&'static Encoding</code></td><td><code>&'static Encoding</code></td></tr> |
| 524 |
//! <tr><td><a href="https://encoding.spec.whatwg.org/#utf-8">UTF-8 encoding</a></td><td><code>UTF_8</code></td><td><code>UTF_8</code></td></tr> |
| 525 |
//! <tr><td><a href="https://encoding.spec.whatwg.org/#concept-encoding-get">get an encoding</a></td><td><code>Encoding::for_label(<var>label</var>)</code></td><td><code>Encoding::for_label(<var>label</var>)</code></td></tr> |
| 526 |
//! <tr><td><a href="https://encoding.spec.whatwg.org/#name">name</a></td><td><code><var>encoding</var>.name()</code></td><td><code><var>encoding</var>.name()</code></td></tr> |
| 527 |
//! <tr><td><a href="https://encoding.spec.whatwg.org/#get-an-output-encoding">get an output encoding</a></td><td><code><var>encoding</var>.output_encoding()</code></td><td><code><var>encoding</var>.output_encoding()</code></td></tr> |
| 528 |
//! <tr><td><a href="https://encoding.spec.whatwg.org/#decode">decode</a></td><td><code>let d = <var>encoding</var>.new_decoder();<br>let res = d.decode_to_<var>*</var>(<var>src</var>, <var>dst</var>, false);<br>// …</br>let last_res = d.decode_to_<var>*</var>(<var>src</var>, <var>dst</var>, true);</code></td><td><code><var>encoding</var>.decode(<var>src</var>)</code></td></tr> |
| 529 |
//! <tr><td><a href="https://encoding.spec.whatwg.org/#utf-8-decode">UTF-8 decode</a></td><td><code>let d = UTF_8.new_decoder_with_bom_removal();<br>let res = d.decode_to_<var>*</var>(<var>src</var>, <var>dst</var>, false);<br>// …</br>let last_res = d.decode_to_<var>*</var>(<var>src</var>, <var>dst</var>, true);</code></td><td><code>UTF_8.decode_with_bom_removal(<var>src</var>)</code></td></tr> |
| 530 |
//! <tr><td><a href="https://encoding.spec.whatwg.org/#utf-8-decode-without-bom">UTF-8 decode without BOM</a></td><td><code>let d = UTF_8.new_decoder_without_bom_handling();<br>let res = d.decode_to_<var>*</var>(<var>src</var>, <var>dst</var>, false);<br>// …</br>let last_res = d.decode_to_<var>*</var>(<var>src</var>, <var>dst</var>, true);</code></td><td><code>UTF_8.decode_without_bom_handling(<var>src</var>)</code></td></tr> |
| 531 |
//! <tr><td><a href="https://encoding.spec.whatwg.org/#utf-8-decode-without-bom-or-fail">UTF-8 decode without BOM or fail</a></td><td><code>let d = UTF_8.new_decoder_without_bom_handling();<br>let res = d.decode_to_<var>*</var>_without_replacement(<var>src</var>, <var>dst</var>, false);<br>// … (fail if malformed)</br>let last_res = d.decode_to_<var>*</var>_without_replacement(<var>src</var>, <var>dst</var>, true);<br>// (fail if malformed)</code></td><td><code>UTF_8.decode_without_bom_handling_and_without_replacement(<var>src</var>)</code></td></tr> |
| 532 |
//! <tr><td><a href="https://encoding.spec.whatwg.org/#encode">encode</a></td><td><code>let e = <var>encoding</var>.new_encoder();<br>let res = e.encode_to_<var>*</var>(<var>src</var>, <var>dst</var>, false);<br>// …</br>let last_res = e.encode_to_<var>*</var>(<var>src</var>, <var>dst</var>, true);</code></td><td><code><var>encoding</var>.encode(<var>src</var>)</code></td></tr> |
| 533 |
//! <tr><td><a href="https://encoding.spec.whatwg.org/#utf-8-encode">UTF-8 encode</a></td><td>Use the UTF-8 nature of Rust strings directly:<br><code><var>write</var>(<var>src</var>.as_bytes());<br>// refill src<br><var>write</var>(<var>src</var>.as_bytes());<br>// refill src<br><var>write</var>(<var>src</var>.as_bytes());<br>// …</code></td><td>Use the UTF-8 nature of Rust strings directly:<br><code><var>src</var>.as_bytes()</code></td></tr> |
| 534 |
//! </tbody> |
| 535 |
//! </table> |
| 536 |
//! |
| 537 |
//! # Compatibility with the rust-encoding API |
| 538 |
//! |
| 539 |
//! The crate |
| 540 |
//! [encoding_rs_compat](https://github.com/hsivonen/encoding_rs_compat/) |
| 541 |
//! is a drop-in replacement for rust-encoding 0.2.32 that implements (most of) |
| 542 |
//! the API of rust-encoding 0.2.32 on top of encoding_rs. |
| 543 |
//! |
| 544 |
//! # Mapping rust-encoding concepts to encoding_rs concepts |
| 545 |
//! |
| 546 |
//! The following table provides a mapping from rust-encoding constructs to |
| 547 |
//! encoding_rs ones. |
| 548 |
//! |
| 549 |
//! <table> |
| 550 |
//! <thead> |
| 551 |
//! <tr><th>rust-encoding</th><th>encoding_rs</th></tr> |
| 552 |
//! </thead> |
| 553 |
//! <tbody> |
| 554 |
//! <tr><td><code>encoding::EncodingRef</code></td><td><code>&'static encoding_rs::Encoding</code></td></tr> |
| 555 |
//! <tr><td><code>encoding::all::<var>WINDOWS_31J</var></code> (not based on the WHATWG name for some encodings)</td><td><code>encoding_rs::<var>SHIFT_JIS</var></code> (always the WHATWG name uppercased and hyphens replaced with underscores)</td></tr> |
| 556 |
//! <tr><td><code>encoding::all::ERROR</code></td><td>Not available because not in the Encoding Standard</td></tr> |
| 557 |
//! <tr><td><code>encoding::all::ASCII</code></td><td>Not available because not in the Encoding Standard</td></tr> |
| 558 |
//! <tr><td><code>encoding::all::ISO_8859_1</code></td><td>Not available because not in the Encoding Standard</td></tr> |
| 559 |
//! <tr><td><code>encoding::all::HZ</code></td><td>Not available because not in the Encoding Standard</td></tr> |
| 560 |
//! <tr><td><code>encoding::label::encoding_from_whatwg_label(<var>string</var>)</code></td><td><code>encoding_rs::Encoding::for_label(<var>string</var>)</code></td></tr> |
| 561 |
//! <tr><td><code><var>enc</var>.whatwg_name()</code> (always lower case)</td><td><code><var>enc</var>.name()</code> (potentially mixed case)</td></tr> |
| 562 |
//! <tr><td><code><var>enc</var>.name()</code></td><td>Not available because not in the Encoding Standard</td></tr> |
| 563 |
//! <tr><td><code>encoding::decode(<var>bytes</var>, encoding::DecoderTrap::Replace, <var>enc</var>)</code></td><td><code><var>enc</var>.decode(<var>bytes</var>)</code></td></tr> |
| 564 |
//! <tr><td><code><var>enc</var>.decode(<var>bytes</var>, encoding::DecoderTrap::Replace)</code></td><td><code><var>enc</var>.decode_without_bom_handling(<var>bytes</var>)</code></td></tr> |
| 565 |
//! <tr><td><code><var>enc</var>.encode(<var>string</var>, encoding::EncoderTrap::NcrEscape)</code></td><td><code><var>enc</var>.encode(<var>string</var>)</code></td></tr> |
| 566 |
//! <tr><td><code><var>enc</var>.raw_decoder()</code></td><td><code><var>enc</var>.new_decoder_without_bom_handling()</code></td></tr> |
| 567 |
//! <tr><td><code><var>enc</var>.raw_encoder()</code></td><td><code><var>enc</var>.new_encoder()</code></td></tr> |
| 568 |
//! <tr><td><code>encoding::RawDecoder</code></td><td><code>encoding_rs::Decoder</code></td></tr> |
| 569 |
//! <tr><td><code>encoding::RawEncoder</code></td><td><code>encoding_rs::Encoder</code></td></tr> |
| 570 |
//! <tr><td><code><var>raw_decoder</var>.raw_feed(<var>src</var>, <var>dst_string</var>)</code></td><td><code><var>dst_string</var>.reserve(<var>decoder</var>.max_utf8_buffer_length_without_replacement(<var>src</var>.len()));<br><var>decoder</var>.decode_to_string_without_replacement(<var>src</var>, <var>dst_string</var>, false)</code></td></tr> |
| 571 |
//! <tr><td><code><var>raw_encoder</var>.raw_feed(<var>src</var>, <var>dst_vec</var>)</code></td><td><code><var>dst_vec</var>.reserve(<var>encoder</var>.max_buffer_length_from_utf8_without_replacement(<var>src</var>.len()));<br><var>encoder</var>.encode_from_utf8_to_vec_without_replacement(<var>src</var>, <var>dst_vec</var>, false)</code></td></tr> |
| 572 |
//! <tr><td><code><var>raw_decoder</var>.raw_finish(<var>dst</var>)</code></td><td><code><var>dst_string</var>.reserve(<var>decoder</var>.max_utf8_buffer_length_without_replacement(0));<br><var>decoder</var>.decode_to_string_without_replacement(b"", <var>dst</var>, true)</code></td></tr> |
| 573 |
//! <tr><td><code><var>raw_encoder</var>.raw_finish(<var>dst</var>)</code></td><td><code><var>dst_vec</var>.reserve(<var>encoder</var>.max_buffer_length_from_utf8_without_replacement(0));<br><var>encoder</var>.encode_from_utf8_to_vec_without_replacement("", <var>dst</var>, true)</code></td></tr> |
| 574 |
//! <tr><td><code>encoding::DecoderTrap::Strict</code></td><td><code>decode*</code> methods that have <code>_without_replacement</code> in their name (and treating the `Malformed` result as fatal).</td></tr> |
| 575 |
//! <tr><td><code>encoding::DecoderTrap::Replace</code></td><td><code>decode*</code> methods that <i>do not</i> have <code>_without_replacement</code> in their name.</td></tr> |
| 576 |
//! <tr><td><code>encoding::DecoderTrap::Ignore</code></td><td>It is a bad idea to ignore errors due to security issues, but this could be implemented using <code>decode*</code> methods that have <code>_without_replacement</code> in their name.</td></tr> |
| 577 |
//! <tr><td><code>encoding::DecoderTrap::Call(DecoderTrapFunc)</code></td><td>Can be implemented using <code>decode*</code> methods that have <code>_without_replacement</code> in their name.</td></tr> |
| 578 |
//! <tr><td><code>encoding::EncoderTrap::Strict</code></td><td><code>encode*</code> methods that have <code>_without_replacement</code> in their name (and treating the `Unmappable` result as fatal).</td></tr> |
| 579 |
//! <tr><td><code>encoding::EncoderTrap::Replace</code></td><td>Can be implemented using <code>encode*</code> methods that have <code>_without_replacement</code> in their name.</td></tr> |
| 580 |
//! <tr><td><code>encoding::EncoderTrap::Ignore</code></td><td>It is a bad idea to ignore errors due to security issues, but this could be implemented using <code>encode*</code> methods that have <code>_without_replacement</code> in their name.</td></tr> |
| 581 |
//! <tr><td><code>encoding::EncoderTrap::NcrEscape</code></td><td><code>encode*</code> methods that <i>do not</i> have <code>_without_replacement</code> in their name.</td></tr> |
| 582 |
//! <tr><td><code>encoding::EncoderTrap::Call(EncoderTrapFunc)</code></td><td>Can be implemented using <code>encode*</code> methods that have <code>_without_replacement</code> in their name.</td></tr> |
| 583 |
//! </tbody> |
| 584 |
//! </table> |
| 585 |
//! |
| 586 |
//! # Relationship with Windows Code Pages |
| 587 |
//! |
| 588 |
//! Despite the Web and browser focus, the encodings defined by the Encoding |
| 589 |
//! Standard and implemented by this crate may be useful for decoding legacy |
| 590 |
//! data that uses Windows code pages. The following table names the single-byte |
| 591 |
//! encodings |
| 592 |
//! that have a closely related Windows code page, the number of the closest |
| 593 |
//! code page, a column indicating whether Windows maps unassigned code points |
| 594 |
//! to the Unicode Private Use Area instead of U+FFFD and a remark number |
| 595 |
//! indicating remarks in the list after the table. |
| 596 |
//! |
| 597 |
//! <table> |
| 598 |
//! <thead> |
| 599 |
//! <tr><th>Encoding</th><th>Code Page</th><th>PUA</th><th>Remarks</th></tr> |
| 600 |
//! </thead> |
| 601 |
//! <tbody> |
| 602 |
//! <tr><td>Shift_JIS</td><td>932</td><td></td><td></td></tr> |
| 603 |
//! <tr><td>GBK</td><td>936</td><td></td><td></td></tr> |
| 604 |
//! <tr><td>EUC-KR</td><td>949</td><td></td><td></td></tr> |
| 605 |
//! <tr><td>Big5</td><td>950</td><td></td><td></td></tr> |
| 606 |
//! <tr><td>IBM866</td><td>866</td><td></td><td></td></tr> |
| 607 |
//! <tr><td>windows-874</td><td>874</td><td>•</td><td></td></tr> |
| 608 |
//! <tr><td>UTF-16LE</td><td>1200</td><td></td><td></td></tr> |
| 609 |
//! <tr><td>UTF-16BE</td><td>1201</td><td></td><td></td></tr> |
| 610 |
//! <tr><td>windows-1250</td><td>1250</td><td></td><td></td></tr> |
| 611 |
//! <tr><td>windows-1251</td><td>1251</td><td></td><td></td></tr> |
| 612 |
//! <tr><td>windows-1252</td><td>1252</td><td></td><td></td></tr> |
| 613 |
//! <tr><td>windows-1253</td><td>1253</td><td>•</td><td></td></tr> |
| 614 |
//! <tr><td>windows-1254</td><td>1254</td><td></td><td></td></tr> |
| 615 |
//! <tr><td>windows-1255</td><td>1255</td><td>•</td><td></td></tr> |
| 616 |
//! <tr><td>windows-1256</td><td>1256</td><td></td><td></td></tr> |
| 617 |
//! <tr><td>windows-1257</td><td>1257</td><td>•</td><td></td></tr> |
| 618 |
//! <tr><td>windows-1258</td><td>1258</td><td></td><td></td></tr> |
| 619 |
//! <tr><td>macintosh</td><td>10000</td><td></td><td>1</td></tr> |
| 620 |
//! <tr><td>x-mac-cyrillic</td><td>10017</td><td></td><td>2</td></tr> |
| 621 |
//! <tr><td>KOI8-R</td><td>20866</td><td></td><td></td></tr> |
| 622 |
//! <tr><td>EUC-JP</td><td>20932</td><td></td><td></td></tr> |
| 623 |
//! <tr><td>KOI8-U</td><td>21866</td><td></td><td></td></tr> |
| 624 |
//! <tr><td>ISO-8859-2</td><td>28592</td><td></td><td></td></tr> |
| 625 |
//! <tr><td>ISO-8859-3</td><td>28593</td><td></td><td></td></tr> |
| 626 |
//! <tr><td>ISO-8859-4</td><td>28594</td><td></td><td></td></tr> |
| 627 |
//! <tr><td>ISO-8859-5</td><td>28595</td><td></td><td></td></tr> |
| 628 |
//! <tr><td>ISO-8859-6</td><td>28596</td><td>•</td><td></td></tr> |
| 629 |
//! <tr><td>ISO-8859-7</td><td>28597</td><td>•</td><td>3</td></tr> |
| 630 |
//! <tr><td>ISO-8859-8</td><td>28598</td><td>•</td><td>4</td></tr> |
| 631 |
//! <tr><td>ISO-8859-13</td><td>28603</td><td>•</td><td></td></tr> |
| 632 |
//! <tr><td>ISO-8859-15</td><td>28605</td><td></td><td></td></tr> |
| 633 |
//! <tr><td>ISO-8859-8-I</td><td>38598</td><td></td><td>5</td></tr> |
| 634 |
//! <tr><td>ISO-2022-JP</td><td>50220</td><td></td><td></td></tr> |
| 635 |
//! <tr><td>gb18030</td><td>54936</td><td></td><td></td></tr> |
| 636 |
//! <tr><td>UTF-8</td><td>65001</td><td></td><td></td></tr> |
| 637 |
//! </tbody> |
| 638 |
//! </table> |
| 639 |
//! |
| 640 |
//! 1. Windows decodes 0xBD to U+2126 OHM SIGN instead of U+03A9 GREEK CAPITAL LETTER OMEGA. |
| 641 |
//! 2. Windows decodes 0xFF to U+00A4 CURRENCY SIGN instead of U+20AC EURO SIGN. |
| 642 |
//! 3. Windows decodes the currency signs at 0xA4 and 0xA5 as well as 0xAA, |
| 643 |
//! which should be U+037A GREEK YPOGEGRAMMENI, to PUA code points. Windows |
| 644 |
//! decodes 0xA1 to U+02BD MODIFIER LETTER REVERSED COMMA instead of U+2018 |
| 645 |
//! LEFT SINGLE QUOTATION MARK and 0xA2 to U+02BC MODIFIER LETTER APOSTROPHE |
| 646 |
//! instead of U+2019 RIGHT SINGLE QUOTATION MARK. |
| 647 |
//! 4. Windows decodes 0xAF to OVERLINE instead of MACRON and 0xFE and 0xFD to PUA instead |
| 648 |
//! of LRM and RLM. |
| 649 |
//! 5. Remarks from the previous item apply. |
| 650 |
//! |
| 651 |
//! The differences between this crate and Windows in the case of multibyte encodings |
| 652 |
//! are not yet fully documented here. The lack of remarks above should not be taken |
| 653 |
//! as indication of lack of differences. |
| 654 |
//! |
| 655 |
//! # Notable Differences from IANA Naming |
| 656 |
//! |
| 657 |
//! In some cases, the Encoding Standard specifies the popular unextended encoding |
| 658 |
//! name where in IANA terms one of the other labels would be more precise considering |
| 659 |
//! the extensions that the Encoding Standard has unified into the encoding. |
| 660 |
//! |
| 661 |
//! <table> |
| 662 |
//! <thead> |
| 663 |
//! <tr><th>Encoding</th><th>IANA</th></tr> |
| 664 |
//! </thead> |
| 665 |
//! <tbody> |
| 666 |
//! <tr><td>Big5</td><td>Big5-HKSCS</td></tr> |
| 667 |
//! <tr><td>EUC-KR</td><td>windows-949</td></tr> |
| 668 |
//! <tr><td>Shift_JIS</td><td>windows-31j</td></tr> |
| 669 |
//! <tr><td>x-mac-cyrillic</td><td>x-mac-ukrainian</td></tr> |
| 670 |
//! </tbody> |
| 671 |
//! </table> |
| 672 |
//! |
| 673 |
//! In other cases where the Encoding Standard unifies unextended and extended |
| 674 |
//! variants of an encoding, the encoding gets the name of the extended |
| 675 |
//! variant. |
| 676 |
//! |
| 677 |
//! <table> |
| 678 |
//! <thead> |
| 679 |
//! <tr><th>IANA</th><th>Unified into Encoding</th></tr> |
| 680 |
//! </thead> |
| 681 |
//! <tbody> |
| 682 |
//! <tr><td>ISO-8859-1</td><td>windows-1252</td></tr> |
| 683 |
//! <tr><td>ISO-8859-9</td><td>windows-1254</td></tr> |
| 684 |
//! <tr><td>TIS-620</td><td>windows-874</td></tr> |
| 685 |
//! </tbody> |
| 686 |
//! </table> |
| 687 |
//! |
| 688 |
//! See the section [_UTF-16LE, UTF-16BE and Unicode Encoding Schemes_](#utf-16le-utf-16be-and-unicode-encoding-schemes) |
| 689 |
//! for discussion about the UTF-16 family. |
| 690 |
|
| 691 |
#![no_std] |
| 692 |
#![cfg_attr(feature = "simd-accel", feature(core_intrinsics, portable_simd))] |
| 693 |
|
| 694 |
#[cfg(feature = "alloc")] |
| 695 |
#[cfg_attr(test, macro_use)] |
| 696 |
extern crate alloc; |
| 697 |
|
| 698 |
extern crate core; |
| 699 |
#[macro_use] |
| 700 |
extern crate cfg_if; |
| 701 |
|
| 702 |
#[cfg(feature = "serde")] |
| 703 |
extern crate serde; |
| 704 |
|
| 705 |
#[cfg(all(test, feature = "serde"))] |
| 706 |
extern crate bincode; |
| 707 |
#[cfg(all(test, feature = "serde"))] |
| 708 |
#[macro_use] |
| 709 |
extern crate serde_derive; |
| 710 |
#[cfg(all(test, feature = "serde"))] |
| 711 |
extern crate serde_json; |
| 712 |
|
| 713 |
#[macro_use] |
| 714 |
mod macros; |
| 715 |
|
| 716 |
#[cfg(all( |
| 717 |
feature = "simd-accel", |
| 718 |
any( |
| 719 |
target_feature = "sse2", |
| 720 |
all(target_endian = "little", target_arch = "aarch64"), |
| 721 |
all(target_endian = "little", target_feature = "neon") |
| 722 |
) |
| 723 |
))] |
| 724 |
mod simd_funcs; |
| 725 |
|
| 726 |
#[cfg(all(test, feature = "alloc"))] |
| 727 |
mod testing; |
| 728 |
|
| 729 |
mod big5; |
| 730 |
mod euc_jp; |
| 731 |
mod euc_kr; |
| 732 |
mod gb18030; |
| 733 |
mod iso_2022_jp; |
| 734 |
mod replacement; |
| 735 |
mod shift_jis; |
| 736 |
mod single_byte; |
| 737 |
mod utf_16; |
| 738 |
mod utf_8; |
| 739 |
mod x_user_defined; |
| 740 |
|
| 741 |
mod ascii; |
| 742 |
mod data; |
| 743 |
mod handles; |
| 744 |
mod variant; |
| 745 |
|
| 746 |
pub mod mem; |
| 747 |
|
| 748 |
use crate::ascii::ascii_valid_up_to; |
| 749 |
use crate::ascii::iso_2022_jp_ascii_valid_up_to; |
| 750 |
use crate::utf_8::utf8_valid_up_to; |
| 751 |
use crate::variant::*; |
| 752 |
|
| 753 |
#[cfg(feature = "alloc")] |
| 754 |
use alloc::borrow::Cow; |
| 755 |
#[cfg(feature = "alloc")] |
| 756 |
use alloc::string::String; |
| 757 |
#[cfg(feature = "alloc")] |
| 758 |
use alloc::vec::Vec; |
| 759 |
use core::cmp::Ordering; |
| 760 |
use core::hash::Hash; |
| 761 |
use core::hash::Hasher; |
| 762 |
|
| 763 |
#[cfg(feature = "serde")] |
| 764 |
use serde::de::Visitor; |
| 765 |
#[cfg(feature = "serde")] |
| 766 |
use serde::{Deserialize, Deserializer, Serialize, Serializer}; |
| 767 |
|
| 768 |
/// This has to be the max length of an NCR instead of max |
| 769 |
/// minus one, because we can't rely on getting the minus |
| 770 |
/// one from the space reserved for the current unmappable, |
| 771 |
/// because the ISO-2022-JP encoder can fill up that space |
| 772 |
/// with a state transition escape. |
| 773 |
const NCR_EXTRA: usize = 10; //  |
| 774 |
|
| 775 |
// BEGIN GENERATED CODE. PLEASE DO NOT EDIT. |
| 776 |
// Instead, please regenerate using generate-encoding-data.py |
| 777 |
|
| 778 |
const LONGEST_LABEL_LENGTH: usize = 19; // cseucpkdfmtjapanese |
| 779 |
|
| 780 |
/// The initializer for the [Big5](static.BIG5.html) encoding. |
| 781 |
/// |
| 782 |
/// For use only for taking the address of this form when |
| 783 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 784 |
/// such as in initializers of other `static`s. If in doubt, |
| 785 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 786 |
/// |
| 787 |
/// This part of the public API will go away if Rust changes |
| 788 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 789 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 790 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 791 |
/// items. |
| 792 |
pub static BIG5_INIT: Encoding = Encoding { |
| 793 |
name: "Big5", |
| 794 |
variant: VariantEncoding::Big5, |
| 795 |
}; |
| 796 |
|
| 797 |
/// The Big5 encoding. |
| 798 |
/// |
| 799 |
/// This is Big5 with HKSCS with mappings to more recent Unicode assignments |
| 800 |
/// instead of the Private Use Area code points that have been used historically. |
| 801 |
/// It is believed to be able to decode existing Web content in a way that makes |
| 802 |
/// sense. |
| 803 |
/// |
| 804 |
/// To avoid form submissions generating data that Web servers don't understand, |
| 805 |
/// the encoder doesn't use the HKSCS byte sequences that precede the unextended |
| 806 |
/// Big5 in the lexical order. |
| 807 |
/// |
| 808 |
/// [Index visualization](https://encoding.spec.whatwg.org/big5.html), |
| 809 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/big5-bmp.html) |
| 810 |
/// |
| 811 |
/// This encoding is designed to be suited for decoding the Windows code page 950 |
| 812 |
/// and its HKSCS patched "951" variant such that the text makes sense, given |
| 813 |
/// assignments that Unicode has made after those encodings used Private Use |
| 814 |
/// Area characters. |
| 815 |
/// |
| 816 |
/// This will change from `static` to `const` if Rust changes |
| 817 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 818 |
/// unique cross-crate, so don't take the address of this |
| 819 |
/// `static`. |
| 820 |
pub static BIG5: &'static Encoding = &BIG5_INIT; |
| 821 |
|
| 822 |
/// The initializer for the [EUC-JP](static.EUC_JP.html) encoding. |
| 823 |
/// |
| 824 |
/// For use only for taking the address of this form when |
| 825 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 826 |
/// such as in initializers of other `static`s. If in doubt, |
| 827 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 828 |
/// |
| 829 |
/// This part of the public API will go away if Rust changes |
| 830 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 831 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 832 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 833 |
/// items. |
| 834 |
pub static EUC_JP_INIT: Encoding = Encoding { |
| 835 |
name: "EUC-JP", |
| 836 |
variant: VariantEncoding::EucJp, |
| 837 |
}; |
| 838 |
|
| 839 |
/// The EUC-JP encoding. |
| 840 |
/// |
| 841 |
/// This is the legacy Unix encoding for Japanese. |
| 842 |
/// |
| 843 |
/// For compatibility with Web servers that don't expect three-byte sequences |
| 844 |
/// in form submissions, the encoder doesn't generate three-byte sequences. |
| 845 |
/// That is, the JIS X 0212 support is decode-only. |
| 846 |
/// |
| 847 |
/// [Index visualization](https://encoding.spec.whatwg.org/euc-jp.html), |
| 848 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/euc-jp-bmp.html) |
| 849 |
/// |
| 850 |
/// This encoding roughly matches the Windows code page 20932. There are error |
| 851 |
/// handling differences and a handful of 2-byte sequences that decode differently. |
| 852 |
/// Additionall, Windows doesn't support 3-byte sequences. |
| 853 |
/// |
| 854 |
/// This will change from `static` to `const` if Rust changes |
| 855 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 856 |
/// unique cross-crate, so don't take the address of this |
| 857 |
/// `static`. |
| 858 |
pub static EUC_JP: &'static Encoding = &EUC_JP_INIT; |
| 859 |
|
| 860 |
/// The initializer for the [EUC-KR](static.EUC_KR.html) encoding. |
| 861 |
/// |
| 862 |
/// For use only for taking the address of this form when |
| 863 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 864 |
/// such as in initializers of other `static`s. If in doubt, |
| 865 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 866 |
/// |
| 867 |
/// This part of the public API will go away if Rust changes |
| 868 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 869 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 870 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 871 |
/// items. |
| 872 |
pub static EUC_KR_INIT: Encoding = Encoding { |
| 873 |
name: "EUC-KR", |
| 874 |
variant: VariantEncoding::EucKr, |
| 875 |
}; |
| 876 |
|
| 877 |
/// The EUC-KR encoding. |
| 878 |
/// |
| 879 |
/// This is the Korean encoding for Windows. It extends the Unix legacy encoding |
| 880 |
/// for Korean, based on KS X 1001 (which also formed the base of MacKorean on Mac OS |
| 881 |
/// Classic), with all the characters from the Hangul Syllables block of Unicode. |
| 882 |
/// |
| 883 |
/// [Index visualization](https://encoding.spec.whatwg.org/euc-kr.html), |
| 884 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/euc-kr-bmp.html) |
| 885 |
/// |
| 886 |
/// This encoding matches the Windows code page 949, except Windows decodes byte 0x80 |
| 887 |
/// to U+0080 and some byte sequences that are error per the Encoding Standard to |
| 888 |
/// the question mark or the Private Use Area. |
| 889 |
/// |
| 890 |
/// This will change from `static` to `const` if Rust changes |
| 891 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 892 |
/// unique cross-crate, so don't take the address of this |
| 893 |
/// `static`. |
| 894 |
pub static EUC_KR: &'static Encoding = &EUC_KR_INIT; |
| 895 |
|
| 896 |
/// The initializer for the [GBK](static.GBK.html) encoding. |
| 897 |
/// |
| 898 |
/// For use only for taking the address of this form when |
| 899 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 900 |
/// such as in initializers of other `static`s. If in doubt, |
| 901 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 902 |
/// |
| 903 |
/// This part of the public API will go away if Rust changes |
| 904 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 905 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 906 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 907 |
/// items. |
| 908 |
pub static GBK_INIT: Encoding = Encoding { |
| 909 |
name: "GBK", |
| 910 |
variant: VariantEncoding::Gbk, |
| 911 |
}; |
| 912 |
|
| 913 |
/// The GBK encoding. |
| 914 |
/// |
| 915 |
/// The decoder for this encoding is the same as the decoder for gb18030. |
| 916 |
/// The encoder side of this encoding is GBK with Windows code page 936 euro |
| 917 |
/// sign behavior. GBK extends GB2312-80 to cover the CJK Unified Ideographs |
| 918 |
/// Unicode block as well as a handful of ideographs from the CJK Unified |
| 919 |
/// Ideographs Extension A and CJK Compatibility Ideographs blocks. |
| 920 |
/// |
| 921 |
/// Unlike e.g. in the case of ISO-8859-1 and windows-1252, GBK encoder wasn't |
| 922 |
/// unified with the gb18030 encoder in the Encoding Standard out of concern |
| 923 |
/// that servers that expect GBK form submissions might not be able to handle |
| 924 |
/// the four-byte sequences. |
| 925 |
/// |
| 926 |
/// [Index visualization for the two-byte sequences](https://encoding.spec.whatwg.org/gb18030.html), |
| 927 |
/// [Visualization of BMP coverage of the two-byte index](https://encoding.spec.whatwg.org/gb18030-bmp.html) |
| 928 |
/// |
| 929 |
/// The encoder of this encoding roughly matches the Windows code page 936. |
| 930 |
/// The decoder side is a superset. |
| 931 |
/// |
| 932 |
/// This will change from `static` to `const` if Rust changes |
| 933 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 934 |
/// unique cross-crate, so don't take the address of this |
| 935 |
/// `static`. |
| 936 |
pub static GBK: &'static Encoding = &GBK_INIT; |
| 937 |
|
| 938 |
/// The initializer for the [IBM866](static.IBM866.html) encoding. |
| 939 |
/// |
| 940 |
/// For use only for taking the address of this form when |
| 941 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 942 |
/// such as in initializers of other `static`s. If in doubt, |
| 943 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 944 |
/// |
| 945 |
/// This part of the public API will go away if Rust changes |
| 946 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 947 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 948 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 949 |
/// items. |
| 950 |
pub static IBM866_INIT: Encoding = Encoding { |
| 951 |
name: "IBM866", |
| 952 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.ibm866, 0x0440, 96, 16), |
| 953 |
}; |
| 954 |
|
| 955 |
/// The IBM866 encoding. |
| 956 |
/// |
| 957 |
/// This the most notable one of the DOS Cyrillic code pages. It has the same |
| 958 |
/// box drawing characters as code page 437, so it can be used for decoding |
| 959 |
/// DOS-era ASCII + box drawing data. |
| 960 |
/// |
| 961 |
/// [Index visualization](https://encoding.spec.whatwg.org/ibm866.html), |
| 962 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/ibm866-bmp.html) |
| 963 |
/// |
| 964 |
/// This encoding matches the Windows code page 866. |
| 965 |
/// |
| 966 |
/// This will change from `static` to `const` if Rust changes |
| 967 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 968 |
/// unique cross-crate, so don't take the address of this |
| 969 |
/// `static`. |
| 970 |
pub static IBM866: &'static Encoding = &IBM866_INIT; |
| 971 |
|
| 972 |
/// The initializer for the [ISO-2022-JP](static.ISO_2022_JP.html) encoding. |
| 973 |
/// |
| 974 |
/// For use only for taking the address of this form when |
| 975 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 976 |
/// such as in initializers of other `static`s. If in doubt, |
| 977 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 978 |
/// |
| 979 |
/// This part of the public API will go away if Rust changes |
| 980 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 981 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 982 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 983 |
/// items. |
| 984 |
pub static ISO_2022_JP_INIT: Encoding = Encoding { |
| 985 |
name: "ISO-2022-JP", |
| 986 |
variant: VariantEncoding::Iso2022Jp, |
| 987 |
}; |
| 988 |
|
| 989 |
/// The ISO-2022-JP encoding. |
| 990 |
/// |
| 991 |
/// This the primary pre-UTF-8 encoding for Japanese email. It uses the ASCII |
| 992 |
/// byte range to encode non-Basic Latin characters. It's the only encoding |
| 993 |
/// supported by this crate whose encoder is stateful. |
| 994 |
/// |
| 995 |
/// [Index visualization](https://encoding.spec.whatwg.org/jis0208.html), |
| 996 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/jis0208-bmp.html) |
| 997 |
/// |
| 998 |
/// This encoding roughly matches the Windows code page 50220. Notably, Windows |
| 999 |
/// uses U+30FB in place of the REPLACEMENT CHARACTER and otherwise differs in |
| 1000 |
/// error handling. |
| 1001 |
/// |
| 1002 |
/// This will change from `static` to `const` if Rust changes |
| 1003 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1004 |
/// unique cross-crate, so don't take the address of this |
| 1005 |
/// `static`. |
| 1006 |
pub static ISO_2022_JP: &'static Encoding = &ISO_2022_JP_INIT; |
| 1007 |
|
| 1008 |
/// The initializer for the [ISO-8859-10](static.ISO_8859_10.html) encoding. |
| 1009 |
/// |
| 1010 |
/// For use only for taking the address of this form when |
| 1011 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1012 |
/// such as in initializers of other `static`s. If in doubt, |
| 1013 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1014 |
/// |
| 1015 |
/// This part of the public API will go away if Rust changes |
| 1016 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1017 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1018 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1019 |
/// items. |
| 1020 |
pub static ISO_8859_10_INIT: Encoding = Encoding { |
| 1021 |
name: "ISO-8859-10", |
| 1022 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_10, 0x00DA, 90, 6), |
| 1023 |
}; |
| 1024 |
|
| 1025 |
/// The ISO-8859-10 encoding. |
| 1026 |
/// |
| 1027 |
/// This is the Nordic part of the ISO/IEC 8859 encoding family. This encoding |
| 1028 |
/// is also known as Latin 6. |
| 1029 |
/// |
| 1030 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-10.html), |
| 1031 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-10-bmp.html) |
| 1032 |
/// |
| 1033 |
/// The Windows code page number for this encoding is 28600, but kernel32.dll |
| 1034 |
/// does not support this encoding. |
| 1035 |
/// |
| 1036 |
/// This will change from `static` to `const` if Rust changes |
| 1037 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1038 |
/// unique cross-crate, so don't take the address of this |
| 1039 |
/// `static`. |
| 1040 |
pub static ISO_8859_10: &'static Encoding = &ISO_8859_10_INIT; |
| 1041 |
|
| 1042 |
/// The initializer for the [ISO-8859-13](static.ISO_8859_13.html) encoding. |
| 1043 |
/// |
| 1044 |
/// For use only for taking the address of this form when |
| 1045 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1046 |
/// such as in initializers of other `static`s. If in doubt, |
| 1047 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1048 |
/// |
| 1049 |
/// This part of the public API will go away if Rust changes |
| 1050 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1051 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1052 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1053 |
/// items. |
| 1054 |
pub static ISO_8859_13_INIT: Encoding = Encoding { |
| 1055 |
name: "ISO-8859-13", |
| 1056 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_13, 0x00DF, 95, 1), |
| 1057 |
}; |
| 1058 |
|
| 1059 |
/// The ISO-8859-13 encoding. |
| 1060 |
/// |
| 1061 |
/// This is the Baltic part of the ISO/IEC 8859 encoding family. This encoding |
| 1062 |
/// is also known as Latin 7. |
| 1063 |
/// |
| 1064 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-13.html), |
| 1065 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-13-bmp.html) |
| 1066 |
/// |
| 1067 |
/// This encoding matches the Windows code page 28603, except Windows decodes |
| 1068 |
/// unassigned code points to the Private Use Area of Unicode. |
| 1069 |
/// |
| 1070 |
/// This will change from `static` to `const` if Rust changes |
| 1071 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1072 |
/// unique cross-crate, so don't take the address of this |
| 1073 |
/// `static`. |
| 1074 |
pub static ISO_8859_13: &'static Encoding = &ISO_8859_13_INIT; |
| 1075 |
|
| 1076 |
/// The initializer for the [ISO-8859-14](static.ISO_8859_14.html) encoding. |
| 1077 |
/// |
| 1078 |
/// For use only for taking the address of this form when |
| 1079 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1080 |
/// such as in initializers of other `static`s. If in doubt, |
| 1081 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1082 |
/// |
| 1083 |
/// This part of the public API will go away if Rust changes |
| 1084 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1085 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1086 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1087 |
/// items. |
| 1088 |
pub static ISO_8859_14_INIT: Encoding = Encoding { |
| 1089 |
name: "ISO-8859-14", |
| 1090 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_14, 0x00DF, 95, 17), |
| 1091 |
}; |
| 1092 |
|
| 1093 |
/// The ISO-8859-14 encoding. |
| 1094 |
/// |
| 1095 |
/// This is the Celtic part of the ISO/IEC 8859 encoding family. This encoding |
| 1096 |
/// is also known as Latin 8. |
| 1097 |
/// |
| 1098 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-14.html), |
| 1099 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-14-bmp.html) |
| 1100 |
/// |
| 1101 |
/// The Windows code page number for this encoding is 28604, but kernel32.dll |
| 1102 |
/// does not support this encoding. |
| 1103 |
/// |
| 1104 |
/// This will change from `static` to `const` if Rust changes |
| 1105 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1106 |
/// unique cross-crate, so don't take the address of this |
| 1107 |
/// `static`. |
| 1108 |
pub static ISO_8859_14: &'static Encoding = &ISO_8859_14_INIT; |
| 1109 |
|
| 1110 |
/// The initializer for the [ISO-8859-15](static.ISO_8859_15.html) encoding. |
| 1111 |
/// |
| 1112 |
/// For use only for taking the address of this form when |
| 1113 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1114 |
/// such as in initializers of other `static`s. If in doubt, |
| 1115 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1116 |
/// |
| 1117 |
/// This part of the public API will go away if Rust changes |
| 1118 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1119 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1120 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1121 |
/// items. |
| 1122 |
pub static ISO_8859_15_INIT: Encoding = Encoding { |
| 1123 |
name: "ISO-8859-15", |
| 1124 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_15, 0x00BF, 63, 65), |
| 1125 |
}; |
| 1126 |
|
| 1127 |
/// The ISO-8859-15 encoding. |
| 1128 |
/// |
| 1129 |
/// This is the revised Western European part of the ISO/IEC 8859 encoding |
| 1130 |
/// family. This encoding is also known as Latin 9. |
| 1131 |
/// |
| 1132 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-15.html), |
| 1133 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-15-bmp.html) |
| 1134 |
/// |
| 1135 |
/// This encoding matches the Windows code page 28605. |
| 1136 |
/// |
| 1137 |
/// This will change from `static` to `const` if Rust changes |
| 1138 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1139 |
/// unique cross-crate, so don't take the address of this |
| 1140 |
/// `static`. |
| 1141 |
pub static ISO_8859_15: &'static Encoding = &ISO_8859_15_INIT; |
| 1142 |
|
| 1143 |
/// The initializer for the [ISO-8859-16](static.ISO_8859_16.html) encoding. |
| 1144 |
/// |
| 1145 |
/// For use only for taking the address of this form when |
| 1146 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1147 |
/// such as in initializers of other `static`s. If in doubt, |
| 1148 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1149 |
/// |
| 1150 |
/// This part of the public API will go away if Rust changes |
| 1151 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1152 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1153 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1154 |
/// items. |
| 1155 |
pub static ISO_8859_16_INIT: Encoding = Encoding { |
| 1156 |
name: "ISO-8859-16", |
| 1157 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_16, 0x00DF, 95, 4), |
| 1158 |
}; |
| 1159 |
|
| 1160 |
/// The ISO-8859-16 encoding. |
| 1161 |
/// |
| 1162 |
/// This is the South-Eastern European part of the ISO/IEC 8859 encoding |
| 1163 |
/// family. This encoding is also known as Latin 10. |
| 1164 |
/// |
| 1165 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-16.html), |
| 1166 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-16-bmp.html) |
| 1167 |
/// |
| 1168 |
/// The Windows code page number for this encoding is 28606, but kernel32.dll |
| 1169 |
/// does not support this encoding. |
| 1170 |
/// |
| 1171 |
/// This will change from `static` to `const` if Rust changes |
| 1172 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1173 |
/// unique cross-crate, so don't take the address of this |
| 1174 |
/// `static`. |
| 1175 |
pub static ISO_8859_16: &'static Encoding = &ISO_8859_16_INIT; |
| 1176 |
|
| 1177 |
/// The initializer for the [ISO-8859-2](static.ISO_8859_2.html) encoding. |
| 1178 |
/// |
| 1179 |
/// For use only for taking the address of this form when |
| 1180 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1181 |
/// such as in initializers of other `static`s. If in doubt, |
| 1182 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1183 |
/// |
| 1184 |
/// This part of the public API will go away if Rust changes |
| 1185 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1186 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1187 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1188 |
/// items. |
| 1189 |
pub static ISO_8859_2_INIT: Encoding = Encoding { |
| 1190 |
name: "ISO-8859-2", |
| 1191 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_2, 0x00DF, 95, 1), |
| 1192 |
}; |
| 1193 |
|
| 1194 |
/// The ISO-8859-2 encoding. |
| 1195 |
/// |
| 1196 |
/// This is the Central European part of the ISO/IEC 8859 encoding family. This encoding is also known as Latin 2. |
| 1197 |
/// |
| 1198 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-2.html), |
| 1199 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-2-bmp.html) |
| 1200 |
/// |
| 1201 |
/// This encoding matches the Windows code page 28592. |
| 1202 |
/// |
| 1203 |
/// This will change from `static` to `const` if Rust changes |
| 1204 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1205 |
/// unique cross-crate, so don't take the address of this |
| 1206 |
/// `static`. |
| 1207 |
pub static ISO_8859_2: &'static Encoding = &ISO_8859_2_INIT; |
| 1208 |
|
| 1209 |
/// The initializer for the [ISO-8859-3](static.ISO_8859_3.html) encoding. |
| 1210 |
/// |
| 1211 |
/// For use only for taking the address of this form when |
| 1212 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1213 |
/// such as in initializers of other `static`s. If in doubt, |
| 1214 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1215 |
/// |
| 1216 |
/// This part of the public API will go away if Rust changes |
| 1217 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1218 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1219 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1220 |
/// items. |
| 1221 |
pub static ISO_8859_3_INIT: Encoding = Encoding { |
| 1222 |
name: "ISO-8859-3", |
| 1223 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_3, 0x00DF, 95, 4), |
| 1224 |
}; |
| 1225 |
|
| 1226 |
/// The ISO-8859-3 encoding. |
| 1227 |
/// |
| 1228 |
/// This is the South European part of the ISO/IEC 8859 encoding family. This encoding is also known as Latin 3. |
| 1229 |
/// |
| 1230 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-3.html), |
| 1231 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-3-bmp.html) |
| 1232 |
/// |
| 1233 |
/// This encoding matches the Windows code page 28593. |
| 1234 |
/// |
| 1235 |
/// This will change from `static` to `const` if Rust changes |
| 1236 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1237 |
/// unique cross-crate, so don't take the address of this |
| 1238 |
/// `static`. |
| 1239 |
pub static ISO_8859_3: &'static Encoding = &ISO_8859_3_INIT; |
| 1240 |
|
| 1241 |
/// The initializer for the [ISO-8859-4](static.ISO_8859_4.html) encoding. |
| 1242 |
/// |
| 1243 |
/// For use only for taking the address of this form when |
| 1244 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1245 |
/// such as in initializers of other `static`s. If in doubt, |
| 1246 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1247 |
/// |
| 1248 |
/// This part of the public API will go away if Rust changes |
| 1249 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1250 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1251 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1252 |
/// items. |
| 1253 |
pub static ISO_8859_4_INIT: Encoding = Encoding { |
| 1254 |
name: "ISO-8859-4", |
| 1255 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_4, 0x00DF, 95, 1), |
| 1256 |
}; |
| 1257 |
|
| 1258 |
/// The ISO-8859-4 encoding. |
| 1259 |
/// |
| 1260 |
/// This is the North European part of the ISO/IEC 8859 encoding family. This encoding is also known as Latin 4. |
| 1261 |
/// |
| 1262 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-4.html), |
| 1263 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-4-bmp.html) |
| 1264 |
/// |
| 1265 |
/// This encoding matches the Windows code page 28594. |
| 1266 |
/// |
| 1267 |
/// This will change from `static` to `const` if Rust changes |
| 1268 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1269 |
/// unique cross-crate, so don't take the address of this |
| 1270 |
/// `static`. |
| 1271 |
pub static ISO_8859_4: &'static Encoding = &ISO_8859_4_INIT; |
| 1272 |
|
| 1273 |
/// The initializer for the [ISO-8859-5](static.ISO_8859_5.html) encoding. |
| 1274 |
/// |
| 1275 |
/// For use only for taking the address of this form when |
| 1276 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1277 |
/// such as in initializers of other `static`s. If in doubt, |
| 1278 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1279 |
/// |
| 1280 |
/// This part of the public API will go away if Rust changes |
| 1281 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1282 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1283 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1284 |
/// items. |
| 1285 |
pub static ISO_8859_5_INIT: Encoding = Encoding { |
| 1286 |
name: "ISO-8859-5", |
| 1287 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_5, 0x040E, 46, 66), |
| 1288 |
}; |
| 1289 |
|
| 1290 |
/// The ISO-8859-5 encoding. |
| 1291 |
/// |
| 1292 |
/// This is the Cyrillic part of the ISO/IEC 8859 encoding family. |
| 1293 |
/// |
| 1294 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-5.html), |
| 1295 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-5-bmp.html) |
| 1296 |
/// |
| 1297 |
/// This encoding matches the Windows code page 28595. |
| 1298 |
/// |
| 1299 |
/// This will change from `static` to `const` if Rust changes |
| 1300 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1301 |
/// unique cross-crate, so don't take the address of this |
| 1302 |
/// `static`. |
| 1303 |
pub static ISO_8859_5: &'static Encoding = &ISO_8859_5_INIT; |
| 1304 |
|
| 1305 |
/// The initializer for the [ISO-8859-6](static.ISO_8859_6.html) encoding. |
| 1306 |
/// |
| 1307 |
/// For use only for taking the address of this form when |
| 1308 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1309 |
/// such as in initializers of other `static`s. If in doubt, |
| 1310 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1311 |
/// |
| 1312 |
/// This part of the public API will go away if Rust changes |
| 1313 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1314 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1315 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1316 |
/// items. |
| 1317 |
pub static ISO_8859_6_INIT: Encoding = Encoding { |
| 1318 |
name: "ISO-8859-6", |
| 1319 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_6, 0x0621, 65, 26), |
| 1320 |
}; |
| 1321 |
|
| 1322 |
/// The ISO-8859-6 encoding. |
| 1323 |
/// |
| 1324 |
/// This is the Arabic part of the ISO/IEC 8859 encoding family. |
| 1325 |
/// |
| 1326 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-6.html), |
| 1327 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-6-bmp.html) |
| 1328 |
/// |
| 1329 |
/// This encoding matches the Windows code page 28596, except Windows decodes |
| 1330 |
/// unassigned code points to the Private Use Area of Unicode. |
| 1331 |
/// |
| 1332 |
/// This will change from `static` to `const` if Rust changes |
| 1333 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1334 |
/// unique cross-crate, so don't take the address of this |
| 1335 |
/// `static`. |
| 1336 |
pub static ISO_8859_6: &'static Encoding = &ISO_8859_6_INIT; |
| 1337 |
|
| 1338 |
/// The initializer for the [ISO-8859-7](static.ISO_8859_7.html) encoding. |
| 1339 |
/// |
| 1340 |
/// For use only for taking the address of this form when |
| 1341 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1342 |
/// such as in initializers of other `static`s. If in doubt, |
| 1343 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1344 |
/// |
| 1345 |
/// This part of the public API will go away if Rust changes |
| 1346 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1347 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1348 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1349 |
/// items. |
| 1350 |
pub static ISO_8859_7_INIT: Encoding = Encoding { |
| 1351 |
name: "ISO-8859-7", |
| 1352 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_7, 0x03A3, 83, 44), |
| 1353 |
}; |
| 1354 |
|
| 1355 |
/// The ISO-8859-7 encoding. |
| 1356 |
/// |
| 1357 |
/// This is the Greek part of the ISO/IEC 8859 encoding family. |
| 1358 |
/// |
| 1359 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-7.html), |
| 1360 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-7-bmp.html) |
| 1361 |
/// |
| 1362 |
/// This encoding roughly matches the Windows code page 28597. Windows decodes |
| 1363 |
/// unassigned code points, the currency signs at 0xA4 and 0xA5 as well as |
| 1364 |
/// 0xAA, which should be U+037A GREEK YPOGEGRAMMENI, to the Private Use Area |
| 1365 |
/// of Unicode. Windows decodes 0xA1 to U+02BD MODIFIER LETTER REVERSED COMMA |
| 1366 |
/// instead of U+2018 LEFT SINGLE QUOTATION MARK and 0xA2 to U+02BC MODIFIER |
| 1367 |
/// LETTER APOSTROPHE instead of U+2019 RIGHT SINGLE QUOTATION MARK. |
| 1368 |
/// |
| 1369 |
/// This will change from `static` to `const` if Rust changes |
| 1370 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1371 |
/// unique cross-crate, so don't take the address of this |
| 1372 |
/// `static`. |
| 1373 |
pub static ISO_8859_7: &'static Encoding = &ISO_8859_7_INIT; |
| 1374 |
|
| 1375 |
/// The initializer for the [ISO-8859-8](static.ISO_8859_8.html) encoding. |
| 1376 |
/// |
| 1377 |
/// For use only for taking the address of this form when |
| 1378 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1379 |
/// such as in initializers of other `static`s. If in doubt, |
| 1380 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1381 |
/// |
| 1382 |
/// This part of the public API will go away if Rust changes |
| 1383 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1384 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1385 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1386 |
/// items. |
| 1387 |
pub static ISO_8859_8_INIT: Encoding = Encoding { |
| 1388 |
name: "ISO-8859-8", |
| 1389 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_8, 0x05D0, 96, 27), |
| 1390 |
}; |
| 1391 |
|
| 1392 |
/// The ISO-8859-8 encoding. |
| 1393 |
/// |
| 1394 |
/// This is the Hebrew part of the ISO/IEC 8859 encoding family in visual order. |
| 1395 |
/// |
| 1396 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-8.html), |
| 1397 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-8-bmp.html) |
| 1398 |
/// |
| 1399 |
/// This encoding roughly matches the Windows code page 28598. Windows decodes |
| 1400 |
/// 0xAF to OVERLINE instead of MACRON and 0xFE and 0xFD to the Private Use |
| 1401 |
/// Area instead of LRM and RLM. Windows decodes unassigned code points to |
| 1402 |
/// the private use area. |
| 1403 |
/// |
| 1404 |
/// This will change from `static` to `const` if Rust changes |
| 1405 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1406 |
/// unique cross-crate, so don't take the address of this |
| 1407 |
/// `static`. |
| 1408 |
pub static ISO_8859_8: &'static Encoding = &ISO_8859_8_INIT; |
| 1409 |
|
| 1410 |
/// The initializer for the [ISO-8859-8-I](static.ISO_8859_8_I.html) encoding. |
| 1411 |
/// |
| 1412 |
/// For use only for taking the address of this form when |
| 1413 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1414 |
/// such as in initializers of other `static`s. If in doubt, |
| 1415 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1416 |
/// |
| 1417 |
/// This part of the public API will go away if Rust changes |
| 1418 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1419 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1420 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1421 |
/// items. |
| 1422 |
pub static ISO_8859_8_I_INIT: Encoding = Encoding { |
| 1423 |
name: "ISO-8859-8-I", |
| 1424 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.iso_8859_8, 0x05D0, 96, 27), |
| 1425 |
}; |
| 1426 |
|
| 1427 |
/// The ISO-8859-8-I encoding. |
| 1428 |
/// |
| 1429 |
/// This is the Hebrew part of the ISO/IEC 8859 encoding family in logical order. |
| 1430 |
/// |
| 1431 |
/// [Index visualization](https://encoding.spec.whatwg.org/iso-8859-8.html), |
| 1432 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/iso-8859-8-bmp.html) |
| 1433 |
/// |
| 1434 |
/// This encoding roughly matches the Windows code page 38598. Windows decodes |
| 1435 |
/// 0xAF to OVERLINE instead of MACRON and 0xFE and 0xFD to the Private Use |
| 1436 |
/// Area instead of LRM and RLM. Windows decodes unassigned code points to |
| 1437 |
/// the private use area. |
| 1438 |
/// |
| 1439 |
/// This will change from `static` to `const` if Rust changes |
| 1440 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1441 |
/// unique cross-crate, so don't take the address of this |
| 1442 |
/// `static`. |
| 1443 |
pub static ISO_8859_8_I: &'static Encoding = &ISO_8859_8_I_INIT; |
| 1444 |
|
| 1445 |
/// The initializer for the [KOI8-R](static.KOI8_R.html) encoding. |
| 1446 |
/// |
| 1447 |
/// For use only for taking the address of this form when |
| 1448 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1449 |
/// such as in initializers of other `static`s. If in doubt, |
| 1450 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1451 |
/// |
| 1452 |
/// This part of the public API will go away if Rust changes |
| 1453 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1454 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1455 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1456 |
/// items. |
| 1457 |
pub static KOI8_R_INIT: Encoding = Encoding { |
| 1458 |
name: "KOI8-R", |
| 1459 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.koi8_r, 0x044E, 64, 1), |
| 1460 |
}; |
| 1461 |
|
| 1462 |
/// The KOI8-R encoding. |
| 1463 |
/// |
| 1464 |
/// This is an encoding for Russian from [RFC 1489](https://tools.ietf.org/html/rfc1489). |
| 1465 |
/// |
| 1466 |
/// [Index visualization](https://encoding.spec.whatwg.org/koi8-r.html), |
| 1467 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/koi8-r-bmp.html) |
| 1468 |
/// |
| 1469 |
/// This encoding matches the Windows code page 20866. |
| 1470 |
/// |
| 1471 |
/// This will change from `static` to `const` if Rust changes |
| 1472 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1473 |
/// unique cross-crate, so don't take the address of this |
| 1474 |
/// `static`. |
| 1475 |
pub static KOI8_R: &'static Encoding = &KOI8_R_INIT; |
| 1476 |
|
| 1477 |
/// The initializer for the [KOI8-U](static.KOI8_U.html) encoding. |
| 1478 |
/// |
| 1479 |
/// For use only for taking the address of this form when |
| 1480 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1481 |
/// such as in initializers of other `static`s. If in doubt, |
| 1482 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1483 |
/// |
| 1484 |
/// This part of the public API will go away if Rust changes |
| 1485 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1486 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1487 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1488 |
/// items. |
| 1489 |
pub static KOI8_U_INIT: Encoding = Encoding { |
| 1490 |
name: "KOI8-U", |
| 1491 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.koi8_u, 0x044E, 64, 1), |
| 1492 |
}; |
| 1493 |
|
| 1494 |
/// The KOI8-U encoding. |
| 1495 |
/// |
| 1496 |
/// This is an encoding for Ukrainian adapted from KOI8-R. |
| 1497 |
/// |
| 1498 |
/// [Index visualization](https://encoding.spec.whatwg.org/koi8-u.html), |
| 1499 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/koi8-u-bmp.html) |
| 1500 |
/// |
| 1501 |
/// This encoding matches the Windows code page 21866. |
| 1502 |
/// |
| 1503 |
/// This will change from `static` to `const` if Rust changes |
| 1504 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1505 |
/// unique cross-crate, so don't take the address of this |
| 1506 |
/// `static`. |
| 1507 |
pub static KOI8_U: &'static Encoding = &KOI8_U_INIT; |
| 1508 |
|
| 1509 |
/// The initializer for the [Shift_JIS](static.SHIFT_JIS.html) encoding. |
| 1510 |
/// |
| 1511 |
/// For use only for taking the address of this form when |
| 1512 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1513 |
/// such as in initializers of other `static`s. If in doubt, |
| 1514 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1515 |
/// |
| 1516 |
/// This part of the public API will go away if Rust changes |
| 1517 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1518 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1519 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1520 |
/// items. |
| 1521 |
pub static SHIFT_JIS_INIT: Encoding = Encoding { |
| 1522 |
name: "Shift_JIS", |
| 1523 |
variant: VariantEncoding::ShiftJis, |
| 1524 |
}; |
| 1525 |
|
| 1526 |
/// The Shift_JIS encoding. |
| 1527 |
/// |
| 1528 |
/// This is the Japanese encoding for Windows. |
| 1529 |
/// |
| 1530 |
/// [Index visualization](https://encoding.spec.whatwg.org/shift_jis.html), |
| 1531 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/shift_jis-bmp.html) |
| 1532 |
/// |
| 1533 |
/// This encoding matches the Windows code page 932, except Windows decodes some byte |
| 1534 |
/// sequences that are error per the Encoding Standard to the question mark or the |
| 1535 |
/// Private Use Area and generally uses U+30FB in place of the REPLACEMENT CHARACTER. |
| 1536 |
/// |
| 1537 |
/// This will change from `static` to `const` if Rust changes |
| 1538 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1539 |
/// unique cross-crate, so don't take the address of this |
| 1540 |
/// `static`. |
| 1541 |
pub static SHIFT_JIS: &'static Encoding = &SHIFT_JIS_INIT; |
| 1542 |
|
| 1543 |
/// The initializer for the [UTF-16BE](static.UTF_16BE.html) encoding. |
| 1544 |
/// |
| 1545 |
/// For use only for taking the address of this form when |
| 1546 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1547 |
/// such as in initializers of other `static`s. If in doubt, |
| 1548 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1549 |
/// |
| 1550 |
/// This part of the public API will go away if Rust changes |
| 1551 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1552 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1553 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1554 |
/// items. |
| 1555 |
pub static UTF_16BE_INIT: Encoding = Encoding { |
| 1556 |
name: "UTF-16BE", |
| 1557 |
variant: VariantEncoding::Utf16Be, |
| 1558 |
}; |
| 1559 |
|
| 1560 |
/// The UTF-16BE encoding. |
| 1561 |
/// |
| 1562 |
/// This decode-only encoding uses 16-bit code units due to Unicode originally |
| 1563 |
/// having been designed as a 16-bit reportoire. In the absence of a byte order |
| 1564 |
/// mark the big endian byte order is assumed. |
| 1565 |
/// |
| 1566 |
/// There is no corresponding encoder in this crate or in the Encoding |
| 1567 |
/// Standard. The output encoding of this encoding is UTF-8. |
| 1568 |
/// |
| 1569 |
/// This encoding matches the Windows code page 1201. |
| 1570 |
/// |
| 1571 |
/// This will change from `static` to `const` if Rust changes |
| 1572 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1573 |
/// unique cross-crate, so don't take the address of this |
| 1574 |
/// `static`. |
| 1575 |
pub static UTF_16BE: &'static Encoding = &UTF_16BE_INIT; |
| 1576 |
|
| 1577 |
/// The initializer for the [UTF-16LE](static.UTF_16LE.html) encoding. |
| 1578 |
/// |
| 1579 |
/// For use only for taking the address of this form when |
| 1580 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1581 |
/// such as in initializers of other `static`s. If in doubt, |
| 1582 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1583 |
/// |
| 1584 |
/// This part of the public API will go away if Rust changes |
| 1585 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1586 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1587 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1588 |
/// items. |
| 1589 |
pub static UTF_16LE_INIT: Encoding = Encoding { |
| 1590 |
name: "UTF-16LE", |
| 1591 |
variant: VariantEncoding::Utf16Le, |
| 1592 |
}; |
| 1593 |
|
| 1594 |
/// The UTF-16LE encoding. |
| 1595 |
/// |
| 1596 |
/// This decode-only encoding uses 16-bit code units due to Unicode originally |
| 1597 |
/// having been designed as a 16-bit reportoire. In the absence of a byte order |
| 1598 |
/// mark the little endian byte order is assumed. |
| 1599 |
/// |
| 1600 |
/// There is no corresponding encoder in this crate or in the Encoding |
| 1601 |
/// Standard. The output encoding of this encoding is UTF-8. |
| 1602 |
/// |
| 1603 |
/// This encoding matches the Windows code page 1200. |
| 1604 |
/// |
| 1605 |
/// This will change from `static` to `const` if Rust changes |
| 1606 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1607 |
/// unique cross-crate, so don't take the address of this |
| 1608 |
/// `static`. |
| 1609 |
pub static UTF_16LE: &'static Encoding = &UTF_16LE_INIT; |
| 1610 |
|
| 1611 |
/// The initializer for the [UTF-8](static.UTF_8.html) encoding. |
| 1612 |
/// |
| 1613 |
/// For use only for taking the address of this form when |
| 1614 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1615 |
/// such as in initializers of other `static`s. If in doubt, |
| 1616 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1617 |
/// |
| 1618 |
/// This part of the public API will go away if Rust changes |
| 1619 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1620 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1621 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1622 |
/// items. |
| 1623 |
pub static UTF_8_INIT: Encoding = Encoding { |
| 1624 |
name: "UTF-8", |
| 1625 |
variant: VariantEncoding::Utf8, |
| 1626 |
}; |
| 1627 |
|
| 1628 |
/// The UTF-8 encoding. |
| 1629 |
/// |
| 1630 |
/// This is the encoding that should be used for all new development it can |
| 1631 |
/// represent all of Unicode. |
| 1632 |
/// |
| 1633 |
/// This encoding matches the Windows code page 65001, except Windows differs |
| 1634 |
/// in the number of errors generated for some erroneous byte sequences. |
| 1635 |
/// |
| 1636 |
/// This will change from `static` to `const` if Rust changes |
| 1637 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1638 |
/// unique cross-crate, so don't take the address of this |
| 1639 |
/// `static`. |
| 1640 |
pub static UTF_8: &'static Encoding = &UTF_8_INIT; |
| 1641 |
|
| 1642 |
/// The initializer for the [gb18030](static.GB18030.html) encoding. |
| 1643 |
/// |
| 1644 |
/// For use only for taking the address of this form when |
| 1645 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1646 |
/// such as in initializers of other `static`s. If in doubt, |
| 1647 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1648 |
/// |
| 1649 |
/// This part of the public API will go away if Rust changes |
| 1650 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1651 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1652 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1653 |
/// items. |
| 1654 |
pub static GB18030_INIT: Encoding = Encoding { |
| 1655 |
name: "gb18030", |
| 1656 |
variant: VariantEncoding::Gb18030, |
| 1657 |
}; |
| 1658 |
|
| 1659 |
/// The gb18030 encoding. |
| 1660 |
/// |
| 1661 |
/// This encoding matches GB18030-2005 except the two-byte sequence 0xA3 0xA0 |
| 1662 |
/// maps to U+3000 for compatibility with existing Web content. As a result, |
| 1663 |
/// this encoding can represent all of Unicode except for the private-use |
| 1664 |
/// character U+E5E5. |
| 1665 |
/// |
| 1666 |
/// [Index visualization for the two-byte sequences](https://encoding.spec.whatwg.org/gb18030.html), |
| 1667 |
/// [Visualization of BMP coverage of the two-byte index](https://encoding.spec.whatwg.org/gb18030-bmp.html) |
| 1668 |
/// |
| 1669 |
/// This encoding matches the Windows code page 54936. |
| 1670 |
/// |
| 1671 |
/// This will change from `static` to `const` if Rust changes |
| 1672 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1673 |
/// unique cross-crate, so don't take the address of this |
| 1674 |
/// `static`. |
| 1675 |
pub static GB18030: &'static Encoding = &GB18030_INIT; |
| 1676 |
|
| 1677 |
/// The initializer for the [macintosh](static.MACINTOSH.html) encoding. |
| 1678 |
/// |
| 1679 |
/// For use only for taking the address of this form when |
| 1680 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1681 |
/// such as in initializers of other `static`s. If in doubt, |
| 1682 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1683 |
/// |
| 1684 |
/// This part of the public API will go away if Rust changes |
| 1685 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1686 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1687 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1688 |
/// items. |
| 1689 |
pub static MACINTOSH_INIT: Encoding = Encoding { |
| 1690 |
name: "macintosh", |
| 1691 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.macintosh, 0x00CD, 106, 3), |
| 1692 |
}; |
| 1693 |
|
| 1694 |
/// The macintosh encoding. |
| 1695 |
/// |
| 1696 |
/// This is the MacRoman encoding from Mac OS Classic. |
| 1697 |
/// |
| 1698 |
/// [Index visualization](https://encoding.spec.whatwg.org/macintosh.html), |
| 1699 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/macintosh-bmp.html) |
| 1700 |
/// |
| 1701 |
/// This encoding matches the Windows code page 10000, except Windows decodes |
| 1702 |
/// 0xBD to U+2126 OHM SIGN instead of U+03A9 GREEK CAPITAL LETTER OMEGA. |
| 1703 |
/// |
| 1704 |
/// This will change from `static` to `const` if Rust changes |
| 1705 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1706 |
/// unique cross-crate, so don't take the address of this |
| 1707 |
/// `static`. |
| 1708 |
pub static MACINTOSH: &'static Encoding = &MACINTOSH_INIT; |
| 1709 |
|
| 1710 |
/// The initializer for the [replacement](static.REPLACEMENT.html) encoding. |
| 1711 |
/// |
| 1712 |
/// For use only for taking the address of this form when |
| 1713 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1714 |
/// such as in initializers of other `static`s. If in doubt, |
| 1715 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1716 |
/// |
| 1717 |
/// This part of the public API will go away if Rust changes |
| 1718 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1719 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1720 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1721 |
/// items. |
| 1722 |
pub static REPLACEMENT_INIT: Encoding = Encoding { |
| 1723 |
name: "replacement", |
| 1724 |
variant: VariantEncoding::Replacement, |
| 1725 |
}; |
| 1726 |
|
| 1727 |
/// The replacement encoding. |
| 1728 |
/// |
| 1729 |
/// This decode-only encoding decodes all non-zero-length streams to a single |
| 1730 |
/// REPLACEMENT CHARACTER. Its purpose is to avoid the use of an |
| 1731 |
/// ASCII-compatible fallback encoding (typically windows-1252) for some |
| 1732 |
/// encodings that are no longer supported by the Web Platform and that |
| 1733 |
/// would be dangerous to treat as ASCII-compatible. |
| 1734 |
/// |
| 1735 |
/// There is no corresponding encoder. The output encoding of this encoding |
| 1736 |
/// is UTF-8. |
| 1737 |
/// |
| 1738 |
/// This encoding does not have a Windows code page number. |
| 1739 |
/// |
| 1740 |
/// This will change from `static` to `const` if Rust changes |
| 1741 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1742 |
/// unique cross-crate, so don't take the address of this |
| 1743 |
/// `static`. |
| 1744 |
pub static REPLACEMENT: &'static Encoding = &REPLACEMENT_INIT; |
| 1745 |
|
| 1746 |
/// The initializer for the [windows-1250](static.WINDOWS_1250.html) encoding. |
| 1747 |
/// |
| 1748 |
/// For use only for taking the address of this form when |
| 1749 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1750 |
/// such as in initializers of other `static`s. If in doubt, |
| 1751 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1752 |
/// |
| 1753 |
/// This part of the public API will go away if Rust changes |
| 1754 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1755 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1756 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1757 |
/// items. |
| 1758 |
pub static WINDOWS_1250_INIT: Encoding = Encoding { |
| 1759 |
name: "windows-1250", |
| 1760 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.windows_1250, 0x00DC, 92, 2), |
| 1761 |
}; |
| 1762 |
|
| 1763 |
/// The windows-1250 encoding. |
| 1764 |
/// |
| 1765 |
/// This is the Central European encoding for Windows. |
| 1766 |
/// |
| 1767 |
/// [Index visualization](https://encoding.spec.whatwg.org/windows-1250.html), |
| 1768 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/windows-1250-bmp.html) |
| 1769 |
/// |
| 1770 |
/// This encoding matches the Windows code page 1250. |
| 1771 |
/// |
| 1772 |
/// This will change from `static` to `const` if Rust changes |
| 1773 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1774 |
/// unique cross-crate, so don't take the address of this |
| 1775 |
/// `static`. |
| 1776 |
pub static WINDOWS_1250: &'static Encoding = &WINDOWS_1250_INIT; |
| 1777 |
|
| 1778 |
/// The initializer for the [windows-1251](static.WINDOWS_1251.html) encoding. |
| 1779 |
/// |
| 1780 |
/// For use only for taking the address of this form when |
| 1781 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1782 |
/// such as in initializers of other `static`s. If in doubt, |
| 1783 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1784 |
/// |
| 1785 |
/// This part of the public API will go away if Rust changes |
| 1786 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1787 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1788 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1789 |
/// items. |
| 1790 |
pub static WINDOWS_1251_INIT: Encoding = Encoding { |
| 1791 |
name: "windows-1251", |
| 1792 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.windows_1251, 0x0410, 64, 64), |
| 1793 |
}; |
| 1794 |
|
| 1795 |
/// The windows-1251 encoding. |
| 1796 |
/// |
| 1797 |
/// This is the Cyrillic encoding for Windows. |
| 1798 |
/// |
| 1799 |
/// [Index visualization](https://encoding.spec.whatwg.org/windows-1251.html), |
| 1800 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/windows-1251-bmp.html) |
| 1801 |
/// |
| 1802 |
/// This encoding matches the Windows code page 1251. |
| 1803 |
/// |
| 1804 |
/// This will change from `static` to `const` if Rust changes |
| 1805 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1806 |
/// unique cross-crate, so don't take the address of this |
| 1807 |
/// `static`. |
| 1808 |
pub static WINDOWS_1251: &'static Encoding = &WINDOWS_1251_INIT; |
| 1809 |
|
| 1810 |
/// The initializer for the [windows-1252](static.WINDOWS_1252.html) encoding. |
| 1811 |
/// |
| 1812 |
/// For use only for taking the address of this form when |
| 1813 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1814 |
/// such as in initializers of other `static`s. If in doubt, |
| 1815 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1816 |
/// |
| 1817 |
/// This part of the public API will go away if Rust changes |
| 1818 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1819 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1820 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1821 |
/// items. |
| 1822 |
pub static WINDOWS_1252_INIT: Encoding = Encoding { |
| 1823 |
name: "windows-1252", |
| 1824 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.windows_1252, 0x00A0, 32, 96), |
| 1825 |
}; |
| 1826 |
|
| 1827 |
/// The windows-1252 encoding. |
| 1828 |
/// |
| 1829 |
/// This is the Western encoding for Windows. It is an extension of ISO-8859-1, |
| 1830 |
/// which is known as Latin 1. |
| 1831 |
/// |
| 1832 |
/// [Index visualization](https://encoding.spec.whatwg.org/windows-1252.html), |
| 1833 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/windows-1252-bmp.html) |
| 1834 |
/// |
| 1835 |
/// This encoding matches the Windows code page 1252. |
| 1836 |
/// |
| 1837 |
/// This will change from `static` to `const` if Rust changes |
| 1838 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1839 |
/// unique cross-crate, so don't take the address of this |
| 1840 |
/// `static`. |
| 1841 |
pub static WINDOWS_1252: &'static Encoding = &WINDOWS_1252_INIT; |
| 1842 |
|
| 1843 |
/// The initializer for the [windows-1253](static.WINDOWS_1253.html) encoding. |
| 1844 |
/// |
| 1845 |
/// For use only for taking the address of this form when |
| 1846 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1847 |
/// such as in initializers of other `static`s. If in doubt, |
| 1848 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1849 |
/// |
| 1850 |
/// This part of the public API will go away if Rust changes |
| 1851 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1852 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1853 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1854 |
/// items. |
| 1855 |
pub static WINDOWS_1253_INIT: Encoding = Encoding { |
| 1856 |
name: "windows-1253", |
| 1857 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.windows_1253, 0x03A3, 83, 44), |
| 1858 |
}; |
| 1859 |
|
| 1860 |
/// The windows-1253 encoding. |
| 1861 |
/// |
| 1862 |
/// This is the Greek encoding for Windows. It is mostly an extension of |
| 1863 |
/// ISO-8859-7, but U+0386 is mapped to a different byte. |
| 1864 |
/// |
| 1865 |
/// [Index visualization](https://encoding.spec.whatwg.org/windows-1253.html), |
| 1866 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/windows-1253-bmp.html) |
| 1867 |
/// |
| 1868 |
/// This encoding matches the Windows code page 1253, except Windows decodes |
| 1869 |
/// unassigned code points to the Private Use Area of Unicode. |
| 1870 |
/// |
| 1871 |
/// This will change from `static` to `const` if Rust changes |
| 1872 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1873 |
/// unique cross-crate, so don't take the address of this |
| 1874 |
/// `static`. |
| 1875 |
pub static WINDOWS_1253: &'static Encoding = &WINDOWS_1253_INIT; |
| 1876 |
|
| 1877 |
/// The initializer for the [windows-1254](static.WINDOWS_1254.html) encoding. |
| 1878 |
/// |
| 1879 |
/// For use only for taking the address of this form when |
| 1880 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1881 |
/// such as in initializers of other `static`s. If in doubt, |
| 1882 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1883 |
/// |
| 1884 |
/// This part of the public API will go away if Rust changes |
| 1885 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1886 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1887 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1888 |
/// items. |
| 1889 |
pub static WINDOWS_1254_INIT: Encoding = Encoding { |
| 1890 |
name: "windows-1254", |
| 1891 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.windows_1254, 0x00DF, 95, 17), |
| 1892 |
}; |
| 1893 |
|
| 1894 |
/// The windows-1254 encoding. |
| 1895 |
/// |
| 1896 |
/// This is the Turkish encoding for Windows. It is an extension of ISO-8859-9, |
| 1897 |
/// which is known as Latin 5. |
| 1898 |
/// |
| 1899 |
/// [Index visualization](https://encoding.spec.whatwg.org/windows-1254.html), |
| 1900 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/windows-1254-bmp.html) |
| 1901 |
/// |
| 1902 |
/// This encoding matches the Windows code page 1254. |
| 1903 |
/// |
| 1904 |
/// This will change from `static` to `const` if Rust changes |
| 1905 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1906 |
/// unique cross-crate, so don't take the address of this |
| 1907 |
/// `static`. |
| 1908 |
pub static WINDOWS_1254: &'static Encoding = &WINDOWS_1254_INIT; |
| 1909 |
|
| 1910 |
/// The initializer for the [windows-1255](static.WINDOWS_1255.html) encoding. |
| 1911 |
/// |
| 1912 |
/// For use only for taking the address of this form when |
| 1913 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1914 |
/// such as in initializers of other `static`s. If in doubt, |
| 1915 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1916 |
/// |
| 1917 |
/// This part of the public API will go away if Rust changes |
| 1918 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1919 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1920 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1921 |
/// items. |
| 1922 |
pub static WINDOWS_1255_INIT: Encoding = Encoding { |
| 1923 |
name: "windows-1255", |
| 1924 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.windows_1255, 0x05D0, 96, 27), |
| 1925 |
}; |
| 1926 |
|
| 1927 |
/// The windows-1255 encoding. |
| 1928 |
/// |
| 1929 |
/// This is the Hebrew encoding for Windows. It is an extension of ISO-8859-8-I, |
| 1930 |
/// except for a currency sign swap. |
| 1931 |
/// |
| 1932 |
/// [Index visualization](https://encoding.spec.whatwg.org/windows-1255.html), |
| 1933 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/windows-1255-bmp.html) |
| 1934 |
/// |
| 1935 |
/// This encoding matches the Windows code page 1255, except Windows decodes |
| 1936 |
/// unassigned code points to the Private Use Area of Unicode. |
| 1937 |
/// |
| 1938 |
/// This will change from `static` to `const` if Rust changes |
| 1939 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1940 |
/// unique cross-crate, so don't take the address of this |
| 1941 |
/// `static`. |
| 1942 |
pub static WINDOWS_1255: &'static Encoding = &WINDOWS_1255_INIT; |
| 1943 |
|
| 1944 |
/// The initializer for the [windows-1256](static.WINDOWS_1256.html) encoding. |
| 1945 |
/// |
| 1946 |
/// For use only for taking the address of this form when |
| 1947 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1948 |
/// such as in initializers of other `static`s. If in doubt, |
| 1949 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1950 |
/// |
| 1951 |
/// This part of the public API will go away if Rust changes |
| 1952 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1953 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1954 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1955 |
/// items. |
| 1956 |
pub static WINDOWS_1256_INIT: Encoding = Encoding { |
| 1957 |
name: "windows-1256", |
| 1958 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.windows_1256, 0x0621, 65, 22), |
| 1959 |
}; |
| 1960 |
|
| 1961 |
/// The windows-1256 encoding. |
| 1962 |
/// |
| 1963 |
/// This is the Arabic encoding for Windows. |
| 1964 |
/// |
| 1965 |
/// [Index visualization](https://encoding.spec.whatwg.org/windows-1256.html), |
| 1966 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/windows-1256-bmp.html) |
| 1967 |
/// |
| 1968 |
/// This encoding matches the Windows code page 1256. |
| 1969 |
/// |
| 1970 |
/// This will change from `static` to `const` if Rust changes |
| 1971 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1972 |
/// unique cross-crate, so don't take the address of this |
| 1973 |
/// `static`. |
| 1974 |
pub static WINDOWS_1256: &'static Encoding = &WINDOWS_1256_INIT; |
| 1975 |
|
| 1976 |
/// The initializer for the [windows-1257](static.WINDOWS_1257.html) encoding. |
| 1977 |
/// |
| 1978 |
/// For use only for taking the address of this form when |
| 1979 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 1980 |
/// such as in initializers of other `static`s. If in doubt, |
| 1981 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 1982 |
/// |
| 1983 |
/// This part of the public API will go away if Rust changes |
| 1984 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 1985 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 1986 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 1987 |
/// items. |
| 1988 |
pub static WINDOWS_1257_INIT: Encoding = Encoding { |
| 1989 |
name: "windows-1257", |
| 1990 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.windows_1257, 0x00DF, 95, 1), |
| 1991 |
}; |
| 1992 |
|
| 1993 |
/// The windows-1257 encoding. |
| 1994 |
/// |
| 1995 |
/// This is the Baltic encoding for Windows. |
| 1996 |
/// |
| 1997 |
/// [Index visualization](https://encoding.spec.whatwg.org/windows-1257.html), |
| 1998 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/windows-1257-bmp.html) |
| 1999 |
/// |
| 2000 |
/// This encoding matches the Windows code page 1257, except Windows decodes |
| 2001 |
/// unassigned code points to the Private Use Area of Unicode. |
| 2002 |
/// |
| 2003 |
/// This will change from `static` to `const` if Rust changes |
| 2004 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 2005 |
/// unique cross-crate, so don't take the address of this |
| 2006 |
/// `static`. |
| 2007 |
pub static WINDOWS_1257: &'static Encoding = &WINDOWS_1257_INIT; |
| 2008 |
|
| 2009 |
/// The initializer for the [windows-1258](static.WINDOWS_1258.html) encoding. |
| 2010 |
/// |
| 2011 |
/// For use only for taking the address of this form when |
| 2012 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 2013 |
/// such as in initializers of other `static`s. If in doubt, |
| 2014 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 2015 |
/// |
| 2016 |
/// This part of the public API will go away if Rust changes |
| 2017 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 2018 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 2019 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 2020 |
/// items. |
| 2021 |
pub static WINDOWS_1258_INIT: Encoding = Encoding { |
| 2022 |
name: "windows-1258", |
| 2023 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.windows_1258, 0x00DF, 95, 4), |
| 2024 |
}; |
| 2025 |
|
| 2026 |
/// The windows-1258 encoding. |
| 2027 |
/// |
| 2028 |
/// This is the Vietnamese encoding for Windows. |
| 2029 |
/// |
| 2030 |
/// [Index visualization](https://encoding.spec.whatwg.org/windows-1258.html), |
| 2031 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/windows-1258-bmp.html) |
| 2032 |
/// |
| 2033 |
/// This encoding matches the Windows code page 1258 when used in the |
| 2034 |
/// non-normalizing mode. Unlike with the other single-byte encodings, the |
| 2035 |
/// result of decoding is not necessarily in Normalization Form C. On the |
| 2036 |
/// other hand, input in the Normalization Form C is not encoded without |
| 2037 |
/// replacement. In general, it's a bad idea to encode to encodings other |
| 2038 |
/// than UTF-8, but this encoding is especially hazardous to encode to. |
| 2039 |
/// |
| 2040 |
/// This will change from `static` to `const` if Rust changes |
| 2041 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 2042 |
/// unique cross-crate, so don't take the address of this |
| 2043 |
/// `static`. |
| 2044 |
pub static WINDOWS_1258: &'static Encoding = &WINDOWS_1258_INIT; |
| 2045 |
|
| 2046 |
/// The initializer for the [windows-874](static.WINDOWS_874.html) encoding. |
| 2047 |
/// |
| 2048 |
/// For use only for taking the address of this form when |
| 2049 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 2050 |
/// such as in initializers of other `static`s. If in doubt, |
| 2051 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 2052 |
/// |
| 2053 |
/// This part of the public API will go away if Rust changes |
| 2054 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 2055 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 2056 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 2057 |
/// items. |
| 2058 |
pub static WINDOWS_874_INIT: Encoding = Encoding { |
| 2059 |
name: "windows-874", |
| 2060 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.windows_874, 0x0E01, 33, 58), |
| 2061 |
}; |
| 2062 |
|
| 2063 |
/// The windows-874 encoding. |
| 2064 |
/// |
| 2065 |
/// This is the Thai encoding for Windows. It is an extension of TIS-620 / ISO-8859-11. |
| 2066 |
/// |
| 2067 |
/// [Index visualization](https://encoding.spec.whatwg.org/windows-874.html), |
| 2068 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/windows-874-bmp.html) |
| 2069 |
/// |
| 2070 |
/// This encoding matches the Windows code page 874, except Windows decodes |
| 2071 |
/// unassigned code points to the Private Use Area of Unicode. |
| 2072 |
/// |
| 2073 |
/// This will change from `static` to `const` if Rust changes |
| 2074 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 2075 |
/// unique cross-crate, so don't take the address of this |
| 2076 |
/// `static`. |
| 2077 |
pub static WINDOWS_874: &'static Encoding = &WINDOWS_874_INIT; |
| 2078 |
|
| 2079 |
/// The initializer for the [x-mac-cyrillic](static.X_MAC_CYRILLIC.html) encoding. |
| 2080 |
/// |
| 2081 |
/// For use only for taking the address of this form when |
| 2082 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 2083 |
/// such as in initializers of other `static`s. If in doubt, |
| 2084 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 2085 |
/// |
| 2086 |
/// This part of the public API will go away if Rust changes |
| 2087 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 2088 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 2089 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 2090 |
/// items. |
| 2091 |
pub static X_MAC_CYRILLIC_INIT: Encoding = Encoding { |
| 2092 |
name: "x-mac-cyrillic", |
| 2093 |
variant: VariantEncoding::SingleByte(&data::SINGLE_BYTE_DATA.x_mac_cyrillic, 0x0430, 96, 31), |
| 2094 |
}; |
| 2095 |
|
| 2096 |
/// The x-mac-cyrillic encoding. |
| 2097 |
/// |
| 2098 |
/// This is the MacUkrainian encoding from Mac OS Classic. |
| 2099 |
/// |
| 2100 |
/// [Index visualization](https://encoding.spec.whatwg.org/x-mac-cyrillic.html), |
| 2101 |
/// [Visualization of BMP coverage](https://encoding.spec.whatwg.org/x-mac-cyrillic-bmp.html) |
| 2102 |
/// |
| 2103 |
/// This encoding matches the Windows code page 10017. |
| 2104 |
/// |
| 2105 |
/// This will change from `static` to `const` if Rust changes |
| 2106 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 2107 |
/// unique cross-crate, so don't take the address of this |
| 2108 |
/// `static`. |
| 2109 |
pub static X_MAC_CYRILLIC: &'static Encoding = &X_MAC_CYRILLIC_INIT; |
| 2110 |
|
| 2111 |
/// The initializer for the [x-user-defined](static.X_USER_DEFINED.html) encoding. |
| 2112 |
/// |
| 2113 |
/// For use only for taking the address of this form when |
| 2114 |
/// Rust prohibits the use of the non-`_INIT` form directly, |
| 2115 |
/// such as in initializers of other `static`s. If in doubt, |
| 2116 |
/// use the corresponding non-`_INIT` reference-typed `static`. |
| 2117 |
/// |
| 2118 |
/// This part of the public API will go away if Rust changes |
| 2119 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 2120 |
/// unique cross-crate or if Rust starts allowing static arrays |
| 2121 |
/// to be initialized with `pub static FOO: &'static Encoding` |
| 2122 |
/// items. |
| 2123 |
pub static X_USER_DEFINED_INIT: Encoding = Encoding { |
| 2124 |
name: "x-user-defined", |
| 2125 |
variant: VariantEncoding::UserDefined, |
| 2126 |
}; |
| 2127 |
|
| 2128 |
/// The x-user-defined encoding. |
| 2129 |
/// |
| 2130 |
/// This encoding offsets the non-ASCII bytes by `0xF700` thereby decoding |
| 2131 |
/// them to the Private Use Area of Unicode. It was used for loading binary |
| 2132 |
/// data into a JavaScript string using `XMLHttpRequest` before XHR supported |
| 2133 |
/// the `"arraybuffer"` response type. |
| 2134 |
/// |
| 2135 |
/// This encoding does not have a Windows code page number. |
| 2136 |
/// |
| 2137 |
/// This will change from `static` to `const` if Rust changes |
| 2138 |
/// to make the referent of `pub const FOO: &'static Encoding` |
| 2139 |
/// unique cross-crate, so don't take the address of this |
| 2140 |
/// `static`. |
| 2141 |
pub static X_USER_DEFINED: &'static Encoding = &X_USER_DEFINED_INIT; |
| 2142 |
|
| 2143 |
static LABELS_SORTED: [&'static str; 228] = [ |
| 2144 |
"l1", |
| 2145 |
"l2", |
| 2146 |
"l3", |
| 2147 |
"l4", |
| 2148 |
"l5", |
| 2149 |
"l6", |
| 2150 |
"l9", |
| 2151 |
"866", |
| 2152 |
"mac", |
| 2153 |
"koi", |
| 2154 |
"gbk", |
| 2155 |
"big5", |
| 2156 |
"utf8", |
| 2157 |
"koi8", |
| 2158 |
"sjis", |
| 2159 |
"ucs-2", |
| 2160 |
"ms932", |
| 2161 |
"cp866", |
| 2162 |
"utf-8", |
| 2163 |
"cp819", |
| 2164 |
"ascii", |
| 2165 |
"x-gbk", |
| 2166 |
"greek", |
| 2167 |
"cp1250", |
| 2168 |
"cp1251", |
| 2169 |
"latin1", |
| 2170 |
"gb2312", |
| 2171 |
"cp1252", |
| 2172 |
"latin2", |
| 2173 |
"cp1253", |
| 2174 |
"latin3", |
| 2175 |
"cp1254", |
| 2176 |
"latin4", |
| 2177 |
"cp1255", |
| 2178 |
"csbig5", |
| 2179 |
"latin5", |
| 2180 |
"utf-16", |
| 2181 |
"cp1256", |
| 2182 |
"ibm866", |
| 2183 |
"latin6", |
| 2184 |
"cp1257", |
| 2185 |
"cp1258", |
| 2186 |
"greek8", |
| 2187 |
"ibm819", |
| 2188 |
"arabic", |
| 2189 |
"visual", |
| 2190 |
"korean", |
| 2191 |
"euc-jp", |
| 2192 |
"koi8-r", |
| 2193 |
"koi8_r", |
| 2194 |
"euc-kr", |
| 2195 |
"x-sjis", |
| 2196 |
"koi8-u", |
| 2197 |
"hebrew", |
| 2198 |
"tis-620", |
| 2199 |
"gb18030", |
| 2200 |
"ksc5601", |
| 2201 |
"gb_2312", |
| 2202 |
"dos-874", |
| 2203 |
"cn-big5", |
| 2204 |
"unicode", |
| 2205 |
"chinese", |
| 2206 |
"logical", |
| 2207 |
"cskoi8r", |
| 2208 |
"cseuckr", |
| 2209 |
"koi8-ru", |
| 2210 |
"x-cp1250", |
| 2211 |
"ksc_5601", |
| 2212 |
"x-cp1251", |
| 2213 |
"iso88591", |
| 2214 |
"csgb2312", |
| 2215 |
"x-cp1252", |
| 2216 |
"iso88592", |
| 2217 |
"x-cp1253", |
| 2218 |
"iso88593", |
| 2219 |
"ecma-114", |
| 2220 |
"x-cp1254", |
| 2221 |
"iso88594", |
| 2222 |
"x-cp1255", |
| 2223 |
"iso88595", |
| 2224 |
"x-x-big5", |
| 2225 |
"x-cp1256", |
| 2226 |
"csibm866", |
| 2227 |
"iso88596", |
| 2228 |
"x-cp1257", |
| 2229 |
"iso88597", |
| 2230 |
"asmo-708", |
| 2231 |
"ecma-118", |
| 2232 |
"elot_928", |
| 2233 |
"x-cp1258", |
| 2234 |
"iso88598", |
| 2235 |
"iso88599", |
| 2236 |
"cyrillic", |
| 2237 |
"utf-16be", |
| 2238 |
"utf-16le", |
| 2239 |
"us-ascii", |
| 2240 |
"ms_kanji", |
| 2241 |
"x-euc-jp", |
| 2242 |
"iso885910", |
| 2243 |
"iso8859-1", |
| 2244 |
"iso885911", |
| 2245 |
"iso8859-2", |
| 2246 |
"iso8859-3", |
| 2247 |
"iso885913", |
| 2248 |
"iso8859-4", |
| 2249 |
"iso885914", |
| 2250 |
"iso8859-5", |
| 2251 |
"iso885915", |
| 2252 |
"iso8859-6", |
| 2253 |
"iso8859-7", |
| 2254 |
"iso8859-8", |
| 2255 |
"iso-ir-58", |
| 2256 |
"iso8859-9", |
| 2257 |
"csunicode", |
| 2258 |
"macintosh", |
| 2259 |
"shift-jis", |
| 2260 |
"shift_jis", |
| 2261 |
"iso-ir-100", |
| 2262 |
"iso8859-10", |
| 2263 |
"iso-ir-110", |
| 2264 |
"gb_2312-80", |
| 2265 |
"iso-8859-1", |
| 2266 |
"iso_8859-1", |
| 2267 |
"iso-ir-101", |
| 2268 |
"iso8859-11", |
| 2269 |
"iso-8859-2", |
| 2270 |
"iso_8859-2", |
| 2271 |
"hz-gb-2312", |
| 2272 |
"iso-8859-3", |
| 2273 |
"iso_8859-3", |
| 2274 |
"iso8859-13", |
| 2275 |
"iso-8859-4", |
| 2276 |
"iso_8859-4", |
| 2277 |
"iso8859-14", |
| 2278 |
"iso-ir-144", |
| 2279 |
"iso-8859-5", |
| 2280 |
"iso_8859-5", |
| 2281 |
"iso8859-15", |
| 2282 |
"iso-8859-6", |
| 2283 |
"iso_8859-6", |
| 2284 |
"iso-ir-126", |
| 2285 |
"iso-8859-7", |
| 2286 |
"iso_8859-7", |
| 2287 |
"iso-ir-127", |
| 2288 |
"iso-ir-157", |
| 2289 |
"iso-8859-8", |
| 2290 |
"iso_8859-8", |
| 2291 |
"iso-ir-138", |
| 2292 |
"iso-ir-148", |
| 2293 |
"iso-8859-9", |
| 2294 |
"iso_8859-9", |
| 2295 |
"iso-ir-109", |
| 2296 |
"iso-ir-149", |
| 2297 |
"big5-hkscs", |
| 2298 |
"csshiftjis", |
| 2299 |
"iso-8859-10", |
| 2300 |
"iso-8859-11", |
| 2301 |
"csisolatin1", |
| 2302 |
"csisolatin2", |
| 2303 |
"iso-8859-13", |
| 2304 |
"csisolatin3", |
| 2305 |
"iso-8859-14", |
| 2306 |
"windows-874", |
| 2307 |
"csisolatin4", |
| 2308 |
"iso-8859-15", |
| 2309 |
"iso_8859-15", |
| 2310 |
"csisolatin5", |
| 2311 |
"iso-8859-16", |
| 2312 |
"csisolatin6", |
| 2313 |
"windows-949", |
| 2314 |
"csisolatin9", |
| 2315 |
"csiso88596e", |
| 2316 |
"csiso88598e", |
| 2317 |
"unicodefffe", |
| 2318 |
"unicodefeff", |
| 2319 |
"csmacintosh", |
| 2320 |
"csiso88596i", |
| 2321 |
"csiso88598i", |
| 2322 |
"windows-31j", |
| 2323 |
"x-mac-roman", |
| 2324 |
"iso-2022-cn", |
| 2325 |
"iso-2022-jp", |
| 2326 |
"csiso2022jp", |
| 2327 |
"iso-2022-kr", |
| 2328 |
"csiso2022kr", |
| 2329 |
"replacement", |
| 2330 |
"windows-1250", |
| 2331 |
"windows-1251", |
| 2332 |
"windows-1252", |
| 2333 |
"windows-1253", |
| 2334 |
"windows-1254", |
| 2335 |
"windows-1255", |
| 2336 |
"windows-1256", |
| 2337 |
"windows-1257", |
| 2338 |
"windows-1258", |
| 2339 |
"iso-8859-6-e", |
| 2340 |
"iso-8859-8-e", |
| 2341 |
"iso-8859-6-i", |
| 2342 |
"iso-8859-8-i", |
| 2343 |
"sun_eu_greek", |
| 2344 |
"csksc56011987", |
| 2345 |
"unicode20utf8", |
| 2346 |
"unicode11utf8", |
| 2347 |
"ks_c_5601-1987", |
| 2348 |
"ansi_x3.4-1968", |
| 2349 |
"ks_c_5601-1989", |
| 2350 |
"x-mac-cyrillic", |
| 2351 |
"x-user-defined", |
| 2352 |
"csiso58gb231280", |
| 2353 |
"iso-10646-ucs-2", |
| 2354 |
"iso_8859-1:1987", |
| 2355 |
"iso_8859-2:1987", |
| 2356 |
"iso_8859-6:1987", |
| 2357 |
"iso_8859-7:1987", |
| 2358 |
"iso_8859-3:1988", |
| 2359 |
"iso_8859-4:1988", |
| 2360 |
"iso_8859-5:1988", |
| 2361 |
"iso_8859-8:1988", |
| 2362 |
"x-unicode20utf8", |
| 2363 |
"iso_8859-9:1989", |
| 2364 |
"csisolatingreek", |
| 2365 |
"x-mac-ukrainian", |
| 2366 |
"iso-2022-cn-ext", |
| 2367 |
"csisolatinarabic", |
| 2368 |
"csisolatinhebrew", |
| 2369 |
"unicode-1-1-utf-8", |
| 2370 |
"csisolatincyrillic", |
| 2371 |
"cseucpkdfmtjapanese", |
| 2372 |
]; |
| 2373 |
|
| 2374 |
static ENCODINGS_IN_LABEL_SORT: [&'static Encoding; 228] = [ |
| 2375 |
&WINDOWS_1252_INIT, |
| 2376 |
&ISO_8859_2_INIT, |
| 2377 |
&ISO_8859_3_INIT, |
| 2378 |
&ISO_8859_4_INIT, |
| 2379 |
&WINDOWS_1254_INIT, |
| 2380 |
&ISO_8859_10_INIT, |
| 2381 |
&ISO_8859_15_INIT, |
| 2382 |
&IBM866_INIT, |
| 2383 |
&MACINTOSH_INIT, |
| 2384 |
&KOI8_R_INIT, |
| 2385 |
&GBK_INIT, |
| 2386 |
&BIG5_INIT, |
| 2387 |
&UTF_8_INIT, |
| 2388 |
&KOI8_R_INIT, |
| 2389 |
&SHIFT_JIS_INIT, |
| 2390 |
&UTF_16LE_INIT, |
| 2391 |
&SHIFT_JIS_INIT, |
| 2392 |
&IBM866_INIT, |
| 2393 |
&UTF_8_INIT, |
| 2394 |
&WINDOWS_1252_INIT, |
| 2395 |
&WINDOWS_1252_INIT, |
| 2396 |
&GBK_INIT, |
| 2397 |
&ISO_8859_7_INIT, |
| 2398 |
&WINDOWS_1250_INIT, |
| 2399 |
&WINDOWS_1251_INIT, |
| 2400 |
&WINDOWS_1252_INIT, |
| 2401 |
&GBK_INIT, |
| 2402 |
&WINDOWS_1252_INIT, |
| 2403 |
&ISO_8859_2_INIT, |
| 2404 |
&WINDOWS_1253_INIT, |
| 2405 |
&ISO_8859_3_INIT, |
| 2406 |
&WINDOWS_1254_INIT, |
| 2407 |
&ISO_8859_4_INIT, |
| 2408 |
&WINDOWS_1255_INIT, |
| 2409 |
&BIG5_INIT, |
| 2410 |
&WINDOWS_1254_INIT, |
| 2411 |
&UTF_16LE_INIT, |
| 2412 |
&WINDOWS_1256_INIT, |
| 2413 |
&IBM866_INIT, |
| 2414 |
&ISO_8859_10_INIT, |
| 2415 |
&WINDOWS_1257_INIT, |
| 2416 |
&WINDOWS_1258_INIT, |
| 2417 |
&ISO_8859_7_INIT, |
| 2418 |
&WINDOWS_1252_INIT, |
| 2419 |
&ISO_8859_6_INIT, |
| 2420 |
&ISO_8859_8_INIT, |
| 2421 |
&EUC_KR_INIT, |
| 2422 |
&EUC_JP_INIT, |
| 2423 |
&KOI8_R_INIT, |
| 2424 |
&KOI8_R_INIT, |
| 2425 |
&EUC_KR_INIT, |
| 2426 |
&SHIFT_JIS_INIT, |
| 2427 |
&KOI8_U_INIT, |
| 2428 |
&ISO_8859_8_INIT, |
| 2429 |
&WINDOWS_874_INIT, |
| 2430 |
&GB18030_INIT, |
| 2431 |
&EUC_KR_INIT, |
| 2432 |
&GBK_INIT, |
| 2433 |
&WINDOWS_874_INIT, |
| 2434 |
&BIG5_INIT, |
| 2435 |
&UTF_16LE_INIT, |
| 2436 |
&GBK_INIT, |
| 2437 |
&ISO_8859_8_I_INIT, |
| 2438 |
&KOI8_R_INIT, |
| 2439 |
&EUC_KR_INIT, |
| 2440 |
&KOI8_U_INIT, |
| 2441 |
&WINDOWS_1250_INIT, |
| 2442 |
&EUC_KR_INIT, |
| 2443 |
&WINDOWS_1251_INIT, |
| 2444 |
&WINDOWS_1252_INIT, |
| 2445 |
&GBK_INIT, |
| 2446 |
&WINDOWS_1252_INIT, |
| 2447 |
&ISO_8859_2_INIT, |
| 2448 |
&WINDOWS_1253_INIT, |
| 2449 |
&ISO_8859_3_INIT, |
| 2450 |
&ISO_8859_6_INIT, |
| 2451 |
&WINDOWS_1254_INIT, |
| 2452 |
&ISO_8859_4_INIT, |
| 2453 |
&WINDOWS_1255_INIT, |
| 2454 |
&ISO_8859_5_INIT, |
| 2455 |
&BIG5_INIT, |
| 2456 |
&WINDOWS_1256_INIT, |
| 2457 |
&IBM866_INIT, |
| 2458 |
&ISO_8859_6_INIT, |
| 2459 |
&WINDOWS_1257_INIT, |
| 2460 |
&ISO_8859_7_INIT, |
| 2461 |
&ISO_8859_6_INIT, |
| 2462 |
&ISO_8859_7_INIT, |
| 2463 |
&ISO_8859_7_INIT, |
| 2464 |
&WINDOWS_1258_INIT, |
| 2465 |
&ISO_8859_8_INIT, |
| 2466 |
&WINDOWS_1254_INIT, |
| 2467 |
&ISO_8859_5_INIT, |
| 2468 |
&UTF_16BE_INIT, |
| 2469 |
&UTF_16LE_INIT, |
| 2470 |
&WINDOWS_1252_INIT, |
| 2471 |
&SHIFT_JIS_INIT, |
| 2472 |
&EUC_JP_INIT, |
| 2473 |
&ISO_8859_10_INIT, |
| 2474 |
&WINDOWS_1252_INIT, |
| 2475 |
&WINDOWS_874_INIT, |
| 2476 |
&ISO_8859_2_INIT, |
| 2477 |
&ISO_8859_3_INIT, |
| 2478 |
&ISO_8859_13_INIT, |
| 2479 |
&ISO_8859_4_INIT, |
| 2480 |
&ISO_8859_14_INIT, |
| 2481 |
&ISO_8859_5_INIT, |
| 2482 |
&ISO_8859_15_INIT, |
| 2483 |
&ISO_8859_6_INIT, |
| 2484 |
&ISO_8859_7_INIT, |
| 2485 |
&ISO_8859_8_INIT, |
| 2486 |
&GBK_INIT, |
| 2487 |
&WINDOWS_1254_INIT, |
| 2488 |
&UTF_16LE_INIT, |
| 2489 |
&MACINTOSH_INIT, |
| 2490 |
&SHIFT_JIS_INIT, |
| 2491 |
&SHIFT_JIS_INIT, |
| 2492 |
&WINDOWS_1252_INIT, |
| 2493 |
&ISO_8859_10_INIT, |
| 2494 |
&ISO_8859_4_INIT, |
| 2495 |
&GBK_INIT, |
| 2496 |
&WINDOWS_1252_INIT, |
| 2497 |
&WINDOWS_1252_INIT, |
| 2498 |
&ISO_8859_2_INIT, |
| 2499 |
&WINDOWS_874_INIT, |
| 2500 |
&ISO_8859_2_INIT, |
| 2501 |
&ISO_8859_2_INIT, |
| 2502 |
&REPLACEMENT_INIT, |
| 2503 |
&ISO_8859_3_INIT, |
| 2504 |
&ISO_8859_3_INIT, |
| 2505 |
&ISO_8859_13_INIT, |
| 2506 |
&ISO_8859_4_INIT, |
| 2507 |
&ISO_8859_4_INIT, |
| 2508 |
&ISO_8859_14_INIT, |
| 2509 |
&ISO_8859_5_INIT, |
| 2510 |
&ISO_8859_5_INIT, |
| 2511 |
&ISO_8859_5_INIT, |
| 2512 |
&ISO_8859_15_INIT, |
| 2513 |
&ISO_8859_6_INIT, |
| 2514 |
&ISO_8859_6_INIT, |
| 2515 |
&ISO_8859_7_INIT, |
| 2516 |
&ISO_8859_7_INIT, |
| 2517 |
&ISO_8859_7_INIT, |
| 2518 |
&ISO_8859_6_INIT, |
| 2519 |
&ISO_8859_10_INIT, |
| 2520 |
&ISO_8859_8_INIT, |
| 2521 |
&ISO_8859_8_INIT, |
| 2522 |
&ISO_8859_8_INIT, |
| 2523 |
&WINDOWS_1254_INIT, |
| 2524 |
&WINDOWS_1254_INIT, |
| 2525 |
&WINDOWS_1254_INIT, |
| 2526 |
&ISO_8859_3_INIT, |
| 2527 |
&EUC_KR_INIT, |
| 2528 |
&BIG5_INIT, |
| 2529 |
&SHIFT_JIS_INIT, |
| 2530 |
&ISO_8859_10_INIT, |
| 2531 |
&WINDOWS_874_INIT, |
| 2532 |
&WINDOWS_1252_INIT, |
| 2533 |
&ISO_8859_2_INIT, |
| 2534 |
&ISO_8859_13_INIT, |
| 2535 |
&ISO_8859_3_INIT, |
| 2536 |
&ISO_8859_14_INIT, |
| 2537 |
&WINDOWS_874_INIT, |
| 2538 |
&ISO_8859_4_INIT, |
| 2539 |
&ISO_8859_15_INIT, |
| 2540 |
&ISO_8859_15_INIT, |
| 2541 |
&WINDOWS_1254_INIT, |
| 2542 |
&ISO_8859_16_INIT, |
| 2543 |
&ISO_8859_10_INIT, |
| 2544 |
&EUC_KR_INIT, |
| 2545 |
&ISO_8859_15_INIT, |
| 2546 |
&ISO_8859_6_INIT, |
| 2547 |
&ISO_8859_8_INIT, |
| 2548 |
&UTF_16BE_INIT, |
| 2549 |
&UTF_16LE_INIT, |
| 2550 |
&MACINTOSH_INIT, |
| 2551 |
&ISO_8859_6_INIT, |
| 2552 |
&ISO_8859_8_I_INIT, |
| 2553 |
&SHIFT_JIS_INIT, |
| 2554 |
&MACINTOSH_INIT, |
| 2555 |
&REPLACEMENT_INIT, |
| 2556 |
&ISO_2022_JP_INIT, |
| 2557 |
&ISO_2022_JP_INIT, |
| 2558 |
&REPLACEMENT_INIT, |
| 2559 |
&REPLACEMENT_INIT, |
| 2560 |
&REPLACEMENT_INIT, |
| 2561 |
&WINDOWS_1250_INIT, |
| 2562 |
&WINDOWS_1251_INIT, |
| 2563 |
&WINDOWS_1252_INIT, |
| 2564 |
&WINDOWS_1253_INIT, |
| 2565 |
&WINDOWS_1254_INIT, |
| 2566 |
&WINDOWS_1255_INIT, |
| 2567 |
&WINDOWS_1256_INIT, |
| 2568 |
&WINDOWS_1257_INIT, |
| 2569 |
&WINDOWS_1258_INIT, |
| 2570 |
&ISO_8859_6_INIT, |
| 2571 |
&ISO_8859_8_INIT, |
| 2572 |
&ISO_8859_6_INIT, |
| 2573 |
&ISO_8859_8_I_INIT, |
| 2574 |
&ISO_8859_7_INIT, |
| 2575 |
&EUC_KR_INIT, |
| 2576 |
&UTF_8_INIT, |
| 2577 |
&UTF_8_INIT, |
| 2578 |
&EUC_KR_INIT, |
| 2579 |
&WINDOWS_1252_INIT, |
| 2580 |
&EUC_KR_INIT, |
| 2581 |
&X_MAC_CYRILLIC_INIT, |
| 2582 |
&X_USER_DEFINED_INIT, |
| 2583 |
&GBK_INIT, |
| 2584 |
&UTF_16LE_INIT, |
| 2585 |
&WINDOWS_1252_INIT, |
| 2586 |
&ISO_8859_2_INIT, |
| 2587 |
&ISO_8859_6_INIT, |
| 2588 |
&ISO_8859_7_INIT, |
| 2589 |
&ISO_8859_3_INIT, |
| 2590 |
&ISO_8859_4_INIT, |
| 2591 |
&ISO_8859_5_INIT, |
| 2592 |
&ISO_8859_8_INIT, |
| 2593 |
&UTF_8_INIT, |
| 2594 |
&WINDOWS_1254_INIT, |
| 2595 |
&ISO_8859_7_INIT, |
| 2596 |
&X_MAC_CYRILLIC_INIT, |
| 2597 |
&REPLACEMENT_INIT, |
| 2598 |
&ISO_8859_6_INIT, |
| 2599 |
&ISO_8859_8_INIT, |
| 2600 |
&UTF_8_INIT, |
| 2601 |
&ISO_8859_5_INIT, |
| 2602 |
&EUC_JP_INIT, |
| 2603 |
]; |
| 2604 |
|
| 2605 |
// END GENERATED CODE |
| 2606 |
|
| 2607 |
/// An encoding as defined in the [Encoding Standard][1]. |
| 2608 |
/// |
| 2609 |
/// An _encoding_ defines a mapping from a `u8` sequence to a `char` sequence |
| 2610 |
/// and, in most cases, vice versa. Each encoding has a name, an output |
| 2611 |
/// encoding, and one or more labels. |
| 2612 |
/// |
| 2613 |
/// _Labels_ are ASCII-case-insensitive strings that are used to identify an |
| 2614 |
/// encoding in formats and protocols. The _name_ of the encoding is the |
| 2615 |
/// preferred label in the case appropriate for returning from the |
| 2616 |
/// [`characterSet`][2] property of the `Document` DOM interface. |
| 2617 |
/// |
| 2618 |
/// The _output encoding_ is the encoding used for form submission and URL |
| 2619 |
/// parsing on Web pages in the encoding. This is UTF-8 for the replacement, |
| 2620 |
/// UTF-16LE and UTF-16BE encodings and the encoding itself for other |
| 2621 |
/// encodings. |
| 2622 |
/// |
| 2623 |
/// [1]: https://encoding.spec.whatwg.org/ |
| 2624 |
/// [2]: https://dom.spec.whatwg.org/#dom-document-characterset |
| 2625 |
/// |
| 2626 |
/// # Streaming vs. Non-Streaming |
| 2627 |
/// |
| 2628 |
/// When you have the entire input in a single buffer, you can use the |
| 2629 |
/// methods [`decode()`][3], [`decode_with_bom_removal()`][3], |
| 2630 |
/// [`decode_without_bom_handling()`][5], |
| 2631 |
/// [`decode_without_bom_handling_and_without_replacement()`][6] and |
| 2632 |
/// [`encode()`][7]. (These methods are available to Rust callers only and are |
| 2633 |
/// not available in the C API.) Unlike the rest of the API available to Rust, |
| 2634 |
/// these methods perform heap allocations. You should the `Decoder` and |
| 2635 |
/// `Encoder` objects when your input is split into multiple buffers or when |
| 2636 |
/// you want to control the allocation of the output buffers. |
| 2637 |
/// |
| 2638 |
/// [3]: #method.decode |
| 2639 |
/// [4]: #method.decode_with_bom_removal |
| 2640 |
/// [5]: #method.decode_without_bom_handling |
| 2641 |
/// [6]: #method.decode_without_bom_handling_and_without_replacement |
| 2642 |
/// [7]: #method.encode |
| 2643 |
/// |
| 2644 |
/// # Instances |
| 2645 |
/// |
| 2646 |
/// All instances of `Encoding` are statically allocated and have the `'static` |
| 2647 |
/// lifetime. There is precisely one unique `Encoding` instance for each |
| 2648 |
/// encoding defined in the Encoding Standard. |
| 2649 |
/// |
| 2650 |
/// To obtain a reference to a particular encoding whose identity you know at |
| 2651 |
/// compile time, use a `static` that refers to encoding. There is a `static` |
| 2652 |
/// for each encoding. The `static`s are named in all caps with hyphens |
| 2653 |
/// replaced with underscores (and in C/C++ have `_ENCODING` appended to the |
| 2654 |
/// name). For example, if you know at compile time that you will want to |
| 2655 |
/// decode using the UTF-8 encoding, use the `UTF_8` `static` (`UTF_8_ENCODING` |
| 2656 |
/// in C/C++). |
| 2657 |
/// |
| 2658 |
/// Additionally, there are non-reference-typed forms ending with `_INIT` to |
| 2659 |
/// work around the problem that `static`s of the type `&'static Encoding` |
| 2660 |
/// cannot be used to initialize items of an array whose type is |
| 2661 |
/// `[&'static Encoding; N]`. |
| 2662 |
/// |
| 2663 |
/// If you don't know what encoding you need at compile time and need to |
| 2664 |
/// dynamically get an encoding by label, use |
| 2665 |
/// <code>Encoding::<a href="#method.for_label">for_label</a>(<var>label</var>)</code>. |
| 2666 |
/// |
| 2667 |
/// Instances of `Encoding` can be compared with `==` (in both Rust and in |
| 2668 |
/// C/C++). |
| 2669 |
pub struct Encoding { |
| 2670 |
name: &'static str, |
| 2671 |
variant: VariantEncoding, |
| 2672 |
} |
| 2673 |
|
| 2674 |
impl Encoding { |
| 2675 |
/// Implements the |
| 2676 |
/// [_get an encoding_](https://encoding.spec.whatwg.org/#concept-encoding-get) |
| 2677 |
/// algorithm. |
| 2678 |
/// |
| 2679 |
/// If, after ASCII-lowercasing and removing leading and trailing |
| 2680 |
/// whitespace, the argument matches a label defined in the Encoding |
| 2681 |
/// Standard, `Some(&'static Encoding)` representing the corresponding |
| 2682 |
/// encoding is returned. If there is no match, `None` is returned. |
| 2683 |
/// |
| 2684 |
/// This is the right method to use if the action upon the method returning |
| 2685 |
/// `None` is to use a fallback encoding (e.g. `WINDOWS_1252`) instead. |
| 2686 |
/// When the action upon the method returning `None` is not to proceed with |
| 2687 |
/// a fallback but to refuse processing, `for_label_no_replacement()` is more |
| 2688 |
/// appropriate. |
| 2689 |
/// |
| 2690 |
/// The argument is of type `&[u8]` instead of `&str` to save callers |
| 2691 |
/// that are extracting the label from a non-UTF-8 protocol the trouble |
| 2692 |
/// of conversion to UTF-8. (If you have a `&str`, just call `.as_bytes()` |
| 2693 |
/// on it.) |
| 2694 |
/// |
| 2695 |
/// Available via the C wrapper. |
| 2696 |
/// |
| 2697 |
/// # Example |
| 2698 |
/// ``` |
| 2699 |
/// use encoding_rs::Encoding; |
| 2700 |
/// |
| 2701 |
/// assert_eq!(Some(encoding_rs::UTF_8), Encoding::for_label(b"utf-8")); |
| 2702 |
/// assert_eq!(Some(encoding_rs::UTF_8), Encoding::for_label(b"unicode11utf8")); |
| 2703 |
/// |
| 2704 |
/// assert_eq!(Some(encoding_rs::ISO_8859_2), Encoding::for_label(b"latin2")); |
| 2705 |
/// |
| 2706 |
/// assert_eq!(Some(encoding_rs::UTF_16BE), Encoding::for_label(b"utf-16be")); |
| 2707 |
/// |
| 2708 |
/// assert_eq!(None, Encoding::for_label(b"unrecognized label")); |
| 2709 |
/// ``` |
| 2710 |
pub fn for_label(label: &[u8]) -> Option<&'static Encoding> { |
| 2711 |
let mut trimmed = [0u8; LONGEST_LABEL_LENGTH]; |
| 2712 |
let mut trimmed_pos = 0usize; |
| 2713 |
let mut iter = label.into_iter(); |
| 2714 |
// before |
| 2715 |
loop { |
| 2716 |
match iter.next() { |
| 2717 |
None => { |
| 2718 |
return None; |
| 2719 |
} |
| 2720 |
Some(byte) => { |
| 2721 |
// The characters used in labels are: |
| 2722 |
// a-z (except q, but excluding it below seems excessive) |
| 2723 |
// 0-9 |
| 2724 |
// . _ - : |
| 2725 |
match *byte { |
| 2726 |
0x09u8 | 0x0Au8 | 0x0Cu8 | 0x0Du8 | 0x20u8 => { |
| 2727 |
continue; |
| 2728 |
} |
| 2729 |
b'A'..=b'Z' => { |
| 2730 |
trimmed[trimmed_pos] = *byte + 0x20u8; |
| 2731 |
trimmed_pos = 1usize; |
| 2732 |
break; |
| 2733 |
} |
| 2734 |
b'a'..=b'z' | b'0'..=b'9' | b'-' | b'_' | b':' | b'.' => { |
| 2735 |
trimmed[trimmed_pos] = *byte; |
| 2736 |
trimmed_pos = 1usize; |
| 2737 |
break; |
| 2738 |
} |
| 2739 |
_ => { |
| 2740 |
return None; |
| 2741 |
} |
| 2742 |
} |
| 2743 |
} |
| 2744 |
} |
| 2745 |
} |
| 2746 |
// inside |
| 2747 |
loop { |
| 2748 |
match iter.next() { |
| 2749 |
None => { |
| 2750 |
break; |
| 2751 |
} |
| 2752 |
Some(byte) => { |
| 2753 |
match *byte { |
| 2754 |
0x09u8 | 0x0Au8 | 0x0Cu8 | 0x0Du8 | 0x20u8 => { |
| 2755 |
break; |
| 2756 |
} |
| 2757 |
b'A'..=b'Z' => { |
| 2758 |
if trimmed_pos == LONGEST_LABEL_LENGTH { |
| 2759 |
// There's no encoding with a label this long |
| 2760 |
return None; |
| 2761 |
} |
| 2762 |
trimmed[trimmed_pos] = *byte + 0x20u8; |
| 2763 |
trimmed_pos += 1usize; |
| 2764 |
continue; |
| 2765 |
} |
| 2766 |
b'a'..=b'z' | b'0'..=b'9' | b'-' | b'_' | b':' | b'.' => { |
| 2767 |
if trimmed_pos == LONGEST_LABEL_LENGTH { |
| 2768 |
// There's no encoding with a label this long |
| 2769 |
return None; |
| 2770 |
} |
| 2771 |
trimmed[trimmed_pos] = *byte; |
| 2772 |
trimmed_pos += 1usize; |
| 2773 |
continue; |
| 2774 |
} |
| 2775 |
_ => { |
| 2776 |
return None; |
| 2777 |
} |
| 2778 |
} |
| 2779 |
} |
| 2780 |
} |
| 2781 |
} |
| 2782 |
// after |
| 2783 |
loop { |
| 2784 |
match iter.next() { |
| 2785 |
None => { |
| 2786 |
break; |
| 2787 |
} |
| 2788 |
Some(byte) => { |
| 2789 |
match *byte { |
| 2790 |
0x09u8 | 0x0Au8 | 0x0Cu8 | 0x0Du8 | 0x20u8 => { |
| 2791 |
continue; |
| 2792 |
} |
| 2793 |
_ => { |
| 2794 |
// There's no label with space in the middle |
| 2795 |
return None; |
| 2796 |
} |
| 2797 |
} |
| 2798 |
} |
| 2799 |
} |
| 2800 |
} |
| 2801 |
let candidate = &trimmed[..trimmed_pos]; |
| 2802 |
match LABELS_SORTED.binary_search_by(|probe| { |
| 2803 |
let bytes = probe.as_bytes(); |
| 2804 |
let c = bytes.len().cmp(&candidate.len()); |
| 2805 |
if c != Ordering::Equal { |
| 2806 |
return c; |
| 2807 |
} |
| 2808 |
let probe_iter = bytes.iter().rev(); |
| 2809 |
let candidate_iter = candidate.iter().rev(); |
| 2810 |
probe_iter.cmp(candidate_iter) |
| 2811 |
}) { |
| 2812 |
Ok(i) => Some(ENCODINGS_IN_LABEL_SORT[i]), |
| 2813 |
Err(_) => None, |
| 2814 |
} |
| 2815 |
} |
| 2816 |
|
| 2817 |
/// This method behaves the same as `for_label()`, except when `for_label()` |
| 2818 |
/// would return `Some(REPLACEMENT)`, this method returns `None` instead. |
| 2819 |
/// |
| 2820 |
/// This method is useful in scenarios where a fatal error is required |
| 2821 |
/// upon invalid label, because in those cases the caller typically wishes |
| 2822 |
/// to treat the labels that map to the replacement encoding as fatal |
| 2823 |
/// errors, too. |
| 2824 |
/// |
| 2825 |
/// It is not OK to use this method when the action upon the method returning |
| 2826 |
/// `None` is to use a fallback encoding (e.g. `WINDOWS_1252`). In such a |
| 2827 |
/// case, the `for_label()` method should be used instead in order to avoid |
| 2828 |
/// unsafe fallback for labels that `for_label()` maps to `Some(REPLACEMENT)`. |
| 2829 |
/// |
| 2830 |
/// Available via the C wrapper. |
| 2831 |
#[inline] |
| 2832 |
pub fn for_label_no_replacement(label: &[u8]) -> Option<&'static Encoding> { |
| 2833 |
match Encoding::for_label(label) { |
| 2834 |
None => None, |
| 2835 |
Some(encoding) => { |
| 2836 |
if encoding == REPLACEMENT { |
| 2837 |
None |
| 2838 |
} else { |
| 2839 |
Some(encoding) |
| 2840 |
} |
| 2841 |
} |
| 2842 |
} |
| 2843 |
} |
| 2844 |
|
| 2845 |
/// Performs non-incremental BOM sniffing. |
| 2846 |
/// |
| 2847 |
/// The argument must either be a buffer representing the entire input |
| 2848 |
/// stream (non-streaming case) or a buffer representing at least the first |
| 2849 |
/// three bytes of the input stream (streaming case). |
| 2850 |
/// |
| 2851 |
/// Returns `Some((UTF_8, 3))`, `Some((UTF_16LE, 2))` or |
| 2852 |
/// `Some((UTF_16BE, 2))` if the argument starts with the UTF-8, UTF-16LE |
| 2853 |
/// or UTF-16BE BOM or `None` otherwise. |
| 2854 |
/// |
| 2855 |
/// Available via the C wrapper. |
| 2856 |
#[inline] |
| 2857 |
pub fn for_bom(buffer: &[u8]) -> Option<(&'static Encoding, usize)> { |
| 2858 |
if buffer.starts_with(b"\xEF\xBB\xBF") { |
| 2859 |
Some((UTF_8, 3)) |
| 2860 |
} else if buffer.starts_with(b"\xFF\xFE") { |
| 2861 |
Some((UTF_16LE, 2)) |
| 2862 |
} else if buffer.starts_with(b"\xFE\xFF") { |
| 2863 |
Some((UTF_16BE, 2)) |
| 2864 |
} else { |
| 2865 |
None |
| 2866 |
} |
| 2867 |
} |
| 2868 |
|
| 2869 |
/// Returns the name of this encoding. |
| 2870 |
/// |
| 2871 |
/// This name is appropriate to return as-is from the DOM |
| 2872 |
/// `document.characterSet` property. |
| 2873 |
/// |
| 2874 |
/// Available via the C wrapper. |
| 2875 |
#[inline] |
| 2876 |
pub fn name(&'static self) -> &'static str { |
| 2877 |
self.name |
| 2878 |
} |
| 2879 |
|
| 2880 |
/// Checks whether the _output encoding_ of this encoding can encode every |
| 2881 |
/// `char`. (Only true if the output encoding is UTF-8.) |
| 2882 |
/// |
| 2883 |
/// Available via the C wrapper. |
| 2884 |
#[inline] |
| 2885 |
pub fn can_encode_everything(&'static self) -> bool { |
| 2886 |
self.output_encoding() == UTF_8 |
| 2887 |
} |
| 2888 |
|
| 2889 |
/// Checks whether the bytes 0x00...0x7F map exclusively to the characters |
| 2890 |
/// U+0000...U+007F and vice versa. |
| 2891 |
/// |
| 2892 |
/// Available via the C wrapper. |
| 2893 |
#[inline] |
| 2894 |
pub fn is_ascii_compatible(&'static self) -> bool { |
| 2895 |
!(self == REPLACEMENT || self == UTF_16BE || self == UTF_16LE || self == ISO_2022_JP) |
| 2896 |
} |
| 2897 |
|
| 2898 |
/// Checks whether this encoding maps one byte to one Basic Multilingual |
| 2899 |
/// Plane code point (i.e. byte length equals decoded UTF-16 length) and |
| 2900 |
/// vice versa (for mappable characters). |
| 2901 |
/// |
| 2902 |
/// `true` iff this encoding is on the list of [Legacy single-byte |
| 2903 |
/// encodings](https://encoding.spec.whatwg.org/#legacy-single-byte-encodings) |
| 2904 |
/// in the spec or x-user-defined. |
| 2905 |
/// |
| 2906 |
/// Available via the C wrapper. |
| 2907 |
#[inline] |
| 2908 |
pub fn is_single_byte(&'static self) -> bool { |
| 2909 |
self.variant.is_single_byte() |
| 2910 |
} |
| 2911 |
|
| 2912 |
/// Checks whether the bytes 0x00...0x7F map mostly to the characters |
| 2913 |
/// U+0000...U+007F and vice versa. |
| 2914 |
#[cfg(feature = "alloc")] |
| 2915 |
#[inline] |
| 2916 |
fn is_potentially_borrowable(&'static self) -> bool { |
| 2917 |
!(self == REPLACEMENT || self == UTF_16BE || self == UTF_16LE) |
| 2918 |
} |
| 2919 |
|
| 2920 |
/// Returns the _output encoding_ of this encoding. This is UTF-8 for |
| 2921 |
/// UTF-16BE, UTF-16LE, and replacement and the encoding itself otherwise. |
| 2922 |
/// |
| 2923 |
/// _Note:_ The _output encoding_ concept is needed for form submission and |
| 2924 |
/// error handling in the query strings of URLs in the Web Platform. |
| 2925 |
/// |
| 2926 |
/// Available via the C wrapper. |
| 2927 |
#[inline] |
| 2928 |
pub fn output_encoding(&'static self) -> &'static Encoding { |
| 2929 |
if self == REPLACEMENT || self == UTF_16BE || self == UTF_16LE { |
| 2930 |
UTF_8 |
| 2931 |
} else { |
| 2932 |
self |
| 2933 |
} |
| 2934 |
} |
| 2935 |
|
| 2936 |
/// Decode complete input to `Cow<'a, str>` _with BOM sniffing_ and with |
| 2937 |
/// malformed sequences replaced with the REPLACEMENT CHARACTER when the |
| 2938 |
/// entire input is available as a single buffer (i.e. the end of the |
| 2939 |
/// buffer marks the end of the stream). |
| 2940 |
/// |
| 2941 |
/// The BOM, if any, does not appear in the output. |
| 2942 |
/// |
| 2943 |
/// This method implements the (non-streaming version of) the |
| 2944 |
/// [_decode_](https://encoding.spec.whatwg.org/#decode) spec concept. |
| 2945 |
/// |
| 2946 |
/// The second item in the returned tuple is the encoding that was actually |
| 2947 |
/// used (which may differ from this encoding thanks to BOM sniffing). |
| 2948 |
/// |
| 2949 |
/// The third item in the returned tuple indicates whether there were |
| 2950 |
/// malformed sequences (that were replaced with the REPLACEMENT CHARACTER). |
| 2951 |
/// |
| 2952 |
/// _Note:_ It is wrong to use this when the input buffer represents only |
| 2953 |
/// a segment of the input instead of the whole input. Use `new_decoder()` |
| 2954 |
/// when decoding segmented input. |
| 2955 |
/// |
| 2956 |
/// This method performs a one or two heap allocations for the backing |
| 2957 |
/// buffer of the `String` when unable to borrow. (One allocation if not |
| 2958 |
/// errors and potentially another one in the presence of errors.) The |
| 2959 |
/// first allocation assumes jemalloc and may not be optimal with |
| 2960 |
/// allocators that do not use power-of-two buckets. A borrow is performed |
| 2961 |
/// if decoding UTF-8 and the input is valid UTF-8, if decoding an |
| 2962 |
/// ASCII-compatible encoding and the input is ASCII-only, or when decoding |
| 2963 |
/// ISO-2022-JP and the input is entirely in the ASCII state without state |
| 2964 |
/// transitions. |
| 2965 |
/// |
| 2966 |
/// # Panics |
| 2967 |
/// |
| 2968 |
/// If the size calculation for a heap-allocated backing buffer overflows |
| 2969 |
/// `usize`. |
| 2970 |
/// |
| 2971 |
/// Available to Rust only and only with the `alloc` feature enabled (enabled |
| 2972 |
/// by default). |
| 2973 |
#[cfg(feature = "alloc")] |
| 2974 |
#[inline] |
| 2975 |
pub fn decode<'a>(&'static self, bytes: &'a [u8]) -> (Cow<'a, str>, &'static Encoding, bool) { |
| 2976 |
let (encoding, without_bom) = match Encoding::for_bom(bytes) { |
| 2977 |
Some((encoding, bom_length)) => (encoding, &bytes[bom_length..]), |
| 2978 |
None => (self, bytes), |
| 2979 |
}; |
| 2980 |
let (cow, had_errors) = encoding.decode_without_bom_handling(without_bom); |
| 2981 |
(cow, encoding, had_errors) |
| 2982 |
} |
| 2983 |
|
| 2984 |
/// Decode complete input to `Cow<'a, str>` _with BOM removal_ and with |
| 2985 |
/// malformed sequences replaced with the REPLACEMENT CHARACTER when the |
| 2986 |
/// entire input is available as a single buffer (i.e. the end of the |
| 2987 |
/// buffer marks the end of the stream). |
| 2988 |
/// |
| 2989 |
/// Only an initial byte sequence that is a BOM for this encoding is removed. |
| 2990 |
/// |
| 2991 |
/// When invoked on `UTF_8`, this method implements the (non-streaming |
| 2992 |
/// version of) the |
| 2993 |
/// [_UTF-8 decode_](https://encoding.spec.whatwg.org/#utf-8-decode) spec |
| 2994 |
/// concept. |
| 2995 |
/// |
| 2996 |
/// The second item in the returned pair indicates whether there were |
| 2997 |
/// malformed sequences (that were replaced with the REPLACEMENT CHARACTER). |
| 2998 |
/// |
| 2999 |
/// _Note:_ It is wrong to use this when the input buffer represents only |
| 3000 |
/// a segment of the input instead of the whole input. Use |
| 3001 |
/// `new_decoder_with_bom_removal()` when decoding segmented input. |
| 3002 |
/// |
| 3003 |
/// This method performs a one or two heap allocations for the backing |
| 3004 |
/// buffer of the `String` when unable to borrow. (One allocation if not |
| 3005 |
/// errors and potentially another one in the presence of errors.) The |
| 3006 |
/// first allocation assumes jemalloc and may not be optimal with |
| 3007 |
/// allocators that do not use power-of-two buckets. A borrow is performed |
| 3008 |
/// if decoding UTF-8 and the input is valid UTF-8, if decoding an |
| 3009 |
/// ASCII-compatible encoding and the input is ASCII-only, or when decoding |
| 3010 |
/// ISO-2022-JP and the input is entirely in the ASCII state without state |
| 3011 |
/// transitions. |
| 3012 |
/// |
| 3013 |
/// # Panics |
| 3014 |
/// |
| 3015 |
/// If the size calculation for a heap-allocated backing buffer overflows |
| 3016 |
/// `usize`. |
| 3017 |
/// |
| 3018 |
/// Available to Rust only and only with the `alloc` feature enabled (enabled |
| 3019 |
/// by default). |
| 3020 |
#[cfg(feature = "alloc")] |
| 3021 |
#[inline] |
| 3022 |
pub fn decode_with_bom_removal<'a>(&'static self, bytes: &'a [u8]) -> (Cow<'a, str>, bool) { |
| 3023 |
let without_bom = if self == UTF_8 && bytes.starts_with(b"\xEF\xBB\xBF") { |
| 3024 |
&bytes[3..] |
| 3025 |
} else if (self == UTF_16LE && bytes.starts_with(b"\xFF\xFE")) |
| 3026 |
|| (self == UTF_16BE && bytes.starts_with(b"\xFE\xFF")) |
| 3027 |
{ |
| 3028 |
&bytes[2..] |
| 3029 |
} else { |
| 3030 |
bytes |
| 3031 |
}; |
| 3032 |
self.decode_without_bom_handling(without_bom) |
| 3033 |
} |
| 3034 |
|
| 3035 |
/// Decode complete input to `Cow<'a, str>` _without BOM handling_ and |
| 3036 |
/// with malformed sequences replaced with the REPLACEMENT CHARACTER when |
| 3037 |
/// the entire input is available as a single buffer (i.e. the end of the |
| 3038 |
/// buffer marks the end of the stream). |
| 3039 |
/// |
| 3040 |
/// When invoked on `UTF_8`, this method implements the (non-streaming |
| 3041 |
/// version of) the |
| 3042 |
/// [_UTF-8 decode without BOM_](https://encoding.spec.whatwg.org/#utf-8-decode-without-bom) |
| 3043 |
/// spec concept. |
| 3044 |
/// |
| 3045 |
/// The second item in the returned pair indicates whether there were |
| 3046 |
/// malformed sequences (that were replaced with the REPLACEMENT CHARACTER). |
| 3047 |
/// |
| 3048 |
/// _Note:_ It is wrong to use this when the input buffer represents only |
| 3049 |
/// a segment of the input instead of the whole input. Use |
| 3050 |
/// `new_decoder_without_bom_handling()` when decoding segmented input. |
| 3051 |
/// |
| 3052 |
/// This method performs a one or two heap allocations for the backing |
| 3053 |
/// buffer of the `String` when unable to borrow. (One allocation if not |
| 3054 |
/// errors and potentially another one in the presence of errors.) The |
| 3055 |
/// first allocation assumes jemalloc and may not be optimal with |
| 3056 |
/// allocators that do not use power-of-two buckets. A borrow is performed |
| 3057 |
/// if decoding UTF-8 and the input is valid UTF-8, if decoding an |
| 3058 |
/// ASCII-compatible encoding and the input is ASCII-only, or when decoding |
| 3059 |
/// ISO-2022-JP and the input is entirely in the ASCII state without state |
| 3060 |
/// transitions. |
| 3061 |
/// |
| 3062 |
/// # Panics |
| 3063 |
/// |
| 3064 |
/// If the size calculation for a heap-allocated backing buffer overflows |
| 3065 |
/// `usize`. |
| 3066 |
/// |
| 3067 |
/// Available to Rust only and only with the `alloc` feature enabled (enabled |
| 3068 |
/// by default). |
| 3069 |
#[cfg(feature = "alloc")] |
| 3070 |
pub fn decode_without_bom_handling<'a>(&'static self, bytes: &'a [u8]) -> (Cow<'a, str>, bool) { |
| 3071 |
let (mut decoder, mut string, mut total_read) = if self.is_potentially_borrowable() { |
| 3072 |
let valid_up_to = if self == UTF_8 { |
| 3073 |
utf8_valid_up_to(bytes) |
| 3074 |
} else if self == ISO_2022_JP { |
| 3075 |
iso_2022_jp_ascii_valid_up_to(bytes) |
| 3076 |
} else { |
| 3077 |
ascii_valid_up_to(bytes) |
| 3078 |
}; |
| 3079 |
if valid_up_to == bytes.len() { |
| 3080 |
let str: &str = unsafe { core::str::from_utf8_unchecked(bytes) }; |
| 3081 |
return (Cow::Borrowed(str), false); |
| 3082 |
} |
| 3083 |
let decoder = self.new_decoder_without_bom_handling(); |
| 3084 |
|
| 3085 |
let rounded_without_replacement = checked_next_power_of_two(checked_add( |
| 3086 |
valid_up_to, |
| 3087 |
decoder.max_utf8_buffer_length_without_replacement(bytes.len() - valid_up_to), |
| 3088 |
)); |
| 3089 |
let with_replacement = checked_add( |
| 3090 |
valid_up_to, |
| 3091 |
decoder.max_utf8_buffer_length(bytes.len() - valid_up_to), |
| 3092 |
); |
| 3093 |
let mut string = String::with_capacity( |
| 3094 |
checked_min(rounded_without_replacement, with_replacement).unwrap(), |
| 3095 |
); |
| 3096 |
unsafe { |
| 3097 |
let vec = string.as_mut_vec(); |
| 3098 |
vec.set_len(valid_up_to); |
| 3099 |
core::ptr::copy_nonoverlapping(bytes.as_ptr(), vec.as_mut_ptr(), valid_up_to); |
| 3100 |
} |
| 3101 |
(decoder, string, valid_up_to) |
| 3102 |
} else { |
| 3103 |
let decoder = self.new_decoder_without_bom_handling(); |
| 3104 |
let rounded_without_replacement = checked_next_power_of_two( |
| 3105 |
decoder.max_utf8_buffer_length_without_replacement(bytes.len()), |
| 3106 |
); |
| 3107 |
let with_replacement = decoder.max_utf8_buffer_length(bytes.len()); |
| 3108 |
let string = String::with_capacity( |
| 3109 |
checked_min(rounded_without_replacement, with_replacement).unwrap(), |
| 3110 |
); |
| 3111 |
(decoder, string, 0) |
| 3112 |
}; |
| 3113 |
|
| 3114 |
let mut total_had_errors = false; |
| 3115 |
loop { |
| 3116 |
let (result, read, had_errors) = |
| 3117 |
decoder.decode_to_string(&bytes[total_read..], &mut string, true); |
| 3118 |
total_read += read; |
| 3119 |
total_had_errors |= had_errors; |
| 3120 |
match result { |
| 3121 |
CoderResult::InputEmpty => { |
| 3122 |
debug_assert_eq!(total_read, bytes.len()); |
| 3123 |
return (Cow::Owned(string), total_had_errors); |
| 3124 |
} |
| 3125 |
CoderResult::OutputFull => { |
| 3126 |
// Allocate for the worst case. That is, we should come |
| 3127 |
// here at most once per invocation of this method. |
| 3128 |
let needed = decoder.max_utf8_buffer_length(bytes.len() - total_read); |
| 3129 |
string.reserve(needed.unwrap()); |
| 3130 |
} |
| 3131 |
} |
| 3132 |
} |
| 3133 |
} |
| 3134 |
|
| 3135 |
/// Decode complete input to `Cow<'a, str>` _without BOM handling_ and |
| 3136 |
/// _with malformed sequences treated as fatal_ when the entire input is |
| 3137 |
/// available as a single buffer (i.e. the end of the buffer marks the end |
| 3138 |
/// of the stream). |
| 3139 |
/// |
| 3140 |
/// When invoked on `UTF_8`, this method implements the (non-streaming |
| 3141 |
/// version of) the |
| 3142 |
/// [_UTF-8 decode without BOM or fail_](https://encoding.spec.whatwg.org/#utf-8-decode-without-bom-or-fail) |
| 3143 |
/// spec concept. |
| 3144 |
/// |
| 3145 |
/// Returns `None` if a malformed sequence was encountered and the result |
| 3146 |
/// of the decode as `Some(String)` otherwise. |
| 3147 |
/// |
| 3148 |
/// _Note:_ It is wrong to use this when the input buffer represents only |
| 3149 |
/// a segment of the input instead of the whole input. Use |
| 3150 |
/// `new_decoder_without_bom_handling()` when decoding segmented input. |
| 3151 |
/// |
| 3152 |
/// This method performs a single heap allocation for the backing |
| 3153 |
/// buffer of the `String` when unable to borrow. A borrow is performed if |
| 3154 |
/// decoding UTF-8 and the input is valid UTF-8, if decoding an |
| 3155 |
/// ASCII-compatible encoding and the input is ASCII-only, or when decoding |
| 3156 |
/// ISO-2022-JP and the input is entirely in the ASCII state without state |
| 3157 |
/// transitions. |
| 3158 |
/// |
| 3159 |
/// # Panics |
| 3160 |
/// |
| 3161 |
/// If the size calculation for a heap-allocated backing buffer overflows |
| 3162 |
/// `usize`. |
| 3163 |
/// |
| 3164 |
/// Available to Rust only and only with the `alloc` feature enabled (enabled |
| 3165 |
/// by default). |
| 3166 |
#[cfg(feature = "alloc")] |
| 3167 |
pub fn decode_without_bom_handling_and_without_replacement<'a>( |
| 3168 |
&'static self, |
| 3169 |
bytes: &'a [u8], |
| 3170 |
) -> Option<Cow<'a, str>> { |
| 3171 |
if self == UTF_8 { |
| 3172 |
let valid_up_to = utf8_valid_up_to(bytes); |
| 3173 |
if valid_up_to == bytes.len() { |
| 3174 |
let str: &str = unsafe { core::str::from_utf8_unchecked(bytes) }; |
| 3175 |
return Some(Cow::Borrowed(str)); |
| 3176 |
} |
| 3177 |
return None; |
| 3178 |
} |
| 3179 |
let (mut decoder, mut string, input) = if self.is_potentially_borrowable() { |
| 3180 |
let valid_up_to = if self == ISO_2022_JP { |
| 3181 |
iso_2022_jp_ascii_valid_up_to(bytes) |
| 3182 |
} else { |
| 3183 |
ascii_valid_up_to(bytes) |
| 3184 |
}; |
| 3185 |
if valid_up_to == bytes.len() { |
| 3186 |
let str: &str = unsafe { core::str::from_utf8_unchecked(bytes) }; |
| 3187 |
return Some(Cow::Borrowed(str)); |
| 3188 |
} |
| 3189 |
let decoder = self.new_decoder_without_bom_handling(); |
| 3190 |
let mut string = String::with_capacity( |
| 3191 |
checked_add( |
| 3192 |
valid_up_to, |
| 3193 |
decoder.max_utf8_buffer_length_without_replacement(bytes.len() - valid_up_to), |
| 3194 |
) |
| 3195 |
.unwrap(), |
| 3196 |
); |
| 3197 |
unsafe { |
| 3198 |
let vec = string.as_mut_vec(); |
| 3199 |
vec.set_len(valid_up_to); |
| 3200 |
core::ptr::copy_nonoverlapping(bytes.as_ptr(), vec.as_mut_ptr(), valid_up_to); |
| 3201 |
} |
| 3202 |
(decoder, string, &bytes[valid_up_to..]) |
| 3203 |
} else { |
| 3204 |
let decoder = self.new_decoder_without_bom_handling(); |
| 3205 |
let string = String::with_capacity( |
| 3206 |
decoder |
| 3207 |
.max_utf8_buffer_length_without_replacement(bytes.len()) |
| 3208 |
.unwrap(), |
| 3209 |
); |
| 3210 |
(decoder, string, bytes) |
| 3211 |
}; |
| 3212 |
let (result, read) = decoder.decode_to_string_without_replacement(input, &mut string, true); |
| 3213 |
match result { |
| 3214 |
DecoderResult::InputEmpty => { |
| 3215 |
debug_assert_eq!(read, input.len()); |
| 3216 |
Some(Cow::Owned(string)) |
| 3217 |
} |
| 3218 |
DecoderResult::Malformed(_, _) => None, |
| 3219 |
DecoderResult::OutputFull => unreachable!(), |
| 3220 |
} |
| 3221 |
} |
| 3222 |
|
| 3223 |
/// Encode complete input to `Cow<'a, [u8]>` using the |
| 3224 |
/// [_output encoding_](Encoding::output_encoding) of this encoding with |
| 3225 |
/// unmappable characters replaced with decimal numeric character references |
| 3226 |
/// when the entire input is available as a single buffer (i.e. the end of |
| 3227 |
/// the buffer marks the end of the stream). |
| 3228 |
/// |
| 3229 |
/// This method implements the (non-streaming version of) the |
| 3230 |
/// [_encode_](https://encoding.spec.whatwg.org/#encode) spec concept. For |
| 3231 |
/// the [_UTF-8 encode_](https://encoding.spec.whatwg.org/#utf-8-encode) |
| 3232 |
/// spec concept, it is slightly more efficient to use |
| 3233 |
/// <code><var>string</var>.as_bytes()</code> instead of invoking this |
| 3234 |
/// method on `UTF_8`. |
| 3235 |
/// |
| 3236 |
/// The second item in the returned tuple is the encoding that was actually |
| 3237 |
/// used (*which may differ from this encoding thanks to some encodings |
| 3238 |
/// having UTF-8 as their output encoding*). |
| 3239 |
/// |
| 3240 |
/// The third item in the returned tuple indicates whether there were |
| 3241 |
/// unmappable characters (that were replaced with HTML numeric character |
| 3242 |
/// references). |
| 3243 |
/// |
| 3244 |
/// _Note:_ It is wrong to use this when the input buffer represents only |
| 3245 |
/// a segment of the input instead of the whole input. Use `new_encoder()` |
| 3246 |
/// when encoding segmented output. |
| 3247 |
/// |
| 3248 |
/// When encoding to UTF-8 or when encoding an ASCII-only input to a |
| 3249 |
/// ASCII-compatible encoding, this method returns a borrow of the input |
| 3250 |
/// without a heap allocation. Otherwise, this method performs a single |
| 3251 |
/// heap allocation for the backing buffer of the `Vec<u8>` if there are no |
| 3252 |
/// unmappable characters and potentially multiple heap allocations if |
| 3253 |
/// there are. These allocations are tuned for jemalloc and may not be |
| 3254 |
/// optimal when using a different allocator that doesn't use power-of-two |
| 3255 |
/// buckets. |
| 3256 |
/// |
| 3257 |
/// # Panics |
| 3258 |
/// |
| 3259 |
/// If the size calculation for a heap-allocated backing buffer overflows |
| 3260 |
/// `usize`. |
| 3261 |
/// |
| 3262 |
/// Available to Rust only and only with the `alloc` feature enabled (enabled |
| 3263 |
/// by default). |
| 3264 |
#[cfg(feature = "alloc")] |
| 3265 |
pub fn encode<'a>(&'static self, string: &'a str) -> (Cow<'a, [u8]>, &'static Encoding, bool) { |
| 3266 |
let output_encoding = self.output_encoding(); |
| 3267 |
if output_encoding == UTF_8 { |
| 3268 |
return (Cow::Borrowed(string.as_bytes()), output_encoding, false); |
| 3269 |
} |
| 3270 |
debug_assert!(output_encoding.is_potentially_borrowable()); |
| 3271 |
let bytes = string.as_bytes(); |
| 3272 |
let valid_up_to = if output_encoding == ISO_2022_JP { |
| 3273 |
iso_2022_jp_ascii_valid_up_to(bytes) |
| 3274 |
} else { |
| 3275 |
ascii_valid_up_to(bytes) |
| 3276 |
}; |
| 3277 |
if valid_up_to == bytes.len() { |
| 3278 |
return (Cow::Borrowed(bytes), output_encoding, false); |
| 3279 |
} |
| 3280 |
let mut encoder = output_encoding.new_encoder(); |
| 3281 |
let mut vec: Vec<u8> = Vec::with_capacity( |
| 3282 |
(checked_add( |
| 3283 |
valid_up_to, |
| 3284 |
encoder.max_buffer_length_from_utf8_if_no_unmappables(string.len() - valid_up_to), |
| 3285 |
)) |
| 3286 |
.unwrap() |
| 3287 |
.next_power_of_two(), |
| 3288 |
); |
| 3289 |
unsafe { |
| 3290 |
vec.set_len(valid_up_to); |
| 3291 |
core::ptr::copy_nonoverlapping(bytes.as_ptr(), vec.as_mut_ptr(), valid_up_to); |
| 3292 |
} |
| 3293 |
let mut total_read = valid_up_to; |
| 3294 |
let mut total_had_errors = false; |
| 3295 |
loop { |
| 3296 |
let (result, read, had_errors) = |
| 3297 |
encoder.encode_from_utf8_to_vec(&string[total_read..], &mut vec, true); |
| 3298 |
total_read += read; |
| 3299 |
total_had_errors |= had_errors; |
| 3300 |
match result { |
| 3301 |
CoderResult::InputEmpty => { |
| 3302 |
debug_assert_eq!(total_read, string.len()); |
| 3303 |
return (Cow::Owned(vec), output_encoding, total_had_errors); |
| 3304 |
} |
| 3305 |
CoderResult::OutputFull => { |
| 3306 |
// reserve_exact wants to know how much more on top of current |
| 3307 |
// length--not current capacity. |
| 3308 |
let needed = encoder |
| 3309 |
.max_buffer_length_from_utf8_if_no_unmappables(string.len() - total_read); |
| 3310 |
let rounded = (checked_add(vec.capacity(), needed)) |
| 3311 |
.unwrap() |
| 3312 |
.next_power_of_two(); |
| 3313 |
let additional = rounded - vec.len(); |
| 3314 |
vec.reserve_exact(additional); |
| 3315 |
} |
| 3316 |
} |
| 3317 |
} |
| 3318 |
} |
| 3319 |
|
| 3320 |
fn new_variant_decoder(&'static self) -> VariantDecoder { |
| 3321 |
self.variant.new_variant_decoder() |
| 3322 |
} |
| 3323 |
|
| 3324 |
/// Instantiates a new decoder for this encoding with BOM sniffing enabled. |
| 3325 |
/// |
| 3326 |
/// BOM sniffing may cause the returned decoder to morph into a decoder |
| 3327 |
/// for UTF-8, UTF-16LE or UTF-16BE instead of this encoding. The BOM |
| 3328 |
/// does not appear in the output. |
| 3329 |
/// |
| 3330 |
/// Available via the C wrapper. |
| 3331 |
#[inline] |
| 3332 |
pub fn new_decoder(&'static self) -> Decoder { |
| 3333 |
Decoder::new(self, self.new_variant_decoder(), BomHandling::Sniff) |
| 3334 |
} |
| 3335 |
|
| 3336 |
/// Instantiates a new decoder for this encoding with BOM removal. |
| 3337 |
/// |
| 3338 |
/// If the input starts with bytes that are the BOM for this encoding, |
| 3339 |
/// those bytes are removed. However, the decoder never morphs into a |
| 3340 |
/// decoder for another encoding: A BOM for another encoding is treated as |
| 3341 |
/// (potentially malformed) input to the decoding algorithm for this |
| 3342 |
/// encoding. |
| 3343 |
/// |
| 3344 |
/// Available via the C wrapper. |
| 3345 |
#[inline] |
| 3346 |
pub fn new_decoder_with_bom_removal(&'static self) -> Decoder { |
| 3347 |
Decoder::new(self, self.new_variant_decoder(), BomHandling::Remove) |
| 3348 |
} |
| 3349 |
|
| 3350 |
/// Instantiates a new decoder for this encoding with BOM handling disabled. |
| 3351 |
/// |
| 3352 |
/// If the input starts with bytes that look like a BOM, those bytes are |
| 3353 |
/// not treated as a BOM. (Hence, the decoder never morphs into a decoder |
| 3354 |
/// for another encoding.) |
| 3355 |
/// |
| 3356 |
/// _Note:_ If the caller has performed BOM sniffing on its own but has not |
| 3357 |
/// removed the BOM, the caller should use `new_decoder_with_bom_removal()` |
| 3358 |
/// instead of this method to cause the BOM to be removed. |
| 3359 |
/// |
| 3360 |
/// Available via the C wrapper. |
| 3361 |
#[inline] |
| 3362 |
pub fn new_decoder_without_bom_handling(&'static self) -> Decoder { |
| 3363 |
Decoder::new(self, self.new_variant_decoder(), BomHandling::Off) |
| 3364 |
} |
| 3365 |
|
| 3366 |
/// Instantiates a new encoder for the [_output encoding_](Encoding::output_encoding) |
| 3367 |
/// of this encoding. |
| 3368 |
/// |
| 3369 |
/// _Note:_ The output encoding of UTF-16BE, UTF-16LE, and replacement is UTF-8. There |
| 3370 |
/// is no encoder for UTF-16BE, UTF-16LE, and replacement themselves. |
| 3371 |
/// |
| 3372 |
/// Available via the C wrapper. |
| 3373 |
#[inline] |
| 3374 |
pub fn new_encoder(&'static self) -> Encoder { |
| 3375 |
let enc = self.output_encoding(); |
| 3376 |
enc.variant.new_encoder(enc) |
| 3377 |
} |
| 3378 |
|
| 3379 |
/// Validates UTF-8. |
| 3380 |
/// |
| 3381 |
/// Returns the index of the first byte that makes the input malformed as |
| 3382 |
/// UTF-8 or the length of the slice if the slice is entirely valid. |
| 3383 |
/// |
| 3384 |
/// This is currently faster than the corresponding standard library |
| 3385 |
/// functionality. If this implementation gets upstreamed to the standard |
| 3386 |
/// library, this method may be removed in the future. |
| 3387 |
/// |
| 3388 |
/// Available via the C wrapper. |
| 3389 |
pub fn utf8_valid_up_to(bytes: &[u8]) -> usize { |
| 3390 |
utf8_valid_up_to(bytes) |
| 3391 |
} |
| 3392 |
|
| 3393 |
/// Validates ASCII. |
| 3394 |
/// |
| 3395 |
/// Returns the index of the first byte that makes the input malformed as |
| 3396 |
/// ASCII or the length of the slice if the slice is entirely valid. |
| 3397 |
/// |
| 3398 |
/// Available via the C wrapper. |
| 3399 |
pub fn ascii_valid_up_to(bytes: &[u8]) -> usize { |
| 3400 |
ascii_valid_up_to(bytes) |
| 3401 |
} |
| 3402 |
|
| 3403 |
/// Validates ISO-2022-JP ASCII-state data. |
| 3404 |
/// |
| 3405 |
/// Returns the index of the first byte that makes the input not |
| 3406 |
/// representable in the ASCII state of ISO-2022-JP or the length of the |
| 3407 |
/// slice if the slice is entirely representable in the ASCII state of |
| 3408 |
/// ISO-2022-JP. |
| 3409 |
/// |
| 3410 |
/// Available via the C wrapper. |
| 3411 |
pub fn iso_2022_jp_ascii_valid_up_to(bytes: &[u8]) -> usize { |
| 3412 |
iso_2022_jp_ascii_valid_up_to(bytes) |
| 3413 |
} |
| 3414 |
} |
| 3415 |
|
| 3416 |
impl PartialEq for Encoding { |
| 3417 |
#[inline] |
| 3418 |
fn eq(&self, other: &Encoding) -> bool { |
| 3419 |
(self as *const Encoding) == (other as *const Encoding) |
| 3420 |
} |
| 3421 |
} |
| 3422 |
|
| 3423 |
impl Eq for Encoding {} |
| 3424 |
|
| 3425 |
#[cfg(test)] |
| 3426 |
impl PartialOrd for Encoding { |
| 3427 |
fn partial_cmp(&self, other: &Self) -> Option<Ordering> { |
| 3428 |
(self as *const Encoding as usize).partial_cmp(&(other as *const Encoding as usize)) |
| 3429 |
} |
| 3430 |
} |
| 3431 |
|
| 3432 |
#[cfg(test)] |
| 3433 |
impl Ord for Encoding { |
| 3434 |
fn cmp(&self, other: &Self) -> Ordering { |
| 3435 |
(self as *const Encoding as usize).cmp(&(other as *const Encoding as usize)) |
| 3436 |
} |
| 3437 |
} |
| 3438 |
|
| 3439 |
impl Hash for Encoding { |
| 3440 |
#[inline] |
| 3441 |
fn hash<H: Hasher>(&self, state: &mut H) { |
| 3442 |
(self as *const Encoding).hash(state); |
| 3443 |
} |
| 3444 |
} |
| 3445 |
|
| 3446 |
impl core::fmt::Debug for Encoding { |
| 3447 |
#[inline] |
| 3448 |
fn fmt(&self, f: &mut core::fmt::Formatter) -> core::fmt::Result { |
| 3449 |
write!(f, "Encoding {{ {} }}", self.name) |
| 3450 |
} |
| 3451 |
} |
| 3452 |
|
| 3453 |
#[cfg(feature = "serde")] |
| 3454 |
impl Serialize for Encoding { |
| 3455 |
#[inline] |
| 3456 |
fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error> |
| 3457 |
where |
| 3458 |
S: Serializer, |
| 3459 |
{ |
| 3460 |
serializer.serialize_str(self.name) |
| 3461 |
} |
| 3462 |
} |
| 3463 |
|
| 3464 |
#[cfg(feature = "serde")] |
| 3465 |
struct EncodingVisitor; |
| 3466 |
|
| 3467 |
#[cfg(feature = "serde")] |
| 3468 |
impl<'de> Visitor<'de> for EncodingVisitor { |
| 3469 |
type Value = &'static Encoding; |
| 3470 |
|
| 3471 |
fn expecting(&self, formatter: &mut core::fmt::Formatter) -> core::fmt::Result { |
| 3472 |
formatter.write_str("a valid encoding label") |
| 3473 |
} |
| 3474 |
|
| 3475 |
fn visit_str<E>(self, value: &str) -> Result<&'static Encoding, E> |
| 3476 |
where |
| 3477 |
E: serde::de::Error, |
| 3478 |
{ |
| 3479 |
if let Some(enc) = Encoding::for_label(value.as_bytes()) { |
| 3480 |
Ok(enc) |
| 3481 |
} else { |
| 3482 |
Err(E::custom(alloc::format!( |
| 3483 |
"invalid encoding label: {}", |
| 3484 |
value |
| 3485 |
))) |
| 3486 |
} |
| 3487 |
} |
| 3488 |
} |
| 3489 |
|
| 3490 |
#[cfg(feature = "serde")] |
| 3491 |
impl<'de> Deserialize<'de> for &'static Encoding { |
| 3492 |
fn deserialize<D>(deserializer: D) -> Result<&'static Encoding, D::Error> |
| 3493 |
where |
| 3494 |
D: Deserializer<'de>, |
| 3495 |
{ |
| 3496 |
deserializer.deserialize_str(EncodingVisitor) |
| 3497 |
} |
| 3498 |
} |
| 3499 |
|
| 3500 |
/// Tracks the life cycle of a decoder from BOM sniffing to conversion to end. |
| 3501 |
#[derive(PartialEq, Debug, Copy, Clone)] |
| 3502 |
enum DecoderLifeCycle { |
| 3503 |
/// The decoder has seen no input yet. |
| 3504 |
AtStart, |
| 3505 |
/// The decoder has seen no input yet but expects UTF-8. |
| 3506 |
AtUtf8Start, |
| 3507 |
/// The decoder has seen no input yet but expects UTF-16BE. |
| 3508 |
AtUtf16BeStart, |
| 3509 |
/// The decoder has seen no input yet but expects UTF-16LE. |
| 3510 |
AtUtf16LeStart, |
| 3511 |
/// The decoder has seen EF. |
| 3512 |
SeenUtf8First, |
| 3513 |
/// The decoder has seen EF, BB. |
| 3514 |
SeenUtf8Second, |
| 3515 |
/// The decoder has seen FE. |
| 3516 |
SeenUtf16BeFirst, |
| 3517 |
/// The decoder has seen FF. |
| 3518 |
SeenUtf16LeFirst, |
| 3519 |
/// Saw EF, BB but not BF, there was a buffer boundary after BB and the |
| 3520 |
/// underlying decoder reported EF as an error, so we need to remember to |
| 3521 |
/// push BB before the next buffer. |
| 3522 |
ConvertingWithPendingBB, |
| 3523 |
/// No longer looking for a BOM and EOF not yet seen. |
| 3524 |
Converting, |
| 3525 |
/// EOF has been seen. |
| 3526 |
Finished, |
| 3527 |
} |
| 3528 |
|
| 3529 |
/// Communicate the BOM handling mode. |
| 3530 |
#[derive(Debug, Copy, Clone)] |
| 3531 |
enum BomHandling { |
| 3532 |
/// Don't handle the BOM |
| 3533 |
Off, |
| 3534 |
/// Sniff for UTF-8, UTF-16BE or UTF-16LE BOM |
| 3535 |
Sniff, |
| 3536 |
/// Remove the BOM only if it's the BOM for this encoding |
| 3537 |
Remove, |
| 3538 |
} |
| 3539 |
|
| 3540 |
/// Result of a (potentially partial) decode or encode operation with |
| 3541 |
/// replacement. |
| 3542 |
#[must_use] |
| 3543 |
#[derive(Debug, PartialEq, Eq)] |
| 3544 |
pub enum CoderResult { |
| 3545 |
/// The input was exhausted. |
| 3546 |
/// |
| 3547 |
/// If this result was returned from a call where `last` was `true`, the |
| 3548 |
/// conversion process has completed. Otherwise, the caller should call a |
| 3549 |
/// decode or encode method again with more input. |
| 3550 |
InputEmpty, |
| 3551 |
|
| 3552 |
/// The converter cannot produce another unit of output, because the output |
| 3553 |
/// buffer does not have enough space left. |
| 3554 |
/// |
| 3555 |
/// The caller must provide more output space upon the next call and re-push |
| 3556 |
/// the remaining input to the converter. |
| 3557 |
OutputFull, |
| 3558 |
} |
| 3559 |
|
| 3560 |
/// Result of a (potentially partial) decode operation without replacement. |
| 3561 |
#[must_use] |
| 3562 |
#[derive(Debug, PartialEq, Eq)] |
| 3563 |
pub enum DecoderResult { |
| 3564 |
/// The input was exhausted. |
| 3565 |
/// |
| 3566 |
/// If this result was returned from a call where `last` was `true`, the |
| 3567 |
/// decoding process has completed. Otherwise, the caller should call a |
| 3568 |
/// decode method again with more input. |
| 3569 |
InputEmpty, |
| 3570 |
|
| 3571 |
/// The decoder cannot produce another unit of output, because the output |
| 3572 |
/// buffer does not have enough space left. |
| 3573 |
/// |
| 3574 |
/// The caller must provide more output space upon the next call and re-push |
| 3575 |
/// the remaining input to the decoder. |
| 3576 |
OutputFull, |
| 3577 |
|
| 3578 |
/// The decoder encountered a malformed byte sequence. |
| 3579 |
/// |
| 3580 |
/// The caller must either treat this as a fatal error or must append one |
| 3581 |
/// REPLACEMENT CHARACTER (U+FFFD) to the output and then re-push the |
| 3582 |
/// the remaining input to the decoder. |
| 3583 |
/// |
| 3584 |
/// The first wrapped integer indicates the length of the malformed byte |
| 3585 |
/// sequence. The second wrapped integer indicates the number of bytes |
| 3586 |
/// that were consumed after the malformed sequence. If the second |
| 3587 |
/// integer is zero, the last byte that was consumed is the last byte of |
| 3588 |
/// the malformed sequence. Note that the malformed bytes may have been part |
| 3589 |
/// of an earlier input buffer. |
| 3590 |
/// |
| 3591 |
/// The first wrapped integer can have values 1, 2, 3 or 4. The second |
| 3592 |
/// wrapped integer can have values 0, 1, 2 or 3. The worst-case sum |
| 3593 |
/// of the two is 6, which happens with ISO-2022-JP. |
| 3594 |
Malformed(u8, u8), // u8 instead of usize to avoid useless bloat |
| 3595 |
} |
| 3596 |
|
| 3597 |
/// A converter that decodes a byte stream into Unicode according to a |
| 3598 |
/// character encoding in a streaming (incremental) manner. |
| 3599 |
/// |
| 3600 |
/// The various `decode_*` methods take an input buffer (`src`) and an output |
| 3601 |
/// buffer `dst` both of which are caller-allocated. There are variants for |
| 3602 |
/// both UTF-8 and UTF-16 output buffers. |
| 3603 |
/// |
| 3604 |
/// A `decode_*` method decodes bytes from `src` into Unicode characters stored |
| 3605 |
/// into `dst` until one of the following three things happens: |
| 3606 |
/// |
| 3607 |
/// 1. A malformed byte sequence is encountered (`*_without_replacement` |
| 3608 |
/// variants only). |
| 3609 |
/// |
| 3610 |
/// 2. The output buffer has been filled so near capacity that the decoder |
| 3611 |
/// cannot be sure that processing an additional byte of input wouldn't |
| 3612 |
/// cause so much output that the output buffer would overflow. |
| 3613 |
/// |
| 3614 |
/// 3. All the input bytes have been processed. |
| 3615 |
/// |
| 3616 |
/// The `decode_*` method then returns tuple of a status indicating which one |
| 3617 |
/// of the three reasons to return happened, how many input bytes were read, |
| 3618 |
/// how many output code units (`u8` when decoding into UTF-8 and `u16` |
| 3619 |
/// when decoding to UTF-16) were written (except when decoding into `String`, |
| 3620 |
/// whose length change indicates this), and in the case of the |
| 3621 |
/// variants performing replacement, a boolean indicating whether an error was |
| 3622 |
/// replaced with the REPLACEMENT CHARACTER during the call. |
| 3623 |
/// |
| 3624 |
/// The number of bytes "written" is what's logically written. Garbage may be |
| 3625 |
/// written in the output buffer beyond the point logically written to. |
| 3626 |
/// Therefore, if you wish to decode into an `&mut str`, you should use the |
| 3627 |
/// methods that take an `&mut str` argument instead of the ones that take an |
| 3628 |
/// `&mut [u8]` argument. The former take care of overwriting the trailing |
| 3629 |
/// garbage to ensure the UTF-8 validity of the `&mut str` as a whole, but the |
| 3630 |
/// latter don't. |
| 3631 |
/// |
| 3632 |
/// In the case of the `*_without_replacement` variants, the status is a |
| 3633 |
/// [`DecoderResult`][1] enumeration (possibilities `Malformed`, `OutputFull` and |
| 3634 |
/// `InputEmpty` corresponding to the three cases listed above). |
| 3635 |
/// |
| 3636 |
/// In the case of methods whose name does not end with |
| 3637 |
/// `*_without_replacement`, malformed sequences are automatically replaced |
| 3638 |
/// with the REPLACEMENT CHARACTER and errors do not cause the methods to |
| 3639 |
/// return early. |
| 3640 |
/// |
| 3641 |
/// When decoding to UTF-8, the output buffer must have at least 4 bytes of |
| 3642 |
/// space. When decoding to UTF-16, the output buffer must have at least two |
| 3643 |
/// UTF-16 code units (`u16`) of space. |
| 3644 |
/// |
| 3645 |
/// When decoding to UTF-8 without replacement, the methods are guaranteed |
| 3646 |
/// not to return indicating that more output space is needed if the length |
| 3647 |
/// of the output buffer is at least the length returned by |
| 3648 |
/// [`max_utf8_buffer_length_without_replacement()`][2]. When decoding to UTF-8 |
| 3649 |
/// with replacement, the length of the output buffer that guarantees the |
| 3650 |
/// methods not to return indicating that more output space is needed is given |
| 3651 |
/// by [`max_utf8_buffer_length()`][3]. When decoding to UTF-16 with |
| 3652 |
/// or without replacement, the length of the output buffer that guarantees |
| 3653 |
/// the methods not to return indicating that more output space is needed is |
| 3654 |
/// given by [`max_utf16_buffer_length()`][4]. |
| 3655 |
/// |
| 3656 |
/// The output written into `dst` is guaranteed to be valid UTF-8 or UTF-16, |
| 3657 |
/// and the output after each `decode_*` call is guaranteed to consist of |
| 3658 |
/// complete characters. (I.e. the code unit sequence for the last character is |
| 3659 |
/// guaranteed not to be split across output buffers.) |
| 3660 |
/// |
| 3661 |
/// The boolean argument `last` indicates that the end of the stream is reached |
| 3662 |
/// when all the bytes in `src` have been consumed. |
| 3663 |
/// |
| 3664 |
/// A `Decoder` object can be used to incrementally decode a byte stream. |
| 3665 |
/// |
| 3666 |
/// During the processing of a single stream, the caller must call `decode_*` |
| 3667 |
/// zero or more times with `last` set to `false` and then call `decode_*` at |
| 3668 |
/// least once with `last` set to `true`. If `decode_*` returns `InputEmpty`, |
| 3669 |
/// the processing of the stream has ended. Otherwise, the caller must call |
| 3670 |
/// `decode_*` again with `last` set to `true` (or treat a `Malformed` result as |
| 3671 |
/// a fatal error). |
| 3672 |
/// |
| 3673 |
/// Once the stream has ended, the `Decoder` object must not be used anymore. |
| 3674 |
/// That is, you need to create another one to process another stream. |
| 3675 |
/// |
| 3676 |
/// When the decoder returns `OutputFull` or the decoder returns `Malformed` and |
| 3677 |
/// the caller does not wish to treat it as a fatal error, the input buffer |
| 3678 |
/// `src` may not have been completely consumed. In that case, the caller must |
| 3679 |
/// pass the unconsumed contents of `src` to `decode_*` again upon the next |
| 3680 |
/// call. |
| 3681 |
/// |
| 3682 |
/// [1]: enum.DecoderResult.html |
| 3683 |
/// [2]: #method.max_utf8_buffer_length_without_replacement |
| 3684 |
/// [3]: #method.max_utf8_buffer_length |
| 3685 |
/// [4]: #method.max_utf16_buffer_length |
| 3686 |
/// |
| 3687 |
/// # Infinite loops |
| 3688 |
/// |
| 3689 |
/// When converting with a fixed-size output buffer whose size is too small to |
| 3690 |
/// accommodate one character or (when applicable) one numeric character |
| 3691 |
/// reference of output, an infinite loop ensues. When converting with a |
| 3692 |
/// fixed-size output buffer, it generally makes sense to make the buffer |
| 3693 |
/// fairly large (e.g. couple of kilobytes). |
| 3694 |
pub struct Decoder { |
| 3695 |
encoding: &'static Encoding, |
| 3696 |
variant: VariantDecoder, |
| 3697 |
life_cycle: DecoderLifeCycle, |
| 3698 |
} |
| 3699 |
|
| 3700 |
impl Decoder { |
| 3701 |
fn new(enc: &'static Encoding, decoder: VariantDecoder, sniffing: BomHandling) -> Decoder { |
| 3702 |
Decoder { |
| 3703 |
encoding: enc, |
| 3704 |
variant: decoder, |
| 3705 |
life_cycle: match sniffing { |
| 3706 |
BomHandling::Off => DecoderLifeCycle::Converting, |
| 3707 |
BomHandling::Sniff => DecoderLifeCycle::AtStart, |
| 3708 |
BomHandling::Remove => { |
| 3709 |
if enc == UTF_8 { |
| 3710 |
DecoderLifeCycle::AtUtf8Start |
| 3711 |
} else if enc == UTF_16BE { |
| 3712 |
DecoderLifeCycle::AtUtf16BeStart |
| 3713 |
} else if enc == UTF_16LE { |
| 3714 |
DecoderLifeCycle::AtUtf16LeStart |
| 3715 |
} else { |
| 3716 |
DecoderLifeCycle::Converting |
| 3717 |
} |
| 3718 |
} |
| 3719 |
}, |
| 3720 |
} |
| 3721 |
} |
| 3722 |
|
| 3723 |
/// The `Encoding` this `Decoder` is for. |
| 3724 |
/// |
| 3725 |
/// BOM sniffing can change the return value of this method during the life |
| 3726 |
/// of the decoder. |
| 3727 |
/// |
| 3728 |
/// Available via the C wrapper. |
| 3729 |
#[inline] |
| 3730 |
pub fn encoding(&self) -> &'static Encoding { |
| 3731 |
self.encoding |
| 3732 |
} |
| 3733 |
|
| 3734 |
/// Query the worst-case UTF-8 output size _with replacement_. |
| 3735 |
/// |
| 3736 |
/// Returns the size of the output buffer in UTF-8 code units (`u8`) |
| 3737 |
/// that will not overflow given the current state of the decoder and |
| 3738 |
/// `byte_length` number of additional input bytes when decoding with |
| 3739 |
/// errors handled by outputting a REPLACEMENT CHARACTER for each malformed |
| 3740 |
/// sequence or `None` if `usize` would overflow. |
| 3741 |
/// |
| 3742 |
/// Available via the C wrapper. |
| 3743 |
pub fn max_utf8_buffer_length(&self, byte_length: usize) -> Option<usize> { |
| 3744 |
// Need to consider a) the decoder morphing due to the BOM and b) a partial |
| 3745 |
// BOM getting pushed to the underlying decoder. |
| 3746 |
match self.life_cycle { |
| 3747 |
DecoderLifeCycle::Converting |
| 3748 |
| DecoderLifeCycle::AtUtf8Start |
| 3749 |
| DecoderLifeCycle::AtUtf16LeStart |
| 3750 |
| DecoderLifeCycle::AtUtf16BeStart => { |
| 3751 |
return self.variant.max_utf8_buffer_length(byte_length); |
| 3752 |
} |
| 3753 |
DecoderLifeCycle::AtStart => { |
| 3754 |
if let Some(utf8_bom) = checked_add(3, byte_length.checked_mul(3)) { |
| 3755 |
if let Some(utf16_bom) = checked_add( |
| 3756 |
1, |
| 3757 |
checked_mul(3, checked_div(byte_length.checked_add(1), 2)), |
| 3758 |
) { |
| 3759 |
let utf_bom = core::cmp::max(utf8_bom, utf16_bom); |
| 3760 |
let encoding = self.encoding(); |
| 3761 |
if encoding == UTF_8 || encoding == UTF_16LE || encoding == UTF_16BE { |
| 3762 |
// No need to consider the internal state of the underlying decoder, |
| 3763 |
// because it is at start, because no data has reached it yet. |
| 3764 |
return Some(utf_bom); |
| 3765 |
} else if let Some(non_bom) = |
| 3766 |
self.variant.max_utf8_buffer_length(byte_length) |
| 3767 |
{ |
| 3768 |
return Some(core::cmp::max(utf_bom, non_bom)); |
| 3769 |
} |
| 3770 |
} |
| 3771 |
} |
| 3772 |
} |
| 3773 |
DecoderLifeCycle::SeenUtf8First | DecoderLifeCycle::SeenUtf8Second => { |
| 3774 |
// Add two bytes even when only one byte has been seen, |
| 3775 |
// because the one byte can become a lead byte in multibyte |
| 3776 |
// decoders, but only after the decoder has been queried |
| 3777 |
// for max length, so the decoder's own logic for adding |
| 3778 |
// one for a pending lead cannot work. |
| 3779 |
if let Some(sum) = byte_length.checked_add(2) { |
| 3780 |
if let Some(utf8_bom) = checked_add(3, sum.checked_mul(3)) { |
| 3781 |
if self.encoding() == UTF_8 { |
| 3782 |
// No need to consider the internal state of the underlying decoder, |
| 3783 |
// because it is at start, because no data has reached it yet. |
| 3784 |
return Some(utf8_bom); |
| 3785 |
} else if let Some(non_bom) = self.variant.max_utf8_buffer_length(sum) { |
| 3786 |
return Some(core::cmp::max(utf8_bom, non_bom)); |
| 3787 |
} |
| 3788 |
} |
| 3789 |
} |
| 3790 |
} |
| 3791 |
DecoderLifeCycle::ConvertingWithPendingBB => { |
| 3792 |
if let Some(sum) = byte_length.checked_add(2) { |
| 3793 |
return self.variant.max_utf8_buffer_length(sum); |
| 3794 |
} |
| 3795 |
} |
| 3796 |
DecoderLifeCycle::SeenUtf16LeFirst | DecoderLifeCycle::SeenUtf16BeFirst => { |
| 3797 |
// Add two bytes even when only one byte has been seen, |
| 3798 |
// because the one byte can become a lead byte in multibyte |
| 3799 |
// decoders, but only after the decoder has been queried |
| 3800 |
// for max length, so the decoder's own logic for adding |
| 3801 |
// one for a pending lead cannot work. |
| 3802 |
if let Some(sum) = byte_length.checked_add(2) { |
| 3803 |
if let Some(utf16_bom) = |
| 3804 |
checked_add(1, checked_mul(3, checked_div(sum.checked_add(1), 2))) |
| 3805 |
{ |
| 3806 |
let encoding = self.encoding(); |
| 3807 |
if encoding == UTF_16LE || encoding == UTF_16BE { |
| 3808 |
// No need to consider the internal state of the underlying decoder, |
| 3809 |
// because it is at start, because no data has reached it yet. |
| 3810 |
return Some(utf16_bom); |
| 3811 |
} else if let Some(non_bom) = self.variant.max_utf8_buffer_length(sum) { |
| 3812 |
return Some(core::cmp::max(utf16_bom, non_bom)); |
| 3813 |
} |
| 3814 |
} |
| 3815 |
} |
| 3816 |
} |
| 3817 |
DecoderLifeCycle::Finished => panic!("Must not use a decoder that has finished."), |
| 3818 |
} |
| 3819 |
None |
| 3820 |
} |
| 3821 |
|
| 3822 |
/// Query the worst-case UTF-8 output size _without replacement_. |
| 3823 |
/// |
| 3824 |
/// Returns the size of the output buffer in UTF-8 code units (`u8`) |
| 3825 |
/// that will not overflow given the current state of the decoder and |
| 3826 |
/// `byte_length` number of additional input bytes when decoding without |
| 3827 |
/// replacement error handling or `None` if `usize` would overflow. |
| 3828 |
/// |
| 3829 |
/// Note that this value may be too small for the `_with_replacement` case. |
| 3830 |
/// Use `max_utf8_buffer_length()` for that case. |
| 3831 |
/// |
| 3832 |
/// Available via the C wrapper. |
| 3833 |
pub fn max_utf8_buffer_length_without_replacement(&self, byte_length: usize) -> Option<usize> { |
| 3834 |
// Need to consider a) the decoder morphing due to the BOM and b) a partial |
| 3835 |
// BOM getting pushed to the underlying decoder. |
| 3836 |
match self.life_cycle { |
| 3837 |
DecoderLifeCycle::Converting |
| 3838 |
| DecoderLifeCycle::AtUtf8Start |
| 3839 |
| DecoderLifeCycle::AtUtf16LeStart |
| 3840 |
| DecoderLifeCycle::AtUtf16BeStart => { |
| 3841 |
return self |
| 3842 |
.variant |
| 3843 |
.max_utf8_buffer_length_without_replacement(byte_length); |
| 3844 |
} |
| 3845 |
DecoderLifeCycle::AtStart => { |
| 3846 |
if let Some(utf8_bom) = byte_length.checked_add(3) { |
| 3847 |
if let Some(utf16_bom) = checked_add( |
| 3848 |
1, |
| 3849 |
checked_mul(3, checked_div(byte_length.checked_add(1), 2)), |
| 3850 |
) { |
| 3851 |
let utf_bom = core::cmp::max(utf8_bom, utf16_bom); |
| 3852 |
let encoding = self.encoding(); |
| 3853 |
if encoding == UTF_8 || encoding == UTF_16LE || encoding == UTF_16BE { |
| 3854 |
// No need to consider the internal state of the underlying decoder, |
| 3855 |
// because it is at start, because no data has reached it yet. |
| 3856 |
return Some(utf_bom); |
| 3857 |
} else if let Some(non_bom) = self |
| 3858 |
.variant |
| 3859 |
.max_utf8_buffer_length_without_replacement(byte_length) |
| 3860 |
{ |
| 3861 |
return Some(core::cmp::max(utf_bom, non_bom)); |
| 3862 |
} |
| 3863 |
} |
| 3864 |
} |
| 3865 |
} |
| 3866 |
DecoderLifeCycle::SeenUtf8First | DecoderLifeCycle::SeenUtf8Second => { |
| 3867 |
// Add two bytes even when only one byte has been seen, |
| 3868 |
// because the one byte can become a lead byte in multibyte |
| 3869 |
// decoders, but only after the decoder has been queried |
| 3870 |
// for max length, so the decoder's own logic for adding |
| 3871 |
// one for a pending lead cannot work. |
| 3872 |
if let Some(sum) = byte_length.checked_add(2) { |
| 3873 |
if let Some(utf8_bom) = sum.checked_add(3) { |
| 3874 |
if self.encoding() == UTF_8 { |
| 3875 |
// No need to consider the internal state of the underlying decoder, |
| 3876 |
// because it is at start, because no data has reached it yet. |
| 3877 |
return Some(utf8_bom); |
| 3878 |
} else if let Some(non_bom) = |
| 3879 |
self.variant.max_utf8_buffer_length_without_replacement(sum) |
| 3880 |
{ |
| 3881 |
return Some(core::cmp::max(utf8_bom, non_bom)); |
| 3882 |
} |
| 3883 |
} |
| 3884 |
} |
| 3885 |
} |
| 3886 |
DecoderLifeCycle::ConvertingWithPendingBB => { |
| 3887 |
if let Some(sum) = byte_length.checked_add(2) { |
| 3888 |
return self.variant.max_utf8_buffer_length_without_replacement(sum); |
| 3889 |
} |
| 3890 |
} |
| 3891 |
DecoderLifeCycle::SeenUtf16LeFirst | DecoderLifeCycle::SeenUtf16BeFirst => { |
| 3892 |
// Add two bytes even when only one byte has been seen, |
| 3893 |
// because the one byte can become a lead byte in multibyte |
| 3894 |
// decoders, but only after the decoder has been queried |
| 3895 |
// for max length, so the decoder's own logic for adding |
| 3896 |
// one for a pending lead cannot work. |
| 3897 |
if let Some(sum) = byte_length.checked_add(2) { |
| 3898 |
if let Some(utf16_bom) = |
| 3899 |
checked_add(1, checked_mul(3, checked_div(sum.checked_add(1), 2))) |
| 3900 |
{ |
| 3901 |
let encoding = self.encoding(); |
| 3902 |
if encoding == UTF_16LE || encoding == UTF_16BE { |
| 3903 |
// No need to consider the internal state of the underlying decoder, |
| 3904 |
// because it is at start, because no data has reached it yet. |
| 3905 |
return Some(utf16_bom); |
| 3906 |
} else if let Some(non_bom) = |
| 3907 |
self.variant.max_utf8_buffer_length_without_replacement(sum) |
| 3908 |
{ |
| 3909 |
return Some(core::cmp::max(utf16_bom, non_bom)); |
| 3910 |
} |
| 3911 |
} |
| 3912 |
} |
| 3913 |
} |
| 3914 |
DecoderLifeCycle::Finished => panic!("Must not use a decoder that has finished."), |
| 3915 |
} |
| 3916 |
None |
| 3917 |
} |
| 3918 |
|
| 3919 |
/// Incrementally decode a byte stream into UTF-8 with malformed sequences |
| 3920 |
/// replaced with the REPLACEMENT CHARACTER. |
| 3921 |
/// |
| 3922 |
/// See the documentation of the struct for documentation for `decode_*` |
| 3923 |
/// methods collectively. |
| 3924 |
/// |
| 3925 |
/// Available via the C wrapper. |
| 3926 |
pub fn decode_to_utf8( |
| 3927 |
&mut self, |
| 3928 |
src: &[u8], |
| 3929 |
dst: &mut [u8], |
| 3930 |
last: bool, |
| 3931 |
) -> (CoderResult, usize, usize, bool) { |
| 3932 |
let mut had_errors = false; |
| 3933 |
let mut total_read = 0usize; |
| 3934 |
let mut total_written = 0usize; |
| 3935 |
loop { |
| 3936 |
let (result, read, written) = self.decode_to_utf8_without_replacement( |
| 3937 |
&src[total_read..], |
| 3938 |
&mut dst[total_written..], |
| 3939 |
last, |
| 3940 |
); |
| 3941 |
total_read += read; |
| 3942 |
total_written += written; |
| 3943 |
match result { |
| 3944 |
DecoderResult::InputEmpty => { |
| 3945 |
return ( |
| 3946 |
CoderResult::InputEmpty, |
| 3947 |
total_read, |
| 3948 |
total_written, |
| 3949 |
had_errors, |
| 3950 |
); |
| 3951 |
} |
| 3952 |
DecoderResult::OutputFull => { |
| 3953 |
return ( |
| 3954 |
CoderResult::OutputFull, |
| 3955 |
total_read, |
| 3956 |
total_written, |
| 3957 |
had_errors, |
| 3958 |
); |
| 3959 |
} |
| 3960 |
DecoderResult::Malformed(_, _) => { |
| 3961 |
had_errors = true; |
| 3962 |
// There should always be space for the U+FFFD, because |
| 3963 |
// otherwise we'd have gotten OutputFull already. |
| 3964 |
// XXX: is the above comment actually true for UTF-8 itself? |
| 3965 |
// TODO: Consider having fewer bound checks here. |
| 3966 |
dst[total_written] = 0xEFu8; |
| 3967 |
total_written += 1; |
| 3968 |
dst[total_written] = 0xBFu8; |
| 3969 |
total_written += 1; |
| 3970 |
dst[total_written] = 0xBDu8; |
| 3971 |
total_written += 1; |
| 3972 |
} |
| 3973 |
} |
| 3974 |
} |
| 3975 |
} |
| 3976 |
|
| 3977 |
/// Incrementally decode a byte stream into UTF-8 with malformed sequences |
| 3978 |
/// replaced with the REPLACEMENT CHARACTER with type system signaling |
| 3979 |
/// of UTF-8 validity. |
| 3980 |
/// |
| 3981 |
/// This methods calls `decode_to_utf8` and then zeroes |
| 3982 |
/// out up to three bytes that aren't logically part of the write in order |
| 3983 |
/// to retain the UTF-8 validity even for the unwritten part of the buffer. |
| 3984 |
/// |
| 3985 |
/// See the documentation of the struct for documentation for `decode_*` |
| 3986 |
/// methods collectively. |
| 3987 |
/// |
| 3988 |
/// Available to Rust only. |
| 3989 |
pub fn decode_to_str( |
| 3990 |
&mut self, |
| 3991 |
src: &[u8], |
| 3992 |
dst: &mut str, |
| 3993 |
last: bool, |
| 3994 |
) -> (CoderResult, usize, usize, bool) { |
| 3995 |
let bytes: &mut [u8] = unsafe { dst.as_bytes_mut() }; |
| 3996 |
let (result, read, written, replaced) = self.decode_to_utf8(src, bytes, last); |
| 3997 |
let len = bytes.len(); |
| 3998 |
let mut trail = written; |
| 3999 |
// Non-UTF-8 ASCII-compatible decoders may write up to `MAX_STRIDE_SIZE` |
| 4000 |
// bytes of trailing garbage. No need to optimize non-ASCII-compatible |
| 4001 |
// encodings to avoid overwriting here. |
| 4002 |
if self.encoding != UTF_8 { |
| 4003 |
let max = core::cmp::min(len, trail + ascii::MAX_STRIDE_SIZE); |
| 4004 |
while trail < max { |
| 4005 |
bytes[trail] = 0; |
| 4006 |
trail += 1; |
| 4007 |
} |
| 4008 |
} |
| 4009 |
while trail < len && ((bytes[trail] & 0xC0) == 0x80) { |
| 4010 |
bytes[trail] = 0; |
| 4011 |
trail += 1; |
| 4012 |
} |
| 4013 |
(result, read, written, replaced) |
| 4014 |
} |
| 4015 |
|
| 4016 |
/// Incrementally decode a byte stream into UTF-8 with malformed sequences |
| 4017 |
/// replaced with the REPLACEMENT CHARACTER using a `String` receiver. |
| 4018 |
/// |
| 4019 |
/// Like the others, this method follows the logic that the output buffer is |
| 4020 |
/// caller-allocated. This method treats the capacity of the `String` as |
| 4021 |
/// the output limit. That is, this method guarantees not to cause a |
| 4022 |
/// reallocation of the backing buffer of `String`. |
| 4023 |
/// |
| 4024 |
/// The return value is a tuple that contains the `DecoderResult`, the |
| 4025 |
/// number of bytes read and a boolean indicating whether replacements |
| 4026 |
/// were done. The number of bytes written is signaled via the length of |
| 4027 |
/// the `String` changing. |
| 4028 |
/// |
| 4029 |
/// See the documentation of the struct for documentation for `decode_*` |
| 4030 |
/// methods collectively. |
| 4031 |
/// |
| 4032 |
/// Available to Rust only and only with the `alloc` feature enabled (enabled |
| 4033 |
/// by default). |
| 4034 |
#[cfg(feature = "alloc")] |
| 4035 |
pub fn decode_to_string( |
| 4036 |
&mut self, |
| 4037 |
src: &[u8], |
| 4038 |
dst: &mut String, |
| 4039 |
last: bool, |
| 4040 |
) -> (CoderResult, usize, bool) { |
| 4041 |
unsafe { |
| 4042 |
let vec = dst.as_mut_vec(); |
| 4043 |
let old_len = vec.len(); |
| 4044 |
let capacity = vec.capacity(); |
| 4045 |
vec.set_len(capacity); |
| 4046 |
let (result, read, written, replaced) = |
| 4047 |
self.decode_to_utf8(src, &mut vec[old_len..], last); |
| 4048 |
vec.set_len(old_len + written); |
| 4049 |
(result, read, replaced) |
| 4050 |
} |
| 4051 |
} |
| 4052 |
|
| 4053 |
public_decode_function!(/// Incrementally decode a byte stream into UTF-8 |
| 4054 |
/// _without replacement_. |
| 4055 |
/// |
| 4056 |
/// See the documentation of the struct for |
| 4057 |
/// documentation for `decode_*` methods |
| 4058 |
/// collectively. |
| 4059 |
/// |
| 4060 |
/// Available via the C wrapper. |
| 4061 |
, |
| 4062 |
decode_to_utf8_without_replacement, |
| 4063 |
decode_to_utf8_raw, |
| 4064 |
decode_to_utf8_checking_end, |
| 4065 |
decode_to_utf8_after_one_potential_bom_byte, |
| 4066 |
decode_to_utf8_after_two_potential_bom_bytes, |
| 4067 |
decode_to_utf8_checking_end_with_offset, |
| 4068 |
u8); |
| 4069 |
|
| 4070 |
/// Incrementally decode a byte stream into UTF-8 with type system signaling |
| 4071 |
/// of UTF-8 validity. |
| 4072 |
/// |
| 4073 |
/// This methods calls `decode_to_utf8` and then zeroes out up to three |
| 4074 |
/// bytes that aren't logically part of the write in order to retain the |
| 4075 |
/// UTF-8 validity even for the unwritten part of the buffer. |
| 4076 |
/// |
| 4077 |
/// See the documentation of the struct for documentation for `decode_*` |
| 4078 |
/// methods collectively. |
| 4079 |
/// |
| 4080 |
/// Available to Rust only. |
| 4081 |
pub fn decode_to_str_without_replacement( |
| 4082 |
&mut self, |
| 4083 |
src: &[u8], |
| 4084 |
dst: &mut str, |
| 4085 |
last: bool, |
| 4086 |
) -> (DecoderResult, usize, usize) { |
| 4087 |
let bytes: &mut [u8] = unsafe { dst.as_bytes_mut() }; |
| 4088 |
let (result, read, written) = self.decode_to_utf8_without_replacement(src, bytes, last); |
| 4089 |
let len = bytes.len(); |
| 4090 |
let mut trail = written; |
| 4091 |
// Non-UTF-8 ASCII-compatible decoders may write up to `MAX_STRIDE_SIZE` |
| 4092 |
// bytes of trailing garbage. No need to optimize non-ASCII-compatible |
| 4093 |
// encodings to avoid overwriting here. |
| 4094 |
if self.encoding != UTF_8 { |
| 4095 |
let max = core::cmp::min(len, trail + ascii::MAX_STRIDE_SIZE); |
| 4096 |
while trail < max { |
| 4097 |
bytes[trail] = 0; |
| 4098 |
trail += 1; |
| 4099 |
} |
| 4100 |
} |
| 4101 |
while trail < len && ((bytes[trail] & 0xC0) == 0x80) { |
| 4102 |
bytes[trail] = 0; |
| 4103 |
trail += 1; |
| 4104 |
} |
| 4105 |
(result, read, written) |
| 4106 |
} |
| 4107 |
|
| 4108 |
/// Incrementally decode a byte stream into UTF-8 using a `String` receiver. |
| 4109 |
/// |
| 4110 |
/// Like the others, this method follows the logic that the output buffer is |
| 4111 |
/// caller-allocated. This method treats the capacity of the `String` as |
| 4112 |
/// the output limit. That is, this method guarantees not to cause a |
| 4113 |
/// reallocation of the backing buffer of `String`. |
| 4114 |
/// |
| 4115 |
/// The return value is a pair that contains the `DecoderResult` and the |
| 4116 |
/// number of bytes read. The number of bytes written is signaled via |
| 4117 |
/// the length of the `String` changing. |
| 4118 |
/// |
| 4119 |
/// See the documentation of the struct for documentation for `decode_*` |
| 4120 |
/// methods collectively. |
| 4121 |
/// |
| 4122 |
/// Available to Rust only and only with the `alloc` feature enabled (enabled |
| 4123 |
/// by default). |
| 4124 |
#[cfg(feature = "alloc")] |
| 4125 |
pub fn decode_to_string_without_replacement( |
| 4126 |
&mut self, |
| 4127 |
src: &[u8], |
| 4128 |
dst: &mut String, |
| 4129 |
last: bool, |
| 4130 |
) -> (DecoderResult, usize) { |
| 4131 |
unsafe { |
| 4132 |
let vec = dst.as_mut_vec(); |
| 4133 |
let old_len = vec.len(); |
| 4134 |
let capacity = vec.capacity(); |
| 4135 |
vec.set_len(capacity); |
| 4136 |
let (result, read, written) = |
| 4137 |
self.decode_to_utf8_without_replacement(src, &mut vec[old_len..], last); |
| 4138 |
vec.set_len(old_len + written); |
| 4139 |
(result, read) |
| 4140 |
} |
| 4141 |
} |
| 4142 |
|
| 4143 |
/// Query the worst-case UTF-16 output size (with or without replacement). |
| 4144 |
/// |
| 4145 |
/// Returns the size of the output buffer in UTF-16 code units (`u16`) |
| 4146 |
/// that will not overflow given the current state of the decoder and |
| 4147 |
/// `byte_length` number of additional input bytes or `None` if `usize` |
| 4148 |
/// would overflow. |
| 4149 |
/// |
| 4150 |
/// Since the REPLACEMENT CHARACTER fits into one UTF-16 code unit, the |
| 4151 |
/// return value of this method applies also in the |
| 4152 |
/// `_without_replacement` case. |
| 4153 |
/// |
| 4154 |
/// Available via the C wrapper. |
| 4155 |
pub fn max_utf16_buffer_length(&self, byte_length: usize) -> Option<usize> { |
| 4156 |
// Need to consider a) the decoder morphing due to the BOM and b) a partial |
| 4157 |
// BOM getting pushed to the underlying decoder. |
| 4158 |
match self.life_cycle { |
| 4159 |
DecoderLifeCycle::Converting |
| 4160 |
| DecoderLifeCycle::AtUtf8Start |
| 4161 |
| DecoderLifeCycle::AtUtf16LeStart |
| 4162 |
| DecoderLifeCycle::AtUtf16BeStart => { |
| 4163 |
return self.variant.max_utf16_buffer_length(byte_length); |
| 4164 |
} |
| 4165 |
DecoderLifeCycle::AtStart => { |
| 4166 |
if let Some(utf8_bom) = byte_length.checked_add(1) { |
| 4167 |
if let Some(utf16_bom) = |
| 4168 |
checked_add(1, checked_div(byte_length.checked_add(1), 2)) |
| 4169 |
{ |
| 4170 |
let utf_bom = core::cmp::max(utf8_bom, utf16_bom); |
| 4171 |
let encoding = self.encoding(); |
| 4172 |
if encoding == UTF_8 || encoding == UTF_16LE || encoding == UTF_16BE { |
| 4173 |
// No need to consider the internal state of the underlying decoder, |
| 4174 |
// because it is at start, because no data has reached it yet. |
| 4175 |
return Some(utf_bom); |
| 4176 |
} else if let Some(non_bom) = |
| 4177 |
self.variant.max_utf16_buffer_length(byte_length) |
| 4178 |
{ |
| 4179 |
return Some(core::cmp::max(utf_bom, non_bom)); |
| 4180 |
} |
| 4181 |
} |
| 4182 |
} |
| 4183 |
} |
| 4184 |
DecoderLifeCycle::SeenUtf8First | DecoderLifeCycle::SeenUtf8Second => { |
| 4185 |
// Add two bytes even when only one byte has been seen, |
| 4186 |
// because the one byte can become a lead byte in multibyte |
| 4187 |
// decoders, but only after the decoder has been queried |
| 4188 |
// for max length, so the decoder's own logic for adding |
| 4189 |
// one for a pending lead cannot work. |
| 4190 |
if let Some(sum) = byte_length.checked_add(2) { |
| 4191 |
if let Some(utf8_bom) = sum.checked_add(1) { |
| 4192 |
if self.encoding() == UTF_8 { |
| 4193 |
// No need to consider the internal state of the underlying decoder, |
| 4194 |
// because it is at start, because no data has reached it yet. |
| 4195 |
return Some(utf8_bom); |
| 4196 |
} else if let Some(non_bom) = self.variant.max_utf16_buffer_length(sum) { |
| 4197 |
return Some(core::cmp::max(utf8_bom, non_bom)); |
| 4198 |
} |
| 4199 |
} |
| 4200 |
} |
| 4201 |
} |
| 4202 |
DecoderLifeCycle::ConvertingWithPendingBB => { |
| 4203 |
if let Some(sum) = byte_length.checked_add(2) { |
| 4204 |
return self.variant.max_utf16_buffer_length(sum); |
| 4205 |
} |
| 4206 |
} |
| 4207 |
DecoderLifeCycle::SeenUtf16LeFirst | DecoderLifeCycle::SeenUtf16BeFirst => { |
| 4208 |
// Add two bytes even when only one byte has been seen, |
| 4209 |
// because the one byte can become a lead byte in multibyte |
| 4210 |
// decoders, but only after the decoder has been queried |
| 4211 |
// for max length, so the decoder's own logic for adding |
| 4212 |
// one for a pending lead cannot work. |
| 4213 |
if let Some(sum) = byte_length.checked_add(2) { |
| 4214 |
if let Some(utf16_bom) = checked_add(1, checked_div(sum.checked_add(1), 2)) { |
| 4215 |
let encoding = self.encoding(); |
| 4216 |
if encoding == UTF_16LE || encoding == UTF_16BE { |
| 4217 |
// No need to consider the internal state of the underlying decoder, |
| 4218 |
// because it is at start, because no data has reached it yet. |
| 4219 |
return Some(utf16_bom); |
| 4220 |
} else if let Some(non_bom) = self.variant.max_utf16_buffer_length(sum) { |
| 4221 |
return Some(core::cmp::max(utf16_bom, non_bom)); |
| 4222 |
} |
| 4223 |
} |
| 4224 |
} |
| 4225 |
} |
| 4226 |
DecoderLifeCycle::Finished => panic!("Must not use a decoder that has finished."), |
| 4227 |
} |
| 4228 |
None |
| 4229 |
} |
| 4230 |
|
| 4231 |
/// Incrementally decode a byte stream into UTF-16 with malformed sequences |
| 4232 |
/// replaced with the REPLACEMENT CHARACTER. |
| 4233 |
/// |
| 4234 |
/// See the documentation of the struct for documentation for `decode_*` |
| 4235 |
/// methods collectively. |
| 4236 |
/// |
| 4237 |
/// Available via the C wrapper. |
| 4238 |
pub fn decode_to_utf16( |
| 4239 |
&mut self, |
| 4240 |
src: &[u8], |
| 4241 |
dst: &mut [u16], |
| 4242 |
last: bool, |
| 4243 |
) -> (CoderResult, usize, usize, bool) { |
| 4244 |
let mut had_errors = false; |
| 4245 |
let mut total_read = 0usize; |
| 4246 |
let mut total_written = 0usize; |
| 4247 |
loop { |
| 4248 |
let (result, read, written) = self.decode_to_utf16_without_replacement( |
| 4249 |
&src[total_read..], |
| 4250 |
&mut dst[total_written..], |
| 4251 |
last, |
| 4252 |
); |
| 4253 |
total_read += read; |
| 4254 |
total_written += written; |
| 4255 |
match result { |
| 4256 |
DecoderResult::InputEmpty => { |
| 4257 |
return ( |
| 4258 |
CoderResult::InputEmpty, |
| 4259 |
total_read, |
| 4260 |
total_written, |
| 4261 |
had_errors, |
| 4262 |
); |
| 4263 |
} |
| 4264 |
DecoderResult::OutputFull => { |
| 4265 |
return ( |
| 4266 |
CoderResult::OutputFull, |
| 4267 |
total_read, |
| 4268 |
total_written, |
| 4269 |
had_errors, |
| 4270 |
); |
| 4271 |
} |
| 4272 |
DecoderResult::Malformed(_, _) => { |
| 4273 |
had_errors = true; |
| 4274 |
// There should always be space for the U+FFFD, because |
| 4275 |
// otherwise we'd have gotten OutputFull already. |
| 4276 |
dst[total_written] = 0xFFFD; |
| 4277 |
total_written += 1; |
| 4278 |
} |
| 4279 |
} |
| 4280 |
} |
| 4281 |
} |
| 4282 |
|
| 4283 |
public_decode_function!(/// Incrementally decode a byte stream into UTF-16 |
| 4284 |
/// _without replacement_. |
| 4285 |
/// |
| 4286 |
/// See the documentation of the struct for |
| 4287 |
/// documentation for `decode_*` methods |
| 4288 |
/// collectively. |
| 4289 |
/// |
| 4290 |
/// Available via the C wrapper. |
| 4291 |
, |
| 4292 |
decode_to_utf16_without_replacement, |
| 4293 |
decode_to_utf16_raw, |
| 4294 |
decode_to_utf16_checking_end, |
| 4295 |
decode_to_utf16_after_one_potential_bom_byte, |
| 4296 |
decode_to_utf16_after_two_potential_bom_bytes, |
| 4297 |
decode_to_utf16_checking_end_with_offset, |
| 4298 |
u16); |
| 4299 |
|
| 4300 |
/// Checks for compatibility with storing Unicode scalar values as unsigned |
| 4301 |
/// bytes taking into account the state of the decoder. |
| 4302 |
/// |
| 4303 |
/// Returns `None` if the decoder is not in a neutral state, including waiting |
| 4304 |
/// for the BOM, or if the encoding is never Latin1-byte-compatible. |
| 4305 |
/// |
| 4306 |
/// Otherwise returns the index of the first byte whose unsigned value doesn't |
| 4307 |
/// directly correspond to the decoded Unicode scalar value, or the length |
| 4308 |
/// of the input if all bytes in the input decode directly to scalar values |
| 4309 |
/// corresponding to the unsigned byte values. |
| 4310 |
/// |
| 4311 |
/// Does not change the state of the decoder. |
| 4312 |
/// |
| 4313 |
/// Do not use this unless you are supporting SpiderMonkey/V8-style string |
| 4314 |
/// storage optimizations. |
| 4315 |
/// |
| 4316 |
/// Available via the C wrapper. |
| 4317 |
pub fn latin1_byte_compatible_up_to(&self, bytes: &[u8]) -> Option<usize> { |
| 4318 |
match self.life_cycle { |
| 4319 |
DecoderLifeCycle::Converting => { |
| 4320 |
return self.variant.latin1_byte_compatible_up_to(bytes); |
| 4321 |
} |
| 4322 |
DecoderLifeCycle::Finished => panic!("Must not use a decoder that has finished."), |
| 4323 |
_ => None, |
| 4324 |
} |
| 4325 |
} |
| 4326 |
} |
| 4327 |
|
| 4328 |
/// Result of a (potentially partial) encode operation without replacement. |
| 4329 |
#[must_use] |
| 4330 |
#[derive(Debug, PartialEq, Eq)] |
| 4331 |
pub enum EncoderResult { |
| 4332 |
/// The input was exhausted. |
| 4333 |
/// |
| 4334 |
/// If this result was returned from a call where `last` was `true`, the |
| 4335 |
/// decoding process has completed. Otherwise, the caller should call a |
| 4336 |
/// decode method again with more input. |
| 4337 |
InputEmpty, |
| 4338 |
|
| 4339 |
/// The encoder cannot produce another unit of output, because the output |
| 4340 |
/// buffer does not have enough space left. |
| 4341 |
/// |
| 4342 |
/// The caller must provide more output space upon the next call and re-push |
| 4343 |
/// the remaining input to the decoder. |
| 4344 |
OutputFull, |
| 4345 |
|
| 4346 |
/// The encoder encountered an unmappable character. |
| 4347 |
/// |
| 4348 |
/// The caller must either treat this as a fatal error or must append |
| 4349 |
/// a placeholder to the output and then re-push the remaining input to the |
| 4350 |
/// encoder. |
| 4351 |
Unmappable(char), |
| 4352 |
} |
| 4353 |
|
| 4354 |
impl EncoderResult { |
| 4355 |
fn unmappable_from_bmp(bmp: u16) -> EncoderResult { |
| 4356 |
EncoderResult::Unmappable(::core::char::from_u32(u32::from(bmp)).unwrap()) |
| 4357 |
} |
| 4358 |
} |
| 4359 |
|
| 4360 |
/// A converter that encodes a Unicode stream into bytes according to a |
| 4361 |
/// character encoding in a streaming (incremental) manner. |
| 4362 |
/// |
| 4363 |
/// The various `encode_*` methods take an input buffer (`src`) and an output |
| 4364 |
/// buffer `dst` both of which are caller-allocated. There are variants for |
| 4365 |
/// both UTF-8 and UTF-16 input buffers. |
| 4366 |
/// |
| 4367 |
/// An `encode_*` method encode characters from `src` into bytes characters |
| 4368 |
/// stored into `dst` until one of the following three things happens: |
| 4369 |
/// |
| 4370 |
/// 1. An unmappable character is encountered (`*_without_replacement` variants |
| 4371 |
/// only). |
| 4372 |
/// |
| 4373 |
/// 2. The output buffer has been filled so near capacity that the decoder |
| 4374 |
/// cannot be sure that processing an additional character of input wouldn't |
| 4375 |
/// cause so much output that the output buffer would overflow. |
| 4376 |
/// |
| 4377 |
/// 3. All the input characters have been processed. |
| 4378 |
/// |
| 4379 |
/// The `encode_*` method then returns tuple of a status indicating which one |
| 4380 |
/// of the three reasons to return happened, how many input code units (`u8` |
| 4381 |
/// when encoding from UTF-8 and `u16` when encoding from UTF-16) were read, |
| 4382 |
/// how many output bytes were written (except when encoding into `Vec<u8>`, |
| 4383 |
/// whose length change indicates this), and in the case of the variants that |
| 4384 |
/// perform replacement, a boolean indicating whether an unmappable |
| 4385 |
/// character was replaced with a numeric character reference during the call. |
| 4386 |
/// |
| 4387 |
/// The number of bytes "written" is what's logically written. Garbage may be |
| 4388 |
/// written in the output buffer beyond the point logically written to. |
| 4389 |
/// |
| 4390 |
/// In the case of the methods whose name ends with |
| 4391 |
/// `*_without_replacement`, the status is an [`EncoderResult`][1] enumeration |
| 4392 |
/// (possibilities `Unmappable`, `OutputFull` and `InputEmpty` corresponding to |
| 4393 |
/// the three cases listed above). |
| 4394 |
/// |
| 4395 |
/// In the case of methods whose name does not end with |
| 4396 |
/// `*_without_replacement`, unmappable characters are automatically replaced |
| 4397 |
/// with the corresponding numeric character references and unmappable |
| 4398 |
/// characters do not cause the methods to return early. |
| 4399 |
/// |
| 4400 |
/// When encoding from UTF-8 without replacement, the methods are guaranteed |
| 4401 |
/// not to return indicating that more output space is needed if the length |
| 4402 |
/// of the output buffer is at least the length returned by |
| 4403 |
/// [`max_buffer_length_from_utf8_without_replacement()`][2]. When encoding from |
| 4404 |
/// UTF-8 with replacement, the length of the output buffer that guarantees the |
| 4405 |
/// methods not to return indicating that more output space is needed in the |
| 4406 |
/// absence of unmappable characters is given by |
| 4407 |
/// [`max_buffer_length_from_utf8_if_no_unmappables()`][3]. When encoding from |
| 4408 |
/// UTF-16 without replacement, the methods are guaranteed not to return |
| 4409 |
/// indicating that more output space is needed if the length of the output |
| 4410 |
/// buffer is at least the length returned by |
| 4411 |
/// [`max_buffer_length_from_utf16_without_replacement()`][4]. When encoding |
| 4412 |
/// from UTF-16 with replacement, the the length of the output buffer that |
| 4413 |
/// guarantees the methods not to return indicating that more output space is |
| 4414 |
/// needed in the absence of unmappable characters is given by |
| 4415 |
/// [`max_buffer_length_from_utf16_if_no_unmappables()`][5]. |
| 4416 |
/// When encoding with replacement, applications are not expected to size the |
| 4417 |
/// buffer for the worst case ahead of time but to resize the buffer if there |
| 4418 |
/// are unmappable characters. This is why max length queries are only available |
| 4419 |
/// for the case where there are no unmappable characters. |
| 4420 |
/// |
| 4421 |
/// When encoding from UTF-8, each `src` buffer _must_ be valid UTF-8. (When |
| 4422 |
/// calling from Rust, the type system takes care of this.) When encoding from |
| 4423 |
/// UTF-16, unpaired surrogates in the input are treated as U+FFFD REPLACEMENT |
| 4424 |
/// CHARACTERS. Therefore, in order for astral characters not to turn into a |
| 4425 |
/// pair of REPLACEMENT CHARACTERS, the caller must ensure that surrogate pairs |
| 4426 |
/// are not split across input buffer boundaries. |
| 4427 |
/// |
| 4428 |
/// After an `encode_*` call returns, the output produced so far, taken as a |
| 4429 |
/// whole from the start of the stream, is guaranteed to consist of a valid |
| 4430 |
/// byte sequence in the target encoding. (I.e. the code unit sequence for a |
| 4431 |
/// character is guaranteed not to be split across output buffers. However, due |
| 4432 |
/// to the stateful nature of ISO-2022-JP, the stream needs to be considered |
| 4433 |
/// from the start for it to be valid. For other encodings, the validity holds |
| 4434 |
/// on a per-output buffer basis.) |
| 4435 |
/// |
| 4436 |
/// The boolean argument `last` indicates that the end of the stream is reached |
| 4437 |
/// when all the characters in `src` have been consumed. This argument is needed |
| 4438 |
/// for ISO-2022-JP and is ignored for other encodings. |
| 4439 |
/// |
| 4440 |
/// An `Encoder` object can be used to incrementally encode a byte stream. |
| 4441 |
/// |
| 4442 |
/// During the processing of a single stream, the caller must call `encode_*` |
| 4443 |
/// zero or more times with `last` set to `false` and then call `encode_*` at |
| 4444 |
/// least once with `last` set to `true`. If `encode_*` returns `InputEmpty`, |
| 4445 |
/// the processing of the stream has ended. Otherwise, the caller must call |
| 4446 |
/// `encode_*` again with `last` set to `true` (or treat an `Unmappable` result |
| 4447 |
/// as a fatal error). |
| 4448 |
/// |
| 4449 |
/// Once the stream has ended, the `Encoder` object must not be used anymore. |
| 4450 |
/// That is, you need to create another one to process another stream. |
| 4451 |
/// |
| 4452 |
/// When the encoder returns `OutputFull` or the encoder returns `Unmappable` |
| 4453 |
/// and the caller does not wish to treat it as a fatal error, the input buffer |
| 4454 |
/// `src` may not have been completely consumed. In that case, the caller must |
| 4455 |
/// pass the unconsumed contents of `src` to `encode_*` again upon the next |
| 4456 |
/// call. |
| 4457 |
/// |
| 4458 |
/// [1]: enum.EncoderResult.html |
| 4459 |
/// [2]: #method.max_buffer_length_from_utf8_without_replacement |
| 4460 |
/// [3]: #method.max_buffer_length_from_utf8_if_no_unmappables |
| 4461 |
/// [4]: #method.max_buffer_length_from_utf16_without_replacement |
| 4462 |
/// [5]: #method.max_buffer_length_from_utf16_if_no_unmappables |
| 4463 |
/// |
| 4464 |
/// # Infinite loops |
| 4465 |
/// |
| 4466 |
/// When converting with a fixed-size output buffer whose size is too small to |
| 4467 |
/// accommodate one character of output, an infinite loop ensues. When |
| 4468 |
/// converting with a fixed-size output buffer, it generally makes sense to |
| 4469 |
/// make the buffer fairly large (e.g. couple of kilobytes). |
| 4470 |
pub struct Encoder { |
| 4471 |
encoding: &'static Encoding, |
| 4472 |
variant: VariantEncoder, |
| 4473 |
} |
| 4474 |
|
| 4475 |
impl Encoder { |
| 4476 |
fn new(enc: &'static Encoding, encoder: VariantEncoder) -> Encoder { |
| 4477 |
Encoder { |
| 4478 |
encoding: enc, |
| 4479 |
variant: encoder, |
| 4480 |
} |
| 4481 |
} |
| 4482 |
|
| 4483 |
/// The `Encoding` this `Encoder` is for. |
| 4484 |
#[inline] |
| 4485 |
pub fn encoding(&self) -> &'static Encoding { |
| 4486 |
self.encoding |
| 4487 |
} |
| 4488 |
|
| 4489 |
/// Returns `true` if this is an ISO-2022-JP encoder that's not in the |
| 4490 |
/// ASCII state and `false` otherwise. |
| 4491 |
#[inline] |
| 4492 |
pub fn has_pending_state(&self) -> bool { |
| 4493 |
self.variant.has_pending_state() |
| 4494 |
} |
| 4495 |
|
| 4496 |
/// Query the worst-case output size when encoding from UTF-8 with |
| 4497 |
/// replacement. |
| 4498 |
/// |
| 4499 |
/// Returns the size of the output buffer in bytes that will not overflow |
| 4500 |
/// given the current state of the encoder and `byte_length` number of |
| 4501 |
/// additional input code units if there are no unmappable characters in |
| 4502 |
/// the input or `None` if `usize` would overflow. |
| 4503 |
/// |
| 4504 |
/// Available via the C wrapper. |
| 4505 |
pub fn max_buffer_length_from_utf8_if_no_unmappables( |
| 4506 |
&self, |
| 4507 |
byte_length: usize, |
| 4508 |
) -> Option<usize> { |
| 4509 |
checked_add( |
| 4510 |
if self.encoding().can_encode_everything() { |
| 4511 |
0 |
| 4512 |
} else { |
| 4513 |
NCR_EXTRA |
| 4514 |
}, |
| 4515 |
self.max_buffer_length_from_utf8_without_replacement(byte_length), |
| 4516 |
) |
| 4517 |
} |
| 4518 |
|
| 4519 |
/// Query the worst-case output size when encoding from UTF-8 without |
| 4520 |
/// replacement. |
| 4521 |
/// |
| 4522 |
/// Returns the size of the output buffer in bytes that will not overflow |
| 4523 |
/// given the current state of the encoder and `byte_length` number of |
| 4524 |
/// additional input code units or `None` if `usize` would overflow. |
| 4525 |
/// |
| 4526 |
/// Available via the C wrapper. |
| 4527 |
pub fn max_buffer_length_from_utf8_without_replacement( |
| 4528 |
&self, |
| 4529 |
byte_length: usize, |
| 4530 |
) -> Option<usize> { |
| 4531 |
self.variant |
| 4532 |
.max_buffer_length_from_utf8_without_replacement(byte_length) |
| 4533 |
} |
| 4534 |
|
| 4535 |
/// Incrementally encode into byte stream from UTF-8 with unmappable |
| 4536 |
/// characters replaced with HTML (decimal) numeric character references. |
| 4537 |
/// |
| 4538 |
/// See the documentation of the struct for documentation for `encode_*` |
| 4539 |
/// methods collectively. |
| 4540 |
/// |
| 4541 |
/// Available via the C wrapper. |
| 4542 |
pub fn encode_from_utf8( |
| 4543 |
&mut self, |
| 4544 |
src: &str, |
| 4545 |
dst: &mut [u8], |
| 4546 |
last: bool, |
| 4547 |
) -> (CoderResult, usize, usize, bool) { |
| 4548 |
let dst_len = dst.len(); |
| 4549 |
let effective_dst_len = if self.encoding().can_encode_everything() { |
| 4550 |
dst_len |
| 4551 |
} else { |
| 4552 |
if dst_len < NCR_EXTRA { |
| 4553 |
if src.is_empty() && !(last && self.has_pending_state()) { |
| 4554 |
return (CoderResult::InputEmpty, 0, 0, false); |
| 4555 |
} |
| 4556 |
return (CoderResult::OutputFull, 0, 0, false); |
| 4557 |
} |
| 4558 |
dst_len - NCR_EXTRA |
| 4559 |
}; |
| 4560 |
let mut had_unmappables = false; |
| 4561 |
let mut total_read = 0usize; |
| 4562 |
let mut total_written = 0usize; |
| 4563 |
loop { |
| 4564 |
let (result, read, written) = self.encode_from_utf8_without_replacement( |
| 4565 |
&src[total_read..], |
| 4566 |
&mut dst[total_written..effective_dst_len], |
| 4567 |
last, |
| 4568 |
); |
| 4569 |
total_read += read; |
| 4570 |
total_written += written; |
| 4571 |
match result { |
| 4572 |
EncoderResult::InputEmpty => { |
| 4573 |
return ( |
| 4574 |
CoderResult::InputEmpty, |
| 4575 |
total_read, |
| 4576 |
total_written, |
| 4577 |
had_unmappables, |
| 4578 |
); |
| 4579 |
} |
| 4580 |
EncoderResult::OutputFull => { |
| 4581 |
return ( |
| 4582 |
CoderResult::OutputFull, |
| 4583 |
total_read, |
| 4584 |
total_written, |
| 4585 |
had_unmappables, |
| 4586 |
); |
| 4587 |
} |
| 4588 |
EncoderResult::Unmappable(unmappable) => { |
| 4589 |
had_unmappables = true; |
| 4590 |
debug_assert!(dst.len() - total_written >= NCR_EXTRA); |
| 4591 |
debug_assert_ne!(self.encoding(), UTF_16BE); |
| 4592 |
debug_assert_ne!(self.encoding(), UTF_16LE); |
| 4593 |
// Additionally, Iso2022JpEncoder is responsible for |
| 4594 |
// transitioning to ASCII when returning with Unmappable. |
| 4595 |
total_written += write_ncr(unmappable, &mut dst[total_written..]); |
| 4596 |
if total_written >= effective_dst_len { |
| 4597 |
if total_read == src.len() && !(last && self.has_pending_state()) { |
| 4598 |
return ( |
| 4599 |
CoderResult::InputEmpty, |
| 4600 |
total_read, |
| 4601 |
total_written, |
| 4602 |
had_unmappables, |
| 4603 |
); |
| 4604 |
} |
| 4605 |
return ( |
| 4606 |
CoderResult::OutputFull, |
| 4607 |
total_read, |
| 4608 |
total_written, |
| 4609 |
had_unmappables, |
| 4610 |
); |
| 4611 |
} |
| 4612 |
} |
| 4613 |
} |
| 4614 |
} |
| 4615 |
} |
| 4616 |
|
| 4617 |
/// Incrementally encode into byte stream from UTF-8 with unmappable |
| 4618 |
/// characters replaced with HTML (decimal) numeric character references. |
| 4619 |
/// |
| 4620 |
/// See the documentation of the struct for documentation for `encode_*` |
| 4621 |
/// methods collectively. |
| 4622 |
/// |
| 4623 |
/// Available to Rust only and only with the `alloc` feature enabled (enabled |
| 4624 |
/// by default). |
| 4625 |
#[cfg(feature = "alloc")] |
| 4626 |
pub fn encode_from_utf8_to_vec( |
| 4627 |
&mut self, |
| 4628 |
src: &str, |
| 4629 |
dst: &mut Vec<u8>, |
| 4630 |
last: bool, |
| 4631 |
) -> (CoderResult, usize, bool) { |
| 4632 |
unsafe { |
| 4633 |
let old_len = dst.len(); |
| 4634 |
let capacity = dst.capacity(); |
| 4635 |
dst.set_len(capacity); |
| 4636 |
let (result, read, written, replaced) = |
| 4637 |
self.encode_from_utf8(src, &mut dst[old_len..], last); |
| 4638 |
dst.set_len(old_len + written); |
| 4639 |
(result, read, replaced) |
| 4640 |
} |
| 4641 |
} |
| 4642 |
|
| 4643 |
/// Incrementally encode into byte stream from UTF-8 _without replacement_. |
| 4644 |
/// |
| 4645 |
/// See the documentation of the struct for documentation for `encode_*` |
| 4646 |
/// methods collectively. |
| 4647 |
/// |
| 4648 |
/// Available via the C wrapper. |
| 4649 |
pub fn encode_from_utf8_without_replacement( |
| 4650 |
&mut self, |
| 4651 |
src: &str, |
| 4652 |
dst: &mut [u8], |
| 4653 |
last: bool, |
| 4654 |
) -> (EncoderResult, usize, usize) { |
| 4655 |
self.variant.encode_from_utf8_raw(src, dst, last) |
| 4656 |
} |
| 4657 |
|
| 4658 |
/// Incrementally encode into byte stream from UTF-8 _without replacement_. |
| 4659 |
/// |
| 4660 |
/// See the documentation of the struct for documentation for `encode_*` |
| 4661 |
/// methods collectively. |
| 4662 |
/// |
| 4663 |
/// Available to Rust only and only with the `alloc` feature enabled (enabled |
| 4664 |
/// by default). |
| 4665 |
#[cfg(feature = "alloc")] |
| 4666 |
pub fn encode_from_utf8_to_vec_without_replacement( |
| 4667 |
&mut self, |
| 4668 |
src: &str, |
| 4669 |
dst: &mut Vec<u8>, |
| 4670 |
last: bool, |
| 4671 |
) -> (EncoderResult, usize) { |
| 4672 |
unsafe { |
| 4673 |
let old_len = dst.len(); |
| 4674 |
let capacity = dst.capacity(); |
| 4675 |
dst.set_len(capacity); |
| 4676 |
let (result, read, written) = |
| 4677 |
self.encode_from_utf8_without_replacement(src, &mut dst[old_len..], last); |
| 4678 |
dst.set_len(old_len + written); |
| 4679 |
(result, read) |
| 4680 |
} |
| 4681 |
} |
| 4682 |
|
| 4683 |
/// Query the worst-case output size when encoding from UTF-16 with |
| 4684 |
/// replacement. |
| 4685 |
/// |
| 4686 |
/// Returns the size of the output buffer in bytes that will not overflow |
| 4687 |
/// given the current state of the encoder and `u16_length` number of |
| 4688 |
/// additional input code units if there are no unmappable characters in |
| 4689 |
/// the input or `None` if `usize` would overflow. |
| 4690 |
/// |
| 4691 |
/// Available via the C wrapper. |
| 4692 |
pub fn max_buffer_length_from_utf16_if_no_unmappables( |
| 4693 |
&self, |
| 4694 |
u16_length: usize, |
| 4695 |
) -> Option<usize> { |
| 4696 |
checked_add( |
| 4697 |
if self.encoding().can_encode_everything() { |
| 4698 |
0 |
| 4699 |
} else { |
| 4700 |
NCR_EXTRA |
| 4701 |
}, |
| 4702 |
self.max_buffer_length_from_utf16_without_replacement(u16_length), |
| 4703 |
) |
| 4704 |
} |
| 4705 |
|
| 4706 |
/// Query the worst-case output size when encoding from UTF-16 without |
| 4707 |
/// replacement. |
| 4708 |
/// |
| 4709 |
/// Returns the size of the output buffer in bytes that will not overflow |
| 4710 |
/// given the current state of the encoder and `u16_length` number of |
| 4711 |
/// additional input code units or `None` if `usize` would overflow. |
| 4712 |
/// |
| 4713 |
/// Available via the C wrapper. |
| 4714 |
pub fn max_buffer_length_from_utf16_without_replacement( |
| 4715 |
&self, |
| 4716 |
u16_length: usize, |
| 4717 |
) -> Option<usize> { |
| 4718 |
self.variant |
| 4719 |
.max_buffer_length_from_utf16_without_replacement(u16_length) |
| 4720 |
} |
| 4721 |
|
| 4722 |
/// Incrementally encode into byte stream from UTF-16 with unmappable |
| 4723 |
/// characters replaced with HTML (decimal) numeric character references. |
| 4724 |
/// |
| 4725 |
/// See the documentation of the struct for documentation for `encode_*` |
| 4726 |
/// methods collectively. |
| 4727 |
/// |
| 4728 |
/// Available via the C wrapper. |
| 4729 |
pub fn encode_from_utf16( |
| 4730 |
&mut self, |
| 4731 |
src: &[u16], |
| 4732 |
dst: &mut [u8], |
| 4733 |
last: bool, |
| 4734 |
) -> (CoderResult, usize, usize, bool) { |
| 4735 |
let dst_len = dst.len(); |
| 4736 |
let effective_dst_len = if self.encoding().can_encode_everything() { |
| 4737 |
dst_len |
| 4738 |
} else { |
| 4739 |
if dst_len < NCR_EXTRA { |
| 4740 |
if src.is_empty() && !(last && self.has_pending_state()) { |
| 4741 |
return (CoderResult::InputEmpty, 0, 0, false); |
| 4742 |
} |
| 4743 |
return (CoderResult::OutputFull, 0, 0, false); |
| 4744 |
} |
| 4745 |
dst_len - NCR_EXTRA |
| 4746 |
}; |
| 4747 |
let mut had_unmappables = false; |
| 4748 |
let mut total_read = 0usize; |
| 4749 |
let mut total_written = 0usize; |
| 4750 |
loop { |
| 4751 |
let (result, read, written) = self.encode_from_utf16_without_replacement( |
| 4752 |
&src[total_read..], |
| 4753 |
&mut dst[total_written..effective_dst_len], |
| 4754 |
last, |
| 4755 |
); |
| 4756 |
total_read += read; |
| 4757 |
total_written += written; |
| 4758 |
match result { |
| 4759 |
EncoderResult::InputEmpty => { |
| 4760 |
return ( |
| 4761 |
CoderResult::InputEmpty, |
| 4762 |
total_read, |
| 4763 |
total_written, |
| 4764 |
had_unmappables, |
| 4765 |
); |
| 4766 |
} |
| 4767 |
EncoderResult::OutputFull => { |
| 4768 |
return ( |
| 4769 |
CoderResult::OutputFull, |
| 4770 |
total_read, |
| 4771 |
total_written, |
| 4772 |
had_unmappables, |
| 4773 |
); |
| 4774 |
} |
| 4775 |
EncoderResult::Unmappable(unmappable) => { |
| 4776 |
had_unmappables = true; |
| 4777 |
debug_assert!(dst.len() - total_written >= NCR_EXTRA); |
| 4778 |
// There are no UTF-16 encoders and even if there were, |
| 4779 |
// they'd never have unmappables. |
| 4780 |
debug_assert_ne!(self.encoding(), UTF_16BE); |
| 4781 |
debug_assert_ne!(self.encoding(), UTF_16LE); |
| 4782 |
// Additionally, Iso2022JpEncoder is responsible for |
| 4783 |
// transitioning to ASCII when returning with Unmappable |
| 4784 |
// from the jis0208 state. That is, when we encode |
| 4785 |
// ISO-2022-JP and come here, the encoder is in either the |
| 4786 |
// ASCII or the Roman state. We are allowed to generate any |
| 4787 |
// printable ASCII excluding \ and ~. |
| 4788 |
total_written += write_ncr(unmappable, &mut dst[total_written..]); |
| 4789 |
if total_written >= effective_dst_len { |
| 4790 |
if total_read == src.len() && !(last && self.has_pending_state()) { |
| 4791 |
return ( |
| 4792 |
CoderResult::InputEmpty, |
| 4793 |
total_read, |
| 4794 |
total_written, |
| 4795 |
had_unmappables, |
| 4796 |
); |
| 4797 |
} |
| 4798 |
return ( |
| 4799 |
CoderResult::OutputFull, |
| 4800 |
total_read, |
| 4801 |
total_written, |
| 4802 |
had_unmappables, |
| 4803 |
); |
| 4804 |
} |
| 4805 |
} |
| 4806 |
} |
| 4807 |
} |
| 4808 |
} |
| 4809 |
|
| 4810 |
/// Incrementally encode into byte stream from UTF-16 _without replacement_. |
| 4811 |
/// |
| 4812 |
/// See the documentation of the struct for documentation for `encode_*` |
| 4813 |
/// methods collectively. |
| 4814 |
/// |
| 4815 |
/// Available via the C wrapper. |
| 4816 |
pub fn encode_from_utf16_without_replacement( |
| 4817 |
&mut self, |
| 4818 |
src: &[u16], |
| 4819 |
dst: &mut [u8], |
| 4820 |
last: bool, |
| 4821 |
) -> (EncoderResult, usize, usize) { |
| 4822 |
self.variant.encode_from_utf16_raw(src, dst, last) |
| 4823 |
} |
| 4824 |
} |
| 4825 |
|
| 4826 |
/// Format an unmappable as NCR without heap allocation. |
| 4827 |
fn write_ncr(unmappable: char, dst: &mut [u8]) -> usize { |
| 4828 |
// len is the number of decimal digits needed to represent unmappable plus |
| 4829 |
// 3 (the length of "&#" and ";"). |
| 4830 |
let mut number = unmappable as u32; |
| 4831 |
let len = if number >= 1_000_000u32 { |
| 4832 |
10usize |
| 4833 |
} else if number >= 100_000u32 { |
| 4834 |
9usize |
| 4835 |
} else if number >= 10_000u32 { |
| 4836 |
8usize |
| 4837 |
} else if number >= 1_000u32 { |
| 4838 |
7usize |
| 4839 |
} else if number >= 100u32 { |
| 4840 |
6usize |
| 4841 |
} else { |
| 4842 |
// Review the outcome of https://github.com/whatwg/encoding/issues/15 |
| 4843 |
// to see if this case is possible |
| 4844 |
5usize |
| 4845 |
}; |
| 4846 |
debug_assert!(number >= 10u32); |
| 4847 |
debug_assert!(len <= dst.len()); |
| 4848 |
let mut pos = len - 1; |
| 4849 |
dst[pos] = b';'; |
| 4850 |
pos -= 1; |
| 4851 |
loop { |
| 4852 |
let rightmost = number % 10; |
| 4853 |
dst[pos] = rightmost as u8 + b'0'; |
| 4854 |
pos -= 1; |
| 4855 |
if number < 10 { |
| 4856 |
break; |
| 4857 |
} |
| 4858 |
number /= 10; |
| 4859 |
} |
| 4860 |
dst[1] = b'#'; |
| 4861 |
dst[0] = b'&'; |
| 4862 |
len |
| 4863 |
} |
| 4864 |
|
| 4865 |
#[inline(always)] |
| 4866 |
fn in_range16(i: u16, start: u16, end: u16) -> bool { |
| 4867 |
i.wrapping_sub(start) < (end - start) |
| 4868 |
} |
| 4869 |
|
| 4870 |
#[inline(always)] |
| 4871 |
fn in_range32(i: u32, start: u32, end: u32) -> bool { |
| 4872 |
i.wrapping_sub(start) < (end - start) |
| 4873 |
} |
| 4874 |
|
| 4875 |
#[inline(always)] |
| 4876 |
fn in_inclusive_range8(i: u8, start: u8, end: u8) -> bool { |
| 4877 |
i.wrapping_sub(start) <= (end - start) |
| 4878 |
} |
| 4879 |
|
| 4880 |
#[inline(always)] |
| 4881 |
fn in_inclusive_range16(i: u16, start: u16, end: u16) -> bool { |
| 4882 |
i.wrapping_sub(start) <= (end - start) |
| 4883 |
} |
| 4884 |
|
| 4885 |
#[inline(always)] |
| 4886 |
fn in_inclusive_range32(i: u32, start: u32, end: u32) -> bool { |
| 4887 |
i.wrapping_sub(start) <= (end - start) |
| 4888 |
} |
| 4889 |
|
| 4890 |
#[inline(always)] |
| 4891 |
fn in_inclusive_range(i: usize, start: usize, end: usize) -> bool { |
| 4892 |
i.wrapping_sub(start) <= (end - start) |
| 4893 |
} |
| 4894 |
|
| 4895 |
#[inline(always)] |
| 4896 |
fn checked_add(num: usize, opt: Option<usize>) -> Option<usize> { |
| 4897 |
if let Some(n) = opt { |
| 4898 |
n.checked_add(num) |
| 4899 |
} else { |
| 4900 |
None |
| 4901 |
} |
| 4902 |
} |
| 4903 |
|
| 4904 |
#[inline(always)] |
| 4905 |
fn checked_add_opt(one: Option<usize>, other: Option<usize>) -> Option<usize> { |
| 4906 |
if let Some(n) = one { |
| 4907 |
checked_add(n, other) |
| 4908 |
} else { |
| 4909 |
None |
| 4910 |
} |
| 4911 |
} |
| 4912 |
|
| 4913 |
#[inline(always)] |
| 4914 |
fn checked_mul(num: usize, opt: Option<usize>) -> Option<usize> { |
| 4915 |
if let Some(n) = opt { |
| 4916 |
n.checked_mul(num) |
| 4917 |
} else { |
| 4918 |
None |
| 4919 |
} |
| 4920 |
} |
| 4921 |
|
| 4922 |
#[inline(always)] |
| 4923 |
fn checked_div(opt: Option<usize>, num: usize) -> Option<usize> { |
| 4924 |
if let Some(n) = opt { |
| 4925 |
n.checked_div(num) |
| 4926 |
} else { |
| 4927 |
None |
| 4928 |
} |
| 4929 |
} |
| 4930 |
|
| 4931 |
#[cfg(feature = "alloc")] |
| 4932 |
#[inline(always)] |
| 4933 |
fn checked_next_power_of_two(opt: Option<usize>) -> Option<usize> { |
| 4934 |
opt.map(|n| n.next_power_of_two()) |
| 4935 |
} |
| 4936 |
|
| 4937 |
#[cfg(feature = "alloc")] |
| 4938 |
#[inline(always)] |
| 4939 |
fn checked_min(one: Option<usize>, other: Option<usize>) -> Option<usize> { |
| 4940 |
if let Some(a) = one { |
| 4941 |
if let Some(b) = other { |
| 4942 |
Some(::core::cmp::min(a, b)) |
| 4943 |
} else { |
| 4944 |
Some(a) |
| 4945 |
} |
| 4946 |
} else { |
| 4947 |
other |
| 4948 |
} |
| 4949 |
} |
| 4950 |
|
| 4951 |
// ############## TESTS ############### |
| 4952 |
|
| 4953 |
#[cfg(all(test, feature = "serde"))] |
| 4954 |
#[derive(Serialize, Deserialize, Debug, PartialEq)] |
| 4955 |
struct Demo { |
| 4956 |
num: u32, |
| 4957 |
name: String, |
| 4958 |
enc: &'static Encoding, |
| 4959 |
} |
| 4960 |
|
| 4961 |
#[cfg(test)] |
| 4962 |
mod test_labels_names; |
| 4963 |
|
| 4964 |
#[cfg(all(test, feature = "alloc"))] |
| 4965 |
mod tests { |
| 4966 |
use super::*; |
| 4967 |
use alloc::borrow::Cow; |
| 4968 |
|
| 4969 |
fn sniff_to_utf16( |
| 4970 |
initial_encoding: &'static Encoding, |
| 4971 |
expected_encoding: &'static Encoding, |
| 4972 |
bytes: &[u8], |
| 4973 |
expect: &[u16], |
| 4974 |
breaks: &[usize], |
| 4975 |
) { |
| 4976 |
let mut decoder = initial_encoding.new_decoder(); |
| 4977 |
|
| 4978 |
let mut dest: Vec<u16> = |
| 4979 |
Vec::with_capacity(decoder.max_utf16_buffer_length(bytes.len()).unwrap()); |
| 4980 |
let capacity = dest.capacity(); |
| 4981 |
dest.resize(capacity, 0u16); |
| 4982 |
|
| 4983 |
let mut total_written = 0usize; |
| 4984 |
let mut start = 0usize; |
| 4985 |
for br in breaks { |
| 4986 |
let (result, read, written, _) = |
| 4987 |
decoder.decode_to_utf16(&bytes[start..*br], &mut dest[total_written..], false); |
| 4988 |
total_written += written; |
| 4989 |
assert_eq!(read, *br - start); |
| 4990 |
match result { |
| 4991 |
CoderResult::InputEmpty => {} |
| 4992 |
CoderResult::OutputFull => { |
| 4993 |
unreachable!(); |
| 4994 |
} |
| 4995 |
} |
| 4996 |
start = *br; |
| 4997 |
} |
| 4998 |
let (result, read, written, _) = |
| 4999 |
decoder.decode_to_utf16(&bytes[start..], &mut dest[total_written..], true); |
| 5000 |
total_written += written; |
| 5001 |
match result { |
| 5002 |
CoderResult::InputEmpty => {} |
| 5003 |
CoderResult::OutputFull => { |
| 5004 |
unreachable!(); |
| 5005 |
} |
| 5006 |
} |
| 5007 |
assert_eq!(read, bytes.len() - start); |
| 5008 |
assert_eq!(total_written, expect.len()); |
| 5009 |
assert_eq!(&dest[..total_written], expect); |
| 5010 |
assert_eq!(decoder.encoding(), expected_encoding); |
| 5011 |
} |
| 5012 |
|
| 5013 |
// Any copyright to the test code below this comment is dedicated to the |
| 5014 |
// Public Domain. http://creativecommons.org/publicdomain/zero/1.0/ |
| 5015 |
|
| 5016 |
#[test] |
| 5017 |
fn test_bom_sniffing() { |
| 5018 |
// ASCII |
| 5019 |
sniff_to_utf16( |
| 5020 |
WINDOWS_1252, |
| 5021 |
WINDOWS_1252, |
| 5022 |
b"\x61\x62", |
| 5023 |
&[0x0061u16, 0x0062u16], |
| 5024 |
&[], |
| 5025 |
); |
| 5026 |
// UTF-8 |
| 5027 |
sniff_to_utf16( |
| 5028 |
WINDOWS_1252, |
| 5029 |
UTF_8, |
| 5030 |
b"\xEF\xBB\xBF\x61\x62", |
| 5031 |
&[0x0061u16, 0x0062u16], |
| 5032 |
&[], |
| 5033 |
); |
| 5034 |
sniff_to_utf16( |
| 5035 |
WINDOWS_1252, |
| 5036 |
UTF_8, |
| 5037 |
b"\xEF\xBB\xBF\x61\x62", |
| 5038 |
&[0x0061u16, 0x0062u16], |
| 5039 |
&[1], |
| 5040 |
); |
| 5041 |
sniff_to_utf16( |
| 5042 |
WINDOWS_1252, |
| 5043 |
UTF_8, |
| 5044 |
b"\xEF\xBB\xBF\x61\x62", |
| 5045 |
&[0x0061u16, 0x0062u16], |
| 5046 |
&[2], |
| 5047 |
); |
| 5048 |
sniff_to_utf16( |
| 5049 |
WINDOWS_1252, |
| 5050 |
UTF_8, |
| 5051 |
b"\xEF\xBB\xBF\x61\x62", |
| 5052 |
&[0x0061u16, 0x0062u16], |
| 5053 |
&[3], |
| 5054 |
); |
| 5055 |
sniff_to_utf16( |
| 5056 |
WINDOWS_1252, |
| 5057 |
UTF_8, |
| 5058 |
b"\xEF\xBB\xBF\x61\x62", |
| 5059 |
&[0x0061u16, 0x0062u16], |
| 5060 |
&[4], |
| 5061 |
); |
| 5062 |
sniff_to_utf16( |
| 5063 |
WINDOWS_1252, |
| 5064 |
UTF_8, |
| 5065 |
b"\xEF\xBB\xBF\x61\x62", |
| 5066 |
&[0x0061u16, 0x0062u16], |
| 5067 |
&[2, 3], |
| 5068 |
); |
| 5069 |
sniff_to_utf16( |
| 5070 |
WINDOWS_1252, |
| 5071 |
UTF_8, |
| 5072 |
b"\xEF\xBB\xBF\x61\x62", |
| 5073 |
&[0x0061u16, 0x0062u16], |
| 5074 |
&[1, 2], |
| 5075 |
); |
| 5076 |
sniff_to_utf16( |
| 5077 |
WINDOWS_1252, |
| 5078 |
UTF_8, |
| 5079 |
b"\xEF\xBB\xBF\x61\x62", |
| 5080 |
&[0x0061u16, 0x0062u16], |
| 5081 |
&[1, 3], |
| 5082 |
); |
| 5083 |
sniff_to_utf16( |
| 5084 |
WINDOWS_1252, |
| 5085 |
UTF_8, |
| 5086 |
b"\xEF\xBB\xBF\x61\x62", |
| 5087 |
&[0x0061u16, 0x0062u16], |
| 5088 |
&[1, 2, 3, 4], |
| 5089 |
); |
| 5090 |
sniff_to_utf16(WINDOWS_1252, UTF_8, b"\xEF\xBB\xBF", &[], &[]); |
| 5091 |
// Not UTF-8 |
| 5092 |
sniff_to_utf16( |
| 5093 |
WINDOWS_1252, |
| 5094 |
WINDOWS_1252, |
| 5095 |
b"\xEF\xBB\x61\x62", |
| 5096 |
&[0x00EFu16, 0x00BBu16, 0x0061u16, 0x0062u16], |
| 5097 |
&[], |
| 5098 |
); |
| 5099 |
sniff_to_utf16( |
| 5100 |
WINDOWS_1252, |
| 5101 |
WINDOWS_1252, |
| 5102 |
b"\xEF\xBB\x61\x62", |
| 5103 |
&[0x00EFu16, 0x00BBu16, 0x0061u16, 0x0062u16], |
| 5104 |
&[1], |
| 5105 |
); |
| 5106 |
sniff_to_utf16( |
| 5107 |
WINDOWS_1252, |
| 5108 |
WINDOWS_1252, |
| 5109 |
b"\xEF\x61\x62", |
| 5110 |
&[0x00EFu16, 0x0061u16, 0x0062u16], |
| 5111 |
&[], |
| 5112 |
); |
| 5113 |
sniff_to_utf16( |
| 5114 |
WINDOWS_1252, |
| 5115 |
WINDOWS_1252, |
| 5116 |
b"\xEF\x61\x62", |
| 5117 |
&[0x00EFu16, 0x0061u16, 0x0062u16], |
| 5118 |
&[1], |
| 5119 |
); |
| 5120 |
sniff_to_utf16( |
| 5121 |
WINDOWS_1252, |
| 5122 |
WINDOWS_1252, |
| 5123 |
b"\xEF\xBB", |
| 5124 |
&[0x00EFu16, 0x00BBu16], |
| 5125 |
&[], |
| 5126 |
); |
| 5127 |
sniff_to_utf16( |
| 5128 |
WINDOWS_1252, |
| 5129 |
WINDOWS_1252, |
| 5130 |
b"\xEF\xBB", |
| 5131 |
&[0x00EFu16, 0x00BBu16], |
| 5132 |
&[1], |
| 5133 |
); |
| 5134 |
sniff_to_utf16(WINDOWS_1252, WINDOWS_1252, b"\xEF", &[0x00EFu16], &[]); |
| 5135 |
// Not UTF-16 |
| 5136 |
sniff_to_utf16( |
| 5137 |
WINDOWS_1252, |
| 5138 |
WINDOWS_1252, |
| 5139 |
b"\xFE\x61\x62", |
| 5140 |
&[0x00FEu16, 0x0061u16, 0x0062u16], |
| 5141 |
&[], |
| 5142 |
); |
| 5143 |
sniff_to_utf16( |
| 5144 |
WINDOWS_1252, |
| 5145 |
WINDOWS_1252, |
| 5146 |
b"\xFE\x61\x62", |
| 5147 |
&[0x00FEu16, 0x0061u16, 0x0062u16], |
| 5148 |
&[1], |
| 5149 |
); |
| 5150 |
sniff_to_utf16(WINDOWS_1252, WINDOWS_1252, b"\xFE", &[0x00FEu16], &[]); |
| 5151 |
sniff_to_utf16( |
| 5152 |
WINDOWS_1252, |
| 5153 |
WINDOWS_1252, |
| 5154 |
b"\xFF\x61\x62", |
| 5155 |
&[0x00FFu16, 0x0061u16, 0x0062u16], |
| 5156 |
&[], |
| 5157 |
); |
| 5158 |
sniff_to_utf16( |
| 5159 |
WINDOWS_1252, |
| 5160 |
WINDOWS_1252, |
| 5161 |
b"\xFF\x61\x62", |
| 5162 |
&[0x00FFu16, 0x0061u16, 0x0062u16], |
| 5163 |
&[1], |
| 5164 |
); |
| 5165 |
sniff_to_utf16(WINDOWS_1252, WINDOWS_1252, b"\xFF", &[0x00FFu16], &[]); |
| 5166 |
// UTF-16 |
| 5167 |
sniff_to_utf16(WINDOWS_1252, UTF_16BE, b"\xFE\xFF", &[], &[]); |
| 5168 |
sniff_to_utf16(WINDOWS_1252, UTF_16BE, b"\xFE\xFF", &[], &[1]); |
| 5169 |
sniff_to_utf16(WINDOWS_1252, UTF_16LE, b"\xFF\xFE", &[], &[]); |
| 5170 |
sniff_to_utf16(WINDOWS_1252, UTF_16LE, b"\xFF\xFE", &[], &[1]); |
| 5171 |
} |
| 5172 |
|
| 5173 |
#[test] |
| 5174 |
fn test_output_encoding() { |
| 5175 |
assert_eq!(REPLACEMENT.output_encoding(), UTF_8); |
| 5176 |
assert_eq!(UTF_16BE.output_encoding(), UTF_8); |
| 5177 |
assert_eq!(UTF_16LE.output_encoding(), UTF_8); |
| 5178 |
assert_eq!(UTF_8.output_encoding(), UTF_8); |
| 5179 |
assert_eq!(WINDOWS_1252.output_encoding(), WINDOWS_1252); |
| 5180 |
assert_eq!(REPLACEMENT.new_encoder().encoding(), UTF_8); |
| 5181 |
assert_eq!(UTF_16BE.new_encoder().encoding(), UTF_8); |
| 5182 |
assert_eq!(UTF_16LE.new_encoder().encoding(), UTF_8); |
| 5183 |
assert_eq!(UTF_8.new_encoder().encoding(), UTF_8); |
| 5184 |
assert_eq!(WINDOWS_1252.new_encoder().encoding(), WINDOWS_1252); |
| 5185 |
} |
| 5186 |
|
| 5187 |
#[test] |
| 5188 |
fn test_label_resolution() { |
| 5189 |
assert_eq!(Encoding::for_label(b"utf-8"), Some(UTF_8)); |
| 5190 |
assert_eq!(Encoding::for_label(b"UTF-8"), Some(UTF_8)); |
| 5191 |
assert_eq!( |
| 5192 |
Encoding::for_label(b" \t \n \x0C \n utf-8 \r \n \t \x0C "), |
| 5193 |
Some(UTF_8) |
| 5194 |
); |
| 5195 |
assert_eq!(Encoding::for_label(b"utf-8 _"), None); |
| 5196 |
assert_eq!(Encoding::for_label(b"bogus"), None); |
| 5197 |
assert_eq!(Encoding::for_label(b"bogusbogusbogusbogus"), None); |
| 5198 |
} |
| 5199 |
|
| 5200 |
#[test] |
| 5201 |
fn test_decode_valid_windows_1257_to_cow() { |
| 5202 |
let (cow, encoding, had_errors) = WINDOWS_1257.decode(b"abc\x80\xE4"); |
| 5203 |
match cow { |
| 5204 |
Cow::Borrowed(_) => unreachable!(), |
| 5205 |
Cow::Owned(s) => { |
| 5206 |
assert_eq!(s, "abc\u{20AC}\u{00E4}"); |
| 5207 |
} |
| 5208 |
} |
| 5209 |
assert_eq!(encoding, WINDOWS_1257); |
| 5210 |
assert!(!had_errors); |
| 5211 |
} |
| 5212 |
|
| 5213 |
#[test] |
| 5214 |
fn test_decode_invalid_windows_1257_to_cow() { |
| 5215 |
let (cow, encoding, had_errors) = WINDOWS_1257.decode(b"abc\x80\xA1\xE4"); |
| 5216 |
match cow { |
| 5217 |
Cow::Borrowed(_) => unreachable!(), |
| 5218 |
Cow::Owned(s) => { |
| 5219 |
assert_eq!(s, "abc\u{20AC}\u{FFFD}\u{00E4}"); |
| 5220 |
} |
| 5221 |
} |
| 5222 |
assert_eq!(encoding, WINDOWS_1257); |
| 5223 |
assert!(had_errors); |
| 5224 |
} |
| 5225 |
|
| 5226 |
#[test] |
| 5227 |
fn test_decode_ascii_only_windows_1257_to_cow() { |
| 5228 |
let (cow, encoding, had_errors) = WINDOWS_1257.decode(b"abc"); |
| 5229 |
match cow { |
| 5230 |
Cow::Borrowed(s) => { |
| 5231 |
assert_eq!(s, "abc"); |
| 5232 |
} |
| 5233 |
Cow::Owned(_) => unreachable!(), |
| 5234 |
} |
| 5235 |
assert_eq!(encoding, WINDOWS_1257); |
| 5236 |
assert!(!had_errors); |
| 5237 |
} |
| 5238 |
|
| 5239 |
#[test] |
| 5240 |
fn test_decode_bomful_valid_utf8_as_windows_1257_to_cow() { |
| 5241 |
let (cow, encoding, had_errors) = WINDOWS_1257.decode(b"\xEF\xBB\xBF\xE2\x82\xAC\xC3\xA4"); |
| 5242 |
match cow { |
| 5243 |
Cow::Borrowed(s) => { |
| 5244 |
assert_eq!(s, "\u{20AC}\u{00E4}"); |
| 5245 |
} |
| 5246 |
Cow::Owned(_) => unreachable!(), |
| 5247 |
} |
| 5248 |
assert_eq!(encoding, UTF_8); |
| 5249 |
assert!(!had_errors); |
| 5250 |
} |
| 5251 |
|
| 5252 |
#[test] |
| 5253 |
fn test_decode_bomful_invalid_utf8_as_windows_1257_to_cow() { |
| 5254 |
let (cow, encoding, had_errors) = |
| 5255 |
WINDOWS_1257.decode(b"\xEF\xBB\xBF\xE2\x82\xAC\x80\xC3\xA4"); |
| 5256 |
match cow { |
| 5257 |
Cow::Borrowed(_) => unreachable!(), |
| 5258 |
Cow::Owned(s) => { |
| 5259 |
assert_eq!(s, "\u{20AC}\u{FFFD}\u{00E4}"); |
| 5260 |
} |
| 5261 |
} |
| 5262 |
assert_eq!(encoding, UTF_8); |
| 5263 |
assert!(had_errors); |
| 5264 |
} |
| 5265 |
|
| 5266 |
#[test] |
| 5267 |
fn test_decode_bomful_valid_utf8_as_utf_8_to_cow() { |
| 5268 |
let (cow, encoding, had_errors) = UTF_8.decode(b"\xEF\xBB\xBF\xE2\x82\xAC\xC3\xA4"); |
| 5269 |
match cow { |
| 5270 |
Cow::Borrowed(s) => { |
| 5271 |
assert_eq!(s, "\u{20AC}\u{00E4}"); |
| 5272 |
} |
| 5273 |
Cow::Owned(_) => unreachable!(), |
| 5274 |
} |
| 5275 |
assert_eq!(encoding, UTF_8); |
| 5276 |
assert!(!had_errors); |
| 5277 |
} |
| 5278 |
|
| 5279 |
#[test] |
| 5280 |
fn test_decode_bomful_invalid_utf8_as_utf_8_to_cow() { |
| 5281 |
let (cow, encoding, had_errors) = UTF_8.decode(b"\xEF\xBB\xBF\xE2\x82\xAC\x80\xC3\xA4"); |
| 5282 |
match cow { |
| 5283 |
Cow::Borrowed(_) => unreachable!(), |
| 5284 |
Cow::Owned(s) => { |
| 5285 |
assert_eq!(s, "\u{20AC}\u{FFFD}\u{00E4}"); |
| 5286 |
} |
| 5287 |
} |
| 5288 |
assert_eq!(encoding, UTF_8); |
| 5289 |
assert!(had_errors); |
| 5290 |
} |
| 5291 |
|
| 5292 |
#[test] |
| 5293 |
fn test_decode_bomful_valid_utf8_as_utf_8_to_cow_with_bom_removal() { |
| 5294 |
let (cow, had_errors) = UTF_8.decode_with_bom_removal(b"\xEF\xBB\xBF\xE2\x82\xAC\xC3\xA4"); |
| 5295 |
match cow { |
| 5296 |
Cow::Borrowed(s) => { |
| 5297 |
assert_eq!(s, "\u{20AC}\u{00E4}"); |
| 5298 |
} |
| 5299 |
Cow::Owned(_) => unreachable!(), |
| 5300 |
} |
| 5301 |
assert!(!had_errors); |
| 5302 |
} |
| 5303 |
|
| 5304 |
#[test] |
| 5305 |
fn test_decode_bomful_valid_utf8_as_windows_1257_to_cow_with_bom_removal() { |
| 5306 |
let (cow, had_errors) = |
| 5307 |
WINDOWS_1257.decode_with_bom_removal(b"\xEF\xBB\xBF\xE2\x82\xAC\xC3\xA4"); |
| 5308 |
match cow { |
| 5309 |
Cow::Borrowed(_) => unreachable!(), |
| 5310 |
Cow::Owned(s) => { |
| 5311 |
assert_eq!( |
| 5312 |
s, |
| 5313 |
"\u{013C}\u{00BB}\u{00E6}\u{0101}\u{201A}\u{00AC}\u{0106}\u{00A4}" |
| 5314 |
); |
| 5315 |
} |
| 5316 |
} |
| 5317 |
assert!(!had_errors); |
| 5318 |
} |
| 5319 |
|
| 5320 |
#[test] |
| 5321 |
fn test_decode_valid_windows_1257_to_cow_with_bom_removal() { |
| 5322 |
let (cow, had_errors) = WINDOWS_1257.decode_with_bom_removal(b"abc\x80\xE4"); |
| 5323 |
match cow { |
| 5324 |
Cow::Borrowed(_) => unreachable!(), |
| 5325 |
Cow::Owned(s) => { |
| 5326 |
assert_eq!(s, "abc\u{20AC}\u{00E4}"); |
| 5327 |
} |
| 5328 |
} |
| 5329 |
assert!(!had_errors); |
| 5330 |
} |
| 5331 |
|
| 5332 |
#[test] |
| 5333 |
fn test_decode_invalid_windows_1257_to_cow_with_bom_removal() { |
| 5334 |
let (cow, had_errors) = WINDOWS_1257.decode_with_bom_removal(b"abc\x80\xA1\xE4"); |
| 5335 |
match cow { |
| 5336 |
Cow::Borrowed(_) => unreachable!(), |
| 5337 |
Cow::Owned(s) => { |
| 5338 |
assert_eq!(s, "abc\u{20AC}\u{FFFD}\u{00E4}"); |
| 5339 |
} |
| 5340 |
} |
| 5341 |
assert!(had_errors); |
| 5342 |
} |
| 5343 |
|
| 5344 |
#[test] |
| 5345 |
fn test_decode_ascii_only_windows_1257_to_cow_with_bom_removal() { |
| 5346 |
let (cow, had_errors) = WINDOWS_1257.decode_with_bom_removal(b"abc"); |
| 5347 |
match cow { |
| 5348 |
Cow::Borrowed(s) => { |
| 5349 |
assert_eq!(s, "abc"); |
| 5350 |
} |
| 5351 |
Cow::Owned(_) => unreachable!(), |
| 5352 |
} |
| 5353 |
assert!(!had_errors); |
| 5354 |
} |
| 5355 |
|
| 5356 |
#[test] |
| 5357 |
fn test_decode_bomful_valid_utf8_to_cow_without_bom_handling() { |
| 5358 |
let (cow, had_errors) = |
| 5359 |
UTF_8.decode_without_bom_handling(b"\xEF\xBB\xBF\xE2\x82\xAC\xC3\xA4"); |
| 5360 |
match cow { |
| 5361 |
Cow::Borrowed(s) => { |
| 5362 |
assert_eq!(s, "\u{FEFF}\u{20AC}\u{00E4}"); |
| 5363 |
} |
| 5364 |
Cow::Owned(_) => unreachable!(), |
| 5365 |
} |
| 5366 |
assert!(!had_errors); |
| 5367 |
} |
| 5368 |
|
| 5369 |
#[test] |
| 5370 |
fn test_decode_bomful_invalid_utf8_to_cow_without_bom_handling() { |
| 5371 |
let (cow, had_errors) = |
| 5372 |
UTF_8.decode_without_bom_handling(b"\xEF\xBB\xBF\xE2\x82\xAC\x80\xC3\xA4"); |
| 5373 |
match cow { |
| 5374 |
Cow::Borrowed(_) => unreachable!(), |
| 5375 |
Cow::Owned(s) => { |
| 5376 |
assert_eq!(s, "\u{FEFF}\u{20AC}\u{FFFD}\u{00E4}"); |
| 5377 |
} |
| 5378 |
} |
| 5379 |
assert!(had_errors); |
| 5380 |
} |
| 5381 |
|
| 5382 |
#[test] |
| 5383 |
fn test_decode_valid_windows_1257_to_cow_without_bom_handling() { |
| 5384 |
let (cow, had_errors) = WINDOWS_1257.decode_without_bom_handling(b"abc\x80\xE4"); |
| 5385 |
match cow { |
| 5386 |
Cow::Borrowed(_) => unreachable!(), |
| 5387 |
Cow::Owned(s) => { |
| 5388 |
assert_eq!(s, "abc\u{20AC}\u{00E4}"); |
| 5389 |
} |
| 5390 |
} |
| 5391 |
assert!(!had_errors); |
| 5392 |
} |
| 5393 |
|
| 5394 |
#[test] |
| 5395 |
fn test_decode_invalid_windows_1257_to_cow_without_bom_handling() { |
| 5396 |
let (cow, had_errors) = WINDOWS_1257.decode_without_bom_handling(b"abc\x80\xA1\xE4"); |
| 5397 |
match cow { |
| 5398 |
Cow::Borrowed(_) => unreachable!(), |
| 5399 |
Cow::Owned(s) => { |
| 5400 |
assert_eq!(s, "abc\u{20AC}\u{FFFD}\u{00E4}"); |
| 5401 |
} |
| 5402 |
} |
| 5403 |
assert!(had_errors); |
| 5404 |
} |
| 5405 |
|
| 5406 |
#[test] |
| 5407 |
fn test_decode_ascii_only_windows_1257_to_cow_without_bom_handling() { |
| 5408 |
let (cow, had_errors) = WINDOWS_1257.decode_without_bom_handling(b"abc"); |
| 5409 |
match cow { |
| 5410 |
Cow::Borrowed(s) => { |
| 5411 |
assert_eq!(s, "abc"); |
| 5412 |
} |
| 5413 |
Cow::Owned(_) => unreachable!(), |
| 5414 |
} |
| 5415 |
assert!(!had_errors); |
| 5416 |
} |
| 5417 |
|
| 5418 |
#[test] |
| 5419 |
fn test_decode_bomful_valid_utf8_to_cow_without_bom_handling_and_without_replacement() { |
| 5420 |
match UTF_8.decode_without_bom_handling_and_without_replacement( |
| 5421 |
b"\xEF\xBB\xBF\xE2\x82\xAC\xC3\xA4", |
| 5422 |
) { |
| 5423 |
Some(cow) => match cow { |
| 5424 |
Cow::Borrowed(s) => { |
| 5425 |
assert_eq!(s, "\u{FEFF}\u{20AC}\u{00E4}"); |
| 5426 |
} |
| 5427 |
Cow::Owned(_) => unreachable!(), |
| 5428 |
}, |
| 5429 |
None => unreachable!(), |
| 5430 |
} |
| 5431 |
} |
| 5432 |
|
| 5433 |
#[test] |
| 5434 |
fn test_decode_bomful_invalid_utf8_to_cow_without_bom_handling_and_without_replacement() { |
| 5435 |
assert!(UTF_8 |
| 5436 |
.decode_without_bom_handling_and_without_replacement( |
| 5437 |
b"\xEF\xBB\xBF\xE2\x82\xAC\x80\xC3\xA4" |
| 5438 |
) |
| 5439 |
.is_none()); |
| 5440 |
} |
| 5441 |
|
| 5442 |
#[test] |
| 5443 |
fn test_decode_valid_windows_1257_to_cow_without_bom_handling_and_without_replacement() { |
| 5444 |
match WINDOWS_1257.decode_without_bom_handling_and_without_replacement(b"abc\x80\xE4") { |
| 5445 |
Some(cow) => match cow { |
| 5446 |
Cow::Borrowed(_) => unreachable!(), |
| 5447 |
Cow::Owned(s) => { |
| 5448 |
assert_eq!(s, "abc\u{20AC}\u{00E4}"); |
| 5449 |
} |
| 5450 |
}, |
| 5451 |
None => unreachable!(), |
| 5452 |
} |
| 5453 |
} |
| 5454 |
|
| 5455 |
#[test] |
| 5456 |
fn test_decode_invalid_windows_1257_to_cow_without_bom_handling_and_without_replacement() { |
| 5457 |
assert!(WINDOWS_1257 |
| 5458 |
.decode_without_bom_handling_and_without_replacement(b"abc\x80\xA1\xE4") |
| 5459 |
.is_none()); |
| 5460 |
} |
| 5461 |
|
| 5462 |
#[test] |
| 5463 |
fn test_decode_ascii_only_windows_1257_to_cow_without_bom_handling_and_without_replacement() { |
| 5464 |
match WINDOWS_1257.decode_without_bom_handling_and_without_replacement(b"abc") { |
| 5465 |
Some(cow) => match cow { |
| 5466 |
Cow::Borrowed(s) => { |
| 5467 |
assert_eq!(s, "abc"); |
| 5468 |
} |
| 5469 |
Cow::Owned(_) => unreachable!(), |
| 5470 |
}, |
| 5471 |
None => unreachable!(), |
| 5472 |
} |
| 5473 |
} |
| 5474 |
|
| 5475 |
#[test] |
| 5476 |
fn test_encode_ascii_only_windows_1257_to_cow() { |
| 5477 |
let (cow, encoding, had_errors) = WINDOWS_1257.encode("abc"); |
| 5478 |
match cow { |
| 5479 |
Cow::Borrowed(s) => { |
| 5480 |
assert_eq!(s, b"abc"); |
| 5481 |
} |
| 5482 |
Cow::Owned(_) => unreachable!(), |
| 5483 |
} |
| 5484 |
assert_eq!(encoding, WINDOWS_1257); |
| 5485 |
assert!(!had_errors); |
| 5486 |
} |
| 5487 |
|
| 5488 |
#[test] |
| 5489 |
fn test_encode_valid_windows_1257_to_cow() { |
| 5490 |
let (cow, encoding, had_errors) = WINDOWS_1257.encode("abc\u{20AC}\u{00E4}"); |
| 5491 |
match cow { |
| 5492 |
Cow::Borrowed(_) => unreachable!(), |
| 5493 |
Cow::Owned(s) => { |
| 5494 |
assert_eq!(s, b"abc\x80\xE4"); |
| 5495 |
} |
| 5496 |
} |
| 5497 |
assert_eq!(encoding, WINDOWS_1257); |
| 5498 |
assert!(!had_errors); |
| 5499 |
} |
| 5500 |
|
| 5501 |
#[test] |
| 5502 |
fn test_utf16_space_with_one_bom_byte() { |
| 5503 |
let mut decoder = UTF_16LE.new_decoder(); |
| 5504 |
let mut dst = [0u16; 12]; |
| 5505 |
{ |
| 5506 |
let needed = decoder.max_utf16_buffer_length(1).unwrap(); |
| 5507 |
let (result, _, _, _) = decoder.decode_to_utf16(b"\xFF", &mut dst[..needed], false); |
| 5508 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5509 |
} |
| 5510 |
{ |
| 5511 |
let needed = decoder.max_utf16_buffer_length(1).unwrap(); |
| 5512 |
let (result, _, _, _) = decoder.decode_to_utf16(b"\xFF", &mut dst[..needed], true); |
| 5513 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5514 |
} |
| 5515 |
} |
| 5516 |
|
| 5517 |
#[test] |
| 5518 |
fn test_utf8_space_with_one_bom_byte() { |
| 5519 |
let mut decoder = UTF_8.new_decoder(); |
| 5520 |
let mut dst = [0u16; 12]; |
| 5521 |
{ |
| 5522 |
let needed = decoder.max_utf16_buffer_length(1).unwrap(); |
| 5523 |
let (result, _, _, _) = decoder.decode_to_utf16(b"\xFF", &mut dst[..needed], false); |
| 5524 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5525 |
} |
| 5526 |
{ |
| 5527 |
let needed = decoder.max_utf16_buffer_length(1).unwrap(); |
| 5528 |
let (result, _, _, _) = decoder.decode_to_utf16(b"\xFF", &mut dst[..needed], true); |
| 5529 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5530 |
} |
| 5531 |
} |
| 5532 |
|
| 5533 |
#[test] |
| 5534 |
fn test_utf16_space_with_two_bom_bytes() { |
| 5535 |
let mut decoder = UTF_16LE.new_decoder(); |
| 5536 |
let mut dst = [0u16; 12]; |
| 5537 |
{ |
| 5538 |
let needed = decoder.max_utf16_buffer_length(1).unwrap(); |
| 5539 |
let (result, _, _, _) = decoder.decode_to_utf16(b"\xEF", &mut dst[..needed], false); |
| 5540 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5541 |
} |
| 5542 |
{ |
| 5543 |
let needed = decoder.max_utf16_buffer_length(1).unwrap(); |
| 5544 |
let (result, _, _, _) = decoder.decode_to_utf16(b"\xBB", &mut dst[..needed], false); |
| 5545 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5546 |
} |
| 5547 |
{ |
| 5548 |
let needed = decoder.max_utf16_buffer_length(1).unwrap(); |
| 5549 |
let (result, _, _, _) = decoder.decode_to_utf16(b"\xFF", &mut dst[..needed], true); |
| 5550 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5551 |
} |
| 5552 |
} |
| 5553 |
|
| 5554 |
#[test] |
| 5555 |
fn test_utf8_space_with_two_bom_bytes() { |
| 5556 |
let mut decoder = UTF_8.new_decoder(); |
| 5557 |
let mut dst = [0u16; 12]; |
| 5558 |
{ |
| 5559 |
let needed = decoder.max_utf16_buffer_length(1).unwrap(); |
| 5560 |
let (result, _, _, _) = decoder.decode_to_utf16(b"\xEF", &mut dst[..needed], false); |
| 5561 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5562 |
} |
| 5563 |
{ |
| 5564 |
let needed = decoder.max_utf16_buffer_length(1).unwrap(); |
| 5565 |
let (result, _, _, _) = decoder.decode_to_utf16(b"\xBB", &mut dst[..needed], false); |
| 5566 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5567 |
} |
| 5568 |
{ |
| 5569 |
let needed = decoder.max_utf16_buffer_length(1).unwrap(); |
| 5570 |
let (result, _, _, _) = decoder.decode_to_utf16(b"\xFF", &mut dst[..needed], true); |
| 5571 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5572 |
} |
| 5573 |
} |
| 5574 |
|
| 5575 |
#[test] |
| 5576 |
fn test_utf16_space_with_one_bom_byte_and_a_second_byte_in_same_call() { |
| 5577 |
let mut decoder = UTF_16LE.new_decoder(); |
| 5578 |
let mut dst = [0u16; 12]; |
| 5579 |
{ |
| 5580 |
let needed = decoder.max_utf16_buffer_length(2).unwrap(); |
| 5581 |
let (result, _, _, _) = decoder.decode_to_utf16(b"\xFF\xFF", &mut dst[..needed], true); |
| 5582 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5583 |
} |
| 5584 |
} |
| 5585 |
|
| 5586 |
#[test] |
| 5587 |
fn test_too_short_buffer_with_iso_2022_jp_ascii_from_utf8() { |
| 5588 |
let mut dst = [0u8; 8]; |
| 5589 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5590 |
{ |
| 5591 |
let (result, _, _, _) = encoder.encode_from_utf8("", &mut dst[..], false); |
| 5592 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5593 |
} |
| 5594 |
{ |
| 5595 |
let (result, _, _, _) = encoder.encode_from_utf8("", &mut dst[..], true); |
| 5596 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5597 |
} |
| 5598 |
} |
| 5599 |
|
| 5600 |
#[test] |
| 5601 |
fn test_too_short_buffer_with_iso_2022_jp_roman_from_utf8() { |
| 5602 |
let mut dst = [0u8; 16]; |
| 5603 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5604 |
{ |
| 5605 |
let (result, _, _, _) = encoder.encode_from_utf8("\u{A5}", &mut dst[..], false); |
| 5606 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5607 |
} |
| 5608 |
{ |
| 5609 |
let (result, _, _, _) = encoder.encode_from_utf8("", &mut dst[..8], false); |
| 5610 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5611 |
} |
| 5612 |
{ |
| 5613 |
let (result, _, _, _) = encoder.encode_from_utf8("", &mut dst[..8], true); |
| 5614 |
assert_eq!(result, CoderResult::OutputFull); |
| 5615 |
} |
| 5616 |
} |
| 5617 |
|
| 5618 |
#[test] |
| 5619 |
fn test_buffer_end_iso_2022_jp_from_utf8() { |
| 5620 |
let mut dst = [0u8; 18]; |
| 5621 |
{ |
| 5622 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5623 |
let (result, _, _, _) = |
| 5624 |
encoder.encode_from_utf8("\u{A5}\u{1F4A9}", &mut dst[..], false); |
| 5625 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5626 |
} |
| 5627 |
{ |
| 5628 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5629 |
let (result, _, _, _) = encoder.encode_from_utf8("\u{A5}\u{1F4A9}", &mut dst[..], true); |
| 5630 |
assert_eq!(result, CoderResult::OutputFull); |
| 5631 |
} |
| 5632 |
{ |
| 5633 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5634 |
let (result, _, _, _) = encoder.encode_from_utf8("\u{1F4A9}", &mut dst[..13], false); |
| 5635 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5636 |
} |
| 5637 |
{ |
| 5638 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5639 |
let (result, _, _, _) = encoder.encode_from_utf8("\u{1F4A9}", &mut dst[..13], true); |
| 5640 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5641 |
} |
| 5642 |
} |
| 5643 |
|
| 5644 |
#[test] |
| 5645 |
fn test_too_short_buffer_with_iso_2022_jp_ascii_from_utf16() { |
| 5646 |
let mut dst = [0u8; 8]; |
| 5647 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5648 |
{ |
| 5649 |
let (result, _, _, _) = encoder.encode_from_utf16(&[0u16; 0], &mut dst[..], false); |
| 5650 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5651 |
} |
| 5652 |
{ |
| 5653 |
let (result, _, _, _) = encoder.encode_from_utf16(&[0u16; 0], &mut dst[..], true); |
| 5654 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5655 |
} |
| 5656 |
} |
| 5657 |
|
| 5658 |
#[test] |
| 5659 |
fn test_too_short_buffer_with_iso_2022_jp_roman_from_utf16() { |
| 5660 |
let mut dst = [0u8; 16]; |
| 5661 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5662 |
{ |
| 5663 |
let (result, _, _, _) = encoder.encode_from_utf16(&[0xA5u16], &mut dst[..], false); |
| 5664 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5665 |
} |
| 5666 |
{ |
| 5667 |
let (result, _, _, _) = encoder.encode_from_utf16(&[0u16; 0], &mut dst[..8], false); |
| 5668 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5669 |
} |
| 5670 |
{ |
| 5671 |
let (result, _, _, _) = encoder.encode_from_utf16(&[0u16; 0], &mut dst[..8], true); |
| 5672 |
assert_eq!(result, CoderResult::OutputFull); |
| 5673 |
} |
| 5674 |
} |
| 5675 |
|
| 5676 |
#[test] |
| 5677 |
fn test_buffer_end_iso_2022_jp_from_utf16() { |
| 5678 |
let mut dst = [0u8; 18]; |
| 5679 |
{ |
| 5680 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5681 |
let (result, _, _, _) = |
| 5682 |
encoder.encode_from_utf16(&[0xA5u16, 0xD83Du16, 0xDCA9u16], &mut dst[..], false); |
| 5683 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5684 |
} |
| 5685 |
{ |
| 5686 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5687 |
let (result, _, _, _) = |
| 5688 |
encoder.encode_from_utf16(&[0xA5u16, 0xD83Du16, 0xDCA9u16], &mut dst[..], true); |
| 5689 |
assert_eq!(result, CoderResult::OutputFull); |
| 5690 |
} |
| 5691 |
{ |
| 5692 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5693 |
let (result, _, _, _) = |
| 5694 |
encoder.encode_from_utf16(&[0xD83Du16, 0xDCA9u16], &mut dst[..13], false); |
| 5695 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5696 |
} |
| 5697 |
{ |
| 5698 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5699 |
let (result, _, _, _) = |
| 5700 |
encoder.encode_from_utf16(&[0xD83Du16, 0xDCA9u16], &mut dst[..13], true); |
| 5701 |
assert_eq!(result, CoderResult::InputEmpty); |
| 5702 |
} |
| 5703 |
} |
| 5704 |
|
| 5705 |
#[test] |
| 5706 |
fn test_buffer_end_utf16be() { |
| 5707 |
let mut decoder = UTF_16BE.new_decoder_without_bom_handling(); |
| 5708 |
let mut dest = [0u8; 4]; |
| 5709 |
|
| 5710 |
assert_eq!( |
| 5711 |
decoder.decode_to_utf8(&[0xD8, 0x00], &mut dest, false), |
| 5712 |
(CoderResult::InputEmpty, 2, 0, false) |
| 5713 |
); |
| 5714 |
|
| 5715 |
let _ = decoder.decode_to_utf8(&[0xD8, 0x00], &mut dest, true); |
| 5716 |
} |
| 5717 |
|
| 5718 |
#[test] |
| 5719 |
fn test_hash() { |
| 5720 |
let mut encodings = ::alloc::collections::btree_set::BTreeSet::new(); |
| 5721 |
encodings.insert(UTF_8); |
| 5722 |
encodings.insert(ISO_2022_JP); |
| 5723 |
assert!(encodings.contains(UTF_8)); |
| 5724 |
assert!(encodings.contains(ISO_2022_JP)); |
| 5725 |
assert!(!encodings.contains(WINDOWS_1252)); |
| 5726 |
encodings.remove(ISO_2022_JP); |
| 5727 |
assert!(!encodings.contains(ISO_2022_JP)); |
| 5728 |
} |
| 5729 |
|
| 5730 |
#[test] |
| 5731 |
fn test_iso_2022_jp_ncr_extra_from_utf16() { |
| 5732 |
let mut dst = [0u8; 17]; |
| 5733 |
{ |
| 5734 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5735 |
let (result, _, _, _) = |
| 5736 |
encoder.encode_from_utf16(&[0x3041u16, 0xFFFFu16], &mut dst[..], true); |
| 5737 |
assert_eq!(result, CoderResult::OutputFull); |
| 5738 |
} |
| 5739 |
} |
| 5740 |
|
| 5741 |
#[test] |
| 5742 |
fn test_iso_2022_jp_ncr_extra_from_utf8() { |
| 5743 |
let mut dst = [0u8; 17]; |
| 5744 |
{ |
| 5745 |
let mut encoder = ISO_2022_JP.new_encoder(); |
| 5746 |
let (result, _, _, _) = |
| 5747 |
encoder.encode_from_utf8("\u{3041}\u{FFFF}", &mut dst[..], true); |
| 5748 |
assert_eq!(result, CoderResult::OutputFull); |
| 5749 |
} |
| 5750 |
} |
| 5751 |
|
| 5752 |
#[test] |
| 5753 |
fn test_max_length_with_bom_to_utf8() { |
| 5754 |
let mut output = [0u8; 20]; |
| 5755 |
let mut decoder = REPLACEMENT.new_decoder(); |
| 5756 |
let input = b"\xEF\xBB\xBFA"; |
| 5757 |
{ |
| 5758 |
let needed = decoder |
| 5759 |
.max_utf8_buffer_length_without_replacement(input.len()) |
| 5760 |
.unwrap(); |
| 5761 |
let (result, read, written) = |
| 5762 |
decoder.decode_to_utf8_without_replacement(input, &mut output[..needed], true); |
| 5763 |
assert_eq!(result, DecoderResult::InputEmpty); |
| 5764 |
assert_eq!(read, input.len()); |
| 5765 |
assert_eq!(written, 1); |
| 5766 |
assert_eq!(output[0], 0x41); |
| 5767 |
} |
| 5768 |
} |
| 5769 |
|
| 5770 |
#[cfg(feature = "serde")] |
| 5771 |
#[test] |
| 5772 |
fn test_serde() { |
| 5773 |
let demo = Demo { |
| 5774 |
num: 42, |
| 5775 |
name: "foo".into(), |
| 5776 |
enc: UTF_8, |
| 5777 |
}; |
| 5778 |
|
| 5779 |
let serialized = serde_json::to_string(&demo).unwrap(); |
| 5780 |
|
| 5781 |
let deserialized: Demo = serde_json::from_str(&serialized).unwrap(); |
| 5782 |
assert_eq!(deserialized, demo); |
| 5783 |
|
| 5784 |
let bincoded = bincode::serialize(&demo).unwrap(); |
| 5785 |
let debincoded: Demo = bincode::deserialize(&bincoded[..]).unwrap(); |
| 5786 |
assert_eq!(debincoded, demo); |
| 5787 |
} |
| 5788 |
|
| 5789 |
#[test] |
| 5790 |
fn test_is_single_byte() { |
| 5791 |
assert!(!BIG5.is_single_byte()); |
| 5792 |
assert!(!EUC_JP.is_single_byte()); |
| 5793 |
assert!(!EUC_KR.is_single_byte()); |
| 5794 |
assert!(!GB18030.is_single_byte()); |
| 5795 |
assert!(!GBK.is_single_byte()); |
| 5796 |
assert!(!REPLACEMENT.is_single_byte()); |
| 5797 |
assert!(!SHIFT_JIS.is_single_byte()); |
| 5798 |
assert!(!UTF_8.is_single_byte()); |
| 5799 |
assert!(!UTF_16BE.is_single_byte()); |
| 5800 |
assert!(!UTF_16LE.is_single_byte()); |
| 5801 |
assert!(!ISO_2022_JP.is_single_byte()); |
| 5802 |
|
| 5803 |
assert!(IBM866.is_single_byte()); |
| 5804 |
assert!(ISO_8859_2.is_single_byte()); |
| 5805 |
assert!(ISO_8859_3.is_single_byte()); |
| 5806 |
assert!(ISO_8859_4.is_single_byte()); |
| 5807 |
assert!(ISO_8859_5.is_single_byte()); |
| 5808 |
assert!(ISO_8859_6.is_single_byte()); |
| 5809 |
assert!(ISO_8859_7.is_single_byte()); |
| 5810 |
assert!(ISO_8859_8.is_single_byte()); |
| 5811 |
assert!(ISO_8859_10.is_single_byte()); |
| 5812 |
assert!(ISO_8859_13.is_single_byte()); |
| 5813 |
assert!(ISO_8859_14.is_single_byte()); |
| 5814 |
assert!(ISO_8859_15.is_single_byte()); |
| 5815 |
assert!(ISO_8859_16.is_single_byte()); |
| 5816 |
assert!(ISO_8859_8_I.is_single_byte()); |
| 5817 |
assert!(KOI8_R.is_single_byte()); |
| 5818 |
assert!(KOI8_U.is_single_byte()); |
| 5819 |
assert!(MACINTOSH.is_single_byte()); |
| 5820 |
assert!(WINDOWS_874.is_single_byte()); |
| 5821 |
assert!(WINDOWS_1250.is_single_byte()); |
| 5822 |
assert!(WINDOWS_1251.is_single_byte()); |
| 5823 |
assert!(WINDOWS_1252.is_single_byte()); |
| 5824 |
assert!(WINDOWS_1253.is_single_byte()); |
| 5825 |
assert!(WINDOWS_1254.is_single_byte()); |
| 5826 |
assert!(WINDOWS_1255.is_single_byte()); |
| 5827 |
assert!(WINDOWS_1256.is_single_byte()); |
| 5828 |
assert!(WINDOWS_1257.is_single_byte()); |
| 5829 |
assert!(WINDOWS_1258.is_single_byte()); |
| 5830 |
assert!(X_MAC_CYRILLIC.is_single_byte()); |
| 5831 |
assert!(X_USER_DEFINED.is_single_byte()); |
| 5832 |
} |
| 5833 |
|
| 5834 |
#[test] |
| 5835 |
fn test_latin1_byte_compatible_up_to() { |
| 5836 |
let buffer = b"a\x81\xB6\xF6\xF0\x82\xB4"; |
| 5837 |
assert_eq!( |
| 5838 |
BIG5.new_decoder_without_bom_handling() |
| 5839 |
.latin1_byte_compatible_up_to(buffer) |
| 5840 |
.unwrap(), |
| 5841 |
1 |
| 5842 |
); |
| 5843 |
assert_eq!( |
| 5844 |
EUC_JP |
| 5845 |
.new_decoder_without_bom_handling() |
| 5846 |
.latin1_byte_compatible_up_to(buffer) |
| 5847 |
.unwrap(), |
| 5848 |
1 |
| 5849 |
); |
| 5850 |
assert_eq!( |
| 5851 |
EUC_KR |
| 5852 |
.new_decoder_without_bom_handling() |
| 5853 |
.latin1_byte_compatible_up_to(buffer) |
| 5854 |
.unwrap(), |
| 5855 |
1 |
| 5856 |
); |
| 5857 |
assert_eq!( |
| 5858 |
GB18030 |
| 5859 |
.new_decoder_without_bom_handling() |
| 5860 |
.latin1_byte_compatible_up_to(buffer) |
| 5861 |
.unwrap(), |
| 5862 |
1 |
| 5863 |
); |
| 5864 |
assert_eq!( |
| 5865 |
GBK.new_decoder_without_bom_handling() |
| 5866 |
.latin1_byte_compatible_up_to(buffer) |
| 5867 |
.unwrap(), |
| 5868 |
1 |
| 5869 |
); |
| 5870 |
assert!(REPLACEMENT |
| 5871 |
.new_decoder_without_bom_handling() |
| 5872 |
.latin1_byte_compatible_up_to(buffer) |
| 5873 |
.is_none()); |
| 5874 |
assert_eq!( |
| 5875 |
SHIFT_JIS |
| 5876 |
.new_decoder_without_bom_handling() |
| 5877 |
.latin1_byte_compatible_up_to(buffer) |
| 5878 |
.unwrap(), |
| 5879 |
1 |
| 5880 |
); |
| 5881 |
assert_eq!( |
| 5882 |
UTF_8 |
| 5883 |
.new_decoder_without_bom_handling() |
| 5884 |
.latin1_byte_compatible_up_to(buffer) |
| 5885 |
.unwrap(), |
| 5886 |
1 |
| 5887 |
); |
| 5888 |
assert!(UTF_16BE |
| 5889 |
.new_decoder_without_bom_handling() |
| 5890 |
.latin1_byte_compatible_up_to(buffer) |
| 5891 |
.is_none()); |
| 5892 |
assert!(UTF_16LE |
| 5893 |
.new_decoder_without_bom_handling() |
| 5894 |
.latin1_byte_compatible_up_to(buffer) |
| 5895 |
.is_none()); |
| 5896 |
assert_eq!( |
| 5897 |
ISO_2022_JP |
| 5898 |
.new_decoder_without_bom_handling() |
| 5899 |
.latin1_byte_compatible_up_to(buffer) |
| 5900 |
.unwrap(), |
| 5901 |
1 |
| 5902 |
); |
| 5903 |
|
| 5904 |
assert_eq!( |
| 5905 |
IBM866 |
| 5906 |
.new_decoder_without_bom_handling() |
| 5907 |
.latin1_byte_compatible_up_to(buffer) |
| 5908 |
.unwrap(), |
| 5909 |
1 |
| 5910 |
); |
| 5911 |
assert_eq!( |
| 5912 |
ISO_8859_2 |
| 5913 |
.new_decoder_without_bom_handling() |
| 5914 |
.latin1_byte_compatible_up_to(buffer) |
| 5915 |
.unwrap(), |
| 5916 |
2 |
| 5917 |
); |
| 5918 |
assert_eq!( |
| 5919 |
ISO_8859_3 |
| 5920 |
.new_decoder_without_bom_handling() |
| 5921 |
.latin1_byte_compatible_up_to(buffer) |
| 5922 |
.unwrap(), |
| 5923 |
2 |
| 5924 |
); |
| 5925 |
assert_eq!( |
| 5926 |
ISO_8859_4 |
| 5927 |
.new_decoder_without_bom_handling() |
| 5928 |
.latin1_byte_compatible_up_to(buffer) |
| 5929 |
.unwrap(), |
| 5930 |
2 |
| 5931 |
); |
| 5932 |
assert_eq!( |
| 5933 |
ISO_8859_5 |
| 5934 |
.new_decoder_without_bom_handling() |
| 5935 |
.latin1_byte_compatible_up_to(buffer) |
| 5936 |
.unwrap(), |
| 5937 |
2 |
| 5938 |
); |
| 5939 |
assert_eq!( |
| 5940 |
ISO_8859_6 |
| 5941 |
.new_decoder_without_bom_handling() |
| 5942 |
.latin1_byte_compatible_up_to(buffer) |
| 5943 |
.unwrap(), |
| 5944 |
2 |
| 5945 |
); |
| 5946 |
assert_eq!( |
| 5947 |
ISO_8859_7 |
| 5948 |
.new_decoder_without_bom_handling() |
| 5949 |
.latin1_byte_compatible_up_to(buffer) |
| 5950 |
.unwrap(), |
| 5951 |
2 |
| 5952 |
); |
| 5953 |
assert_eq!( |
| 5954 |
ISO_8859_8 |
| 5955 |
.new_decoder_without_bom_handling() |
| 5956 |
.latin1_byte_compatible_up_to(buffer) |
| 5957 |
.unwrap(), |
| 5958 |
3 |
| 5959 |
); |
| 5960 |
assert_eq!( |
| 5961 |
ISO_8859_10 |
| 5962 |
.new_decoder_without_bom_handling() |
| 5963 |
.latin1_byte_compatible_up_to(buffer) |
| 5964 |
.unwrap(), |
| 5965 |
2 |
| 5966 |
); |
| 5967 |
assert_eq!( |
| 5968 |
ISO_8859_13 |
| 5969 |
.new_decoder_without_bom_handling() |
| 5970 |
.latin1_byte_compatible_up_to(buffer) |
| 5971 |
.unwrap(), |
| 5972 |
4 |
| 5973 |
); |
| 5974 |
assert_eq!( |
| 5975 |
ISO_8859_14 |
| 5976 |
.new_decoder_without_bom_handling() |
| 5977 |
.latin1_byte_compatible_up_to(buffer) |
| 5978 |
.unwrap(), |
| 5979 |
4 |
| 5980 |
); |
| 5981 |
assert_eq!( |
| 5982 |
ISO_8859_15 |
| 5983 |
.new_decoder_without_bom_handling() |
| 5984 |
.latin1_byte_compatible_up_to(buffer) |
| 5985 |
.unwrap(), |
| 5986 |
6 |
| 5987 |
); |
| 5988 |
assert_eq!( |
| 5989 |
ISO_8859_16 |
| 5990 |
.new_decoder_without_bom_handling() |
| 5991 |
.latin1_byte_compatible_up_to(buffer) |
| 5992 |
.unwrap(), |
| 5993 |
4 |
| 5994 |
); |
| 5995 |
assert_eq!( |
| 5996 |
ISO_8859_8_I |
| 5997 |
.new_decoder_without_bom_handling() |
| 5998 |
.latin1_byte_compatible_up_to(buffer) |
| 5999 |
.unwrap(), |
| 6000 |
3 |
| 6001 |
); |
| 6002 |
assert_eq!( |
| 6003 |
KOI8_R |
| 6004 |
.new_decoder_without_bom_handling() |
| 6005 |
.latin1_byte_compatible_up_to(buffer) |
| 6006 |
.unwrap(), |
| 6007 |
1 |
| 6008 |
); |
| 6009 |
assert_eq!( |
| 6010 |
KOI8_U |
| 6011 |
.new_decoder_without_bom_handling() |
| 6012 |
.latin1_byte_compatible_up_to(buffer) |
| 6013 |
.unwrap(), |
| 6014 |
1 |
| 6015 |
); |
| 6016 |
assert_eq!( |
| 6017 |
MACINTOSH |
| 6018 |
.new_decoder_without_bom_handling() |
| 6019 |
.latin1_byte_compatible_up_to(buffer) |
| 6020 |
.unwrap(), |
| 6021 |
1 |
| 6022 |
); |
| 6023 |
assert_eq!( |
| 6024 |
WINDOWS_874 |
| 6025 |
.new_decoder_without_bom_handling() |
| 6026 |
.latin1_byte_compatible_up_to(buffer) |
| 6027 |
.unwrap(), |
| 6028 |
2 |
| 6029 |
); |
| 6030 |
assert_eq!( |
| 6031 |
WINDOWS_1250 |
| 6032 |
.new_decoder_without_bom_handling() |
| 6033 |
.latin1_byte_compatible_up_to(buffer) |
| 6034 |
.unwrap(), |
| 6035 |
4 |
| 6036 |
); |
| 6037 |
assert_eq!( |
| 6038 |
WINDOWS_1251 |
| 6039 |
.new_decoder_without_bom_handling() |
| 6040 |
.latin1_byte_compatible_up_to(buffer) |
| 6041 |
.unwrap(), |
| 6042 |
1 |
| 6043 |
); |
| 6044 |
assert_eq!( |
| 6045 |
WINDOWS_1252 |
| 6046 |
.new_decoder_without_bom_handling() |
| 6047 |
.latin1_byte_compatible_up_to(buffer) |
| 6048 |
.unwrap(), |
| 6049 |
5 |
| 6050 |
); |
| 6051 |
assert_eq!( |
| 6052 |
WINDOWS_1253 |
| 6053 |
.new_decoder_without_bom_handling() |
| 6054 |
.latin1_byte_compatible_up_to(buffer) |
| 6055 |
.unwrap(), |
| 6056 |
3 |
| 6057 |
); |
| 6058 |
assert_eq!( |
| 6059 |
WINDOWS_1254 |
| 6060 |
.new_decoder_without_bom_handling() |
| 6061 |
.latin1_byte_compatible_up_to(buffer) |
| 6062 |
.unwrap(), |
| 6063 |
4 |
| 6064 |
); |
| 6065 |
assert_eq!( |
| 6066 |
WINDOWS_1255 |
| 6067 |
.new_decoder_without_bom_handling() |
| 6068 |
.latin1_byte_compatible_up_to(buffer) |
| 6069 |
.unwrap(), |
| 6070 |
3 |
| 6071 |
); |
| 6072 |
assert_eq!( |
| 6073 |
WINDOWS_1256 |
| 6074 |
.new_decoder_without_bom_handling() |
| 6075 |
.latin1_byte_compatible_up_to(buffer) |
| 6076 |
.unwrap(), |
| 6077 |
1 |
| 6078 |
); |
| 6079 |
assert_eq!( |
| 6080 |
WINDOWS_1257 |
| 6081 |
.new_decoder_without_bom_handling() |
| 6082 |
.latin1_byte_compatible_up_to(buffer) |
| 6083 |
.unwrap(), |
| 6084 |
4 |
| 6085 |
); |
| 6086 |
assert_eq!( |
| 6087 |
WINDOWS_1258 |
| 6088 |
.new_decoder_without_bom_handling() |
| 6089 |
.latin1_byte_compatible_up_to(buffer) |
| 6090 |
.unwrap(), |
| 6091 |
4 |
| 6092 |
); |
| 6093 |
assert_eq!( |
| 6094 |
X_MAC_CYRILLIC |
| 6095 |
.new_decoder_without_bom_handling() |
| 6096 |
.latin1_byte_compatible_up_to(buffer) |
| 6097 |
.unwrap(), |
| 6098 |
1 |
| 6099 |
); |
| 6100 |
assert_eq!( |
| 6101 |
X_USER_DEFINED |
| 6102 |
.new_decoder_without_bom_handling() |
| 6103 |
.latin1_byte_compatible_up_to(buffer) |
| 6104 |
.unwrap(), |
| 6105 |
1 |
| 6106 |
); |
| 6107 |
|
| 6108 |
assert!(UTF_8 |
| 6109 |
.new_decoder() |
| 6110 |
.latin1_byte_compatible_up_to(buffer) |
| 6111 |
.is_none()); |
| 6112 |
|
| 6113 |
let mut decoder = UTF_8.new_decoder(); |
| 6114 |
let mut output = [0u16; 4]; |
| 6115 |
let _ = decoder.decode_to_utf16(b"\xEF", &mut output, false); |
| 6116 |
assert!(decoder.latin1_byte_compatible_up_to(buffer).is_none()); |
| 6117 |
let _ = decoder.decode_to_utf16(b"\xBB\xBF", &mut output, false); |
| 6118 |
assert_eq!(decoder.latin1_byte_compatible_up_to(buffer), Some(1)); |
| 6119 |
let _ = decoder.decode_to_utf16(b"\xEF", &mut output, false); |
| 6120 |
assert_eq!(decoder.latin1_byte_compatible_up_to(buffer), None); |
| 6121 |
} |
| 6122 |
} |