| 1 |
use std::cmp; |
| 2 |
use std::io; |
| 3 |
use std::io::prelude::*; |
| 4 |
|
| 5 |
use super::{corrupt, GzBuilder, GzHeader, GzHeaderParser}; |
| 6 |
use crate::crc::{Crc, CrcWriter}; |
| 7 |
use crate::zio; |
| 8 |
use crate::{Compress, Compression, Decompress, Status}; |
| 9 |
|
| 10 |
/// A gzip streaming encoder |
| 11 |
/// |
| 12 |
/// This structure exposes a [`Write`] interface that will emit compressed data |
| 13 |
/// to the underlying writer `W`. |
| 14 |
/// |
| 15 |
/// [`Write`]: https://doc.rust-lang.org/std/io/trait.Write.html |
| 16 |
/// |
| 17 |
/// # Examples |
| 18 |
/// |
| 19 |
/// ``` |
| 20 |
/// use std::io::prelude::*; |
| 21 |
/// use flate2::Compression; |
| 22 |
/// use flate2::write::GzEncoder; |
| 23 |
/// |
| 24 |
/// // Vec<u8> implements Write to print the compressed bytes of sample string |
| 25 |
/// # fn main() { |
| 26 |
/// |
| 27 |
/// let mut e = GzEncoder::new(Vec::new(), Compression::default()); |
| 28 |
/// e.write_all(b"Hello World").unwrap(); |
| 29 |
/// println!("{:?}", e.finish().unwrap()); |
| 30 |
/// # } |
| 31 |
/// ``` |
| 32 |
#[derive(Debug)] |
| 33 |
pub struct GzEncoder<W: Write> { |
| 34 |
inner: zio::Writer<W, Compress>, |
| 35 |
crc: Crc, |
| 36 |
crc_bytes_written: usize, |
| 37 |
header: Vec<u8>, |
| 38 |
} |
| 39 |
|
| 40 |
pub fn gz_encoder<W: Write>(header: Vec<u8>, w: W, lvl: Compression) -> GzEncoder<W> { |
| 41 |
GzEncoder { |
| 42 |
inner: zio::Writer::new(w, Compress::new(lvl, false)), |
| 43 |
crc: Crc::new(), |
| 44 |
header, |
| 45 |
crc_bytes_written: 0, |
| 46 |
} |
| 47 |
} |
| 48 |
|
| 49 |
impl<W: Write> GzEncoder<W> { |
| 50 |
/// Creates a new encoder which will use the given compression level. |
| 51 |
/// |
| 52 |
/// The encoder is not configured specially for the emitted header. For |
| 53 |
/// header configuration, see the `GzBuilder` type. |
| 54 |
/// |
| 55 |
/// The data written to the returned encoder will be compressed and then |
| 56 |
/// written to the stream `w`. |
| 57 |
pub fn new(w: W, level: Compression) -> GzEncoder<W> { |
| 58 |
GzBuilder::new().write(w, level) |
| 59 |
} |
| 60 |
|
| 61 |
/// Acquires a reference to the underlying writer. |
| 62 |
pub fn get_ref(&self) -> &W { |
| 63 |
self.inner.get_ref() |
| 64 |
} |
| 65 |
|
| 66 |
/// Acquires a mutable reference to the underlying writer. |
| 67 |
/// |
| 68 |
/// Note that mutation of the writer may result in surprising results if |
| 69 |
/// this encoder is continued to be used. |
| 70 |
pub fn get_mut(&mut self) -> &mut W { |
| 71 |
self.inner.get_mut() |
| 72 |
} |
| 73 |
|
| 74 |
/// Attempt to finish this output stream, writing out final chunks of data. |
| 75 |
/// |
| 76 |
/// Note that this function can only be used once data has finished being |
| 77 |
/// written to the output stream. After this function is called then further |
| 78 |
/// calls to `write` may result in a panic. |
| 79 |
/// |
| 80 |
/// # Panics |
| 81 |
/// |
| 82 |
/// Attempts to write data to this stream may result in a panic after this |
| 83 |
/// function is called. |
| 84 |
/// |
| 85 |
/// # Errors |
| 86 |
/// |
| 87 |
/// This function will perform I/O to complete this stream, and any I/O |
| 88 |
/// errors which occur will be returned from this function. |
| 89 |
pub fn try_finish(&mut self) -> io::Result<()> { |
| 90 |
self.write_header()?; |
| 91 |
self.inner.finish()?; |
| 92 |
|
| 93 |
while self.crc_bytes_written < 8 { |
| 94 |
let (sum, amt) = (self.crc.sum(), self.crc.amount()); |
| 95 |
let buf = [ |
| 96 |
sum as u8, |
| 97 |
(sum >> 8) as u8, |
| 98 |
(sum >> 16) as u8, |
| 99 |
(sum >> 24) as u8, |
| 100 |
amt as u8, |
| 101 |
(amt >> 8) as u8, |
| 102 |
(amt >> 16) as u8, |
| 103 |
(amt >> 24) as u8, |
| 104 |
]; |
| 105 |
let inner = self.inner.get_mut(); |
| 106 |
let n = inner.write(&buf[self.crc_bytes_written..])?; |
| 107 |
self.crc_bytes_written += n; |
| 108 |
} |
| 109 |
Ok(()) |
| 110 |
} |
| 111 |
|
| 112 |
/// Finish encoding this stream, returning the underlying writer once the |
| 113 |
/// encoding is done. |
| 114 |
/// |
| 115 |
/// Note that this function may not be suitable to call in a situation where |
| 116 |
/// the underlying stream is an asynchronous I/O stream. To finish a stream |
| 117 |
/// the `try_finish` (or `shutdown`) method should be used instead. To |
| 118 |
/// re-acquire ownership of a stream it is safe to call this method after |
| 119 |
/// `try_finish` or `shutdown` has returned `Ok`. |
| 120 |
/// |
| 121 |
/// # Errors |
| 122 |
/// |
| 123 |
/// This function will perform I/O to complete this stream, and any I/O |
| 124 |
/// errors which occur will be returned from this function. |
| 125 |
pub fn finish(mut self) -> io::Result<W> { |
| 126 |
self.try_finish()?; |
| 127 |
Ok(self.inner.take_inner()) |
| 128 |
} |
| 129 |
|
| 130 |
fn write_header(&mut self) -> io::Result<()> { |
| 131 |
while !self.header.is_empty() { |
| 132 |
let n = self.inner.get_mut().write(&self.header)?; |
| 133 |
self.header.drain(..n); |
| 134 |
} |
| 135 |
Ok(()) |
| 136 |
} |
| 137 |
} |
| 138 |
|
| 139 |
impl<W: Write> Write for GzEncoder<W> { |
| 140 |
fn write(&mut self, buf: &[u8]) -> io::Result<usize> { |
| 141 |
assert_eq!(self.crc_bytes_written, 0); |
| 142 |
self.write_header()?; |
| 143 |
let n = self.inner.write(buf)?; |
| 144 |
self.crc.update(&buf[..n]); |
| 145 |
Ok(n) |
| 146 |
} |
| 147 |
|
| 148 |
fn flush(&mut self) -> io::Result<()> { |
| 149 |
assert_eq!(self.crc_bytes_written, 0); |
| 150 |
self.write_header()?; |
| 151 |
self.inner.flush() |
| 152 |
} |
| 153 |
} |
| 154 |
|
| 155 |
impl<R: Read + Write> Read for GzEncoder<R> { |
| 156 |
fn read(&mut self, buf: &mut [u8]) -> io::Result<usize> { |
| 157 |
self.get_mut().read(buf) |
| 158 |
} |
| 159 |
} |
| 160 |
|
| 161 |
impl<W: Write> Drop for GzEncoder<W> { |
| 162 |
fn drop(&mut self) { |
| 163 |
if self.inner.is_present() { |
| 164 |
let _ = self.try_finish(); |
| 165 |
} |
| 166 |
} |
| 167 |
} |
| 168 |
|
| 169 |
/// A decoder for a single member of a [gzip file]. |
| 170 |
/// |
| 171 |
/// This structure exposes a [`Write`] interface, receiving compressed data and |
| 172 |
/// writing uncompressed data to the underlying writer. |
| 173 |
/// |
| 174 |
/// After decoding a single member of the gzip data this writer will return the number of bytes up to |
| 175 |
/// to the end of the gzip member and subsequent writes will return Ok(0) allowing the caller to |
| 176 |
/// handle any data following the gzip member. |
| 177 |
/// |
| 178 |
/// To handle gzip files that may have multiple members, see [`MultiGzDecoder`] |
| 179 |
/// or read more |
| 180 |
/// [in the introduction](../index.html#about-multi-member-gzip-files). |
| 181 |
/// |
| 182 |
/// [gzip file]: https://www.rfc-editor.org/rfc/rfc1952#page-5 |
| 183 |
/// [`Write`]: https://doc.rust-lang.org/std/io/trait.Write.html |
| 184 |
/// |
| 185 |
/// # Examples |
| 186 |
/// |
| 187 |
/// ``` |
| 188 |
/// use std::io::prelude::*; |
| 189 |
/// use std::io; |
| 190 |
/// use flate2::Compression; |
| 191 |
/// use flate2::write::{GzEncoder, GzDecoder}; |
| 192 |
/// |
| 193 |
/// # fn main() { |
| 194 |
/// # let mut e = GzEncoder::new(Vec::new(), Compression::default()); |
| 195 |
/// # e.write(b"Hello World").unwrap(); |
| 196 |
/// # let bytes = e.finish().unwrap(); |
| 197 |
/// # assert_eq!("Hello World", decode_writer(bytes).unwrap()); |
| 198 |
/// # } |
| 199 |
/// // Uncompresses a gzip encoded vector of bytes and returns a string or error |
| 200 |
/// // Here Vec<u8> implements Write |
| 201 |
/// fn decode_writer(bytes: Vec<u8>) -> io::Result<String> { |
| 202 |
/// let mut writer = Vec::new(); |
| 203 |
/// let mut decoder = GzDecoder::new(writer); |
| 204 |
/// decoder.write_all(&bytes[..])?; |
| 205 |
/// writer = decoder.finish()?; |
| 206 |
/// let return_string = String::from_utf8(writer).expect("String parsing error"); |
| 207 |
/// Ok(return_string) |
| 208 |
/// } |
| 209 |
/// ``` |
| 210 |
#[derive(Debug)] |
| 211 |
pub struct GzDecoder<W: Write> { |
| 212 |
inner: zio::Writer<CrcWriter<W>, Decompress>, |
| 213 |
crc_bytes: Vec<u8>, |
| 214 |
header_parser: GzHeaderParser, |
| 215 |
} |
| 216 |
|
| 217 |
const CRC_BYTES_LEN: usize = 8; |
| 218 |
|
| 219 |
impl<W: Write> GzDecoder<W> { |
| 220 |
/// Creates a new decoder which will write uncompressed data to the stream. |
| 221 |
/// |
| 222 |
/// When this encoder is dropped or unwrapped the final pieces of data will |
| 223 |
/// be flushed. |
| 224 |
pub fn new(w: W) -> GzDecoder<W> { |
| 225 |
GzDecoder { |
| 226 |
inner: zio::Writer::new(CrcWriter::new(w), Decompress::new(false)), |
| 227 |
crc_bytes: Vec::with_capacity(CRC_BYTES_LEN), |
| 228 |
header_parser: GzHeaderParser::new(), |
| 229 |
} |
| 230 |
} |
| 231 |
|
| 232 |
/// Returns the header associated with this stream. |
| 233 |
pub fn header(&self) -> Option<&GzHeader> { |
| 234 |
self.header_parser.header() |
| 235 |
} |
| 236 |
|
| 237 |
/// Acquires a reference to the underlying writer. |
| 238 |
pub fn get_ref(&self) -> &W { |
| 239 |
self.inner.get_ref().get_ref() |
| 240 |
} |
| 241 |
|
| 242 |
/// Acquires a mutable reference to the underlying writer. |
| 243 |
/// |
| 244 |
/// Note that mutating the output/input state of the stream may corrupt this |
| 245 |
/// object, so care must be taken when using this method. |
| 246 |
pub fn get_mut(&mut self) -> &mut W { |
| 247 |
self.inner.get_mut().get_mut() |
| 248 |
} |
| 249 |
|
| 250 |
/// Attempt to finish this output stream, writing out final chunks of data. |
| 251 |
/// |
| 252 |
/// Note that this function can only be used once data has finished being |
| 253 |
/// written to the output stream. After this function is called then further |
| 254 |
/// calls to `write` may result in a panic. |
| 255 |
/// |
| 256 |
/// # Panics |
| 257 |
/// |
| 258 |
/// Attempts to write data to this stream may result in a panic after this |
| 259 |
/// function is called. |
| 260 |
/// |
| 261 |
/// # Errors |
| 262 |
/// |
| 263 |
/// This function will perform I/O to finish the stream, returning any |
| 264 |
/// errors which happen. |
| 265 |
pub fn try_finish(&mut self) -> io::Result<()> { |
| 266 |
self.finish_and_check_crc()?; |
| 267 |
Ok(()) |
| 268 |
} |
| 269 |
|
| 270 |
/// Consumes this decoder, flushing the output stream. |
| 271 |
/// |
| 272 |
/// This will flush the underlying data stream and then return the contained |
| 273 |
/// writer if the flush succeeded. |
| 274 |
/// |
| 275 |
/// Note that this function may not be suitable to call in a situation where |
| 276 |
/// the underlying stream is an asynchronous I/O stream. To finish a stream |
| 277 |
/// the `try_finish` (or `shutdown`) method should be used instead. To |
| 278 |
/// re-acquire ownership of a stream it is safe to call this method after |
| 279 |
/// `try_finish` or `shutdown` has returned `Ok`. |
| 280 |
/// |
| 281 |
/// # Errors |
| 282 |
/// |
| 283 |
/// This function will perform I/O to complete this stream, and any I/O |
| 284 |
/// errors which occur will be returned from this function. |
| 285 |
pub fn finish(mut self) -> io::Result<W> { |
| 286 |
self.finish_and_check_crc()?; |
| 287 |
Ok(self.inner.take_inner().into_inner()) |
| 288 |
} |
| 289 |
|
| 290 |
fn finish_and_check_crc(&mut self) -> io::Result<()> { |
| 291 |
self.inner.finish()?; |
| 292 |
|
| 293 |
if self.crc_bytes.len() != 8 { |
| 294 |
return Err(corrupt()); |
| 295 |
} |
| 296 |
|
| 297 |
let crc = (self.crc_bytes[0] as u32) |
| 298 |
| ((self.crc_bytes[1] as u32) << 8) |
| 299 |
| ((self.crc_bytes[2] as u32) << 16) |
| 300 |
| ((self.crc_bytes[3] as u32) << 24); |
| 301 |
let amt = (self.crc_bytes[4] as u32) |
| 302 |
| ((self.crc_bytes[5] as u32) << 8) |
| 303 |
| ((self.crc_bytes[6] as u32) << 16) |
| 304 |
| ((self.crc_bytes[7] as u32) << 24); |
| 305 |
if crc != self.inner.get_ref().crc().sum() { |
| 306 |
return Err(corrupt()); |
| 307 |
} |
| 308 |
if amt != self.inner.get_ref().crc().amount() { |
| 309 |
return Err(corrupt()); |
| 310 |
} |
| 311 |
Ok(()) |
| 312 |
} |
| 313 |
} |
| 314 |
|
| 315 |
impl<W: Write> Write for GzDecoder<W> { |
| 316 |
fn write(&mut self, mut buf: &[u8]) -> io::Result<usize> { |
| 317 |
let buflen = buf.len(); |
| 318 |
if self.header().is_none() { |
| 319 |
match self.header_parser.parse(&mut buf) { |
| 320 |
Err(err) => { |
| 321 |
if err.kind() == io::ErrorKind::UnexpectedEof { |
| 322 |
// all data read but header still not complete |
| 323 |
Ok(buflen) |
| 324 |
} else { |
| 325 |
Err(err) |
| 326 |
} |
| 327 |
} |
| 328 |
Ok(_) => { |
| 329 |
debug_assert!(self.header().is_some()); |
| 330 |
// buf now contains the unread part of the original buf |
| 331 |
let n = buflen - buf.len(); |
| 332 |
Ok(n) |
| 333 |
} |
| 334 |
} |
| 335 |
} else { |
| 336 |
let (n, status) = self.inner.write_with_status(buf)?; |
| 337 |
|
| 338 |
if status == Status::StreamEnd && n < buf.len() && self.crc_bytes.len() < 8 { |
| 339 |
let remaining = buf.len() - n; |
| 340 |
let crc_bytes = cmp::min(remaining, CRC_BYTES_LEN - self.crc_bytes.len()); |
| 341 |
self.crc_bytes.extend(&buf[n..n + crc_bytes]); |
| 342 |
return Ok(n + crc_bytes); |
| 343 |
} |
| 344 |
Ok(n) |
| 345 |
} |
| 346 |
} |
| 347 |
|
| 348 |
fn flush(&mut self) -> io::Result<()> { |
| 349 |
self.inner.flush() |
| 350 |
} |
| 351 |
} |
| 352 |
|
| 353 |
impl<W: Read + Write> Read for GzDecoder<W> { |
| 354 |
fn read(&mut self, buf: &mut [u8]) -> io::Result<usize> { |
| 355 |
self.inner.get_mut().get_mut().read(buf) |
| 356 |
} |
| 357 |
} |
| 358 |
|
| 359 |
/// A gzip streaming decoder that decodes a [gzip file] with multiple members. |
| 360 |
/// |
| 361 |
/// This structure exposes a [`Write`] interface that will consume compressed data and |
| 362 |
/// write uncompressed data to the underlying writer. |
| 363 |
/// |
| 364 |
/// A gzip file consists of a series of *members* concatenated one after another. |
| 365 |
/// `MultiGzDecoder` decodes all members of a file and writes them to the |
| 366 |
/// underlying writer one after another. |
| 367 |
/// |
| 368 |
/// To handle members separately, see [GzDecoder] or read more |
| 369 |
/// [in the introduction](../index.html#about-multi-member-gzip-files). |
| 370 |
/// |
| 371 |
/// [gzip file]: https://www.rfc-editor.org/rfc/rfc1952#page-5 |
| 372 |
#[derive(Debug)] |
| 373 |
pub struct MultiGzDecoder<W: Write> { |
| 374 |
inner: GzDecoder<W>, |
| 375 |
} |
| 376 |
|
| 377 |
impl<W: Write> MultiGzDecoder<W> { |
| 378 |
/// Creates a new decoder which will write uncompressed data to the stream. |
| 379 |
/// If the gzip stream contains multiple members all will be decoded. |
| 380 |
pub fn new(w: W) -> MultiGzDecoder<W> { |
| 381 |
MultiGzDecoder { |
| 382 |
inner: GzDecoder::new(w), |
| 383 |
} |
| 384 |
} |
| 385 |
|
| 386 |
/// Returns the header associated with the current member. |
| 387 |
pub fn header(&self) -> Option<&GzHeader> { |
| 388 |
self.inner.header() |
| 389 |
} |
| 390 |
|
| 391 |
/// Acquires a reference to the underlying writer. |
| 392 |
pub fn get_ref(&self) -> &W { |
| 393 |
self.inner.get_ref() |
| 394 |
} |
| 395 |
|
| 396 |
/// Acquires a mutable reference to the underlying writer. |
| 397 |
/// |
| 398 |
/// Note that mutating the output/input state of the stream may corrupt this |
| 399 |
/// object, so care must be taken when using this method. |
| 400 |
pub fn get_mut(&mut self) -> &mut W { |
| 401 |
self.inner.get_mut() |
| 402 |
} |
| 403 |
|
| 404 |
/// Attempt to finish this output stream, writing out final chunks of data. |
| 405 |
/// |
| 406 |
/// Note that this function can only be used once data has finished being |
| 407 |
/// written to the output stream. After this function is called then further |
| 408 |
/// calls to `write` may result in a panic. |
| 409 |
/// |
| 410 |
/// # Panics |
| 411 |
/// |
| 412 |
/// Attempts to write data to this stream may result in a panic after this |
| 413 |
/// function is called. |
| 414 |
/// |
| 415 |
/// # Errors |
| 416 |
/// |
| 417 |
/// This function will perform I/O to finish the stream, returning any |
| 418 |
/// errors which happen. |
| 419 |
pub fn try_finish(&mut self) -> io::Result<()> { |
| 420 |
self.inner.try_finish() |
| 421 |
} |
| 422 |
|
| 423 |
/// Consumes this decoder, flushing the output stream. |
| 424 |
/// |
| 425 |
/// This will flush the underlying data stream and then return the contained |
| 426 |
/// writer if the flush succeeded. |
| 427 |
/// |
| 428 |
/// Note that this function may not be suitable to call in a situation where |
| 429 |
/// the underlying stream is an asynchronous I/O stream. To finish a stream |
| 430 |
/// the `try_finish` (or `shutdown`) method should be used instead. To |
| 431 |
/// re-acquire ownership of a stream it is safe to call this method after |
| 432 |
/// `try_finish` or `shutdown` has returned `Ok`. |
| 433 |
/// |
| 434 |
/// # Errors |
| 435 |
/// |
| 436 |
/// This function will perform I/O to complete this stream, and any I/O |
| 437 |
/// errors which occur will be returned from this function. |
| 438 |
pub fn finish(self) -> io::Result<W> { |
| 439 |
self.inner.finish() |
| 440 |
} |
| 441 |
} |
| 442 |
|
| 443 |
impl<W: Write> Write for MultiGzDecoder<W> { |
| 444 |
fn write(&mut self, buf: &[u8]) -> io::Result<usize> { |
| 445 |
if buf.is_empty() { |
| 446 |
Ok(0) |
| 447 |
} else { |
| 448 |
match self.inner.write(buf) { |
| 449 |
Ok(0) => { |
| 450 |
// When the GzDecoder indicates that it has finished |
| 451 |
// create a new GzDecoder to handle additional data. |
| 452 |
self.inner.try_finish()?; |
| 453 |
let w = self.inner.inner.take_inner().into_inner(); |
| 454 |
self.inner = GzDecoder::new(w); |
| 455 |
self.inner.write(buf) |
| 456 |
} |
| 457 |
res => res, |
| 458 |
} |
| 459 |
} |
| 460 |
} |
| 461 |
|
| 462 |
fn flush(&mut self) -> io::Result<()> { |
| 463 |
self.inner.flush() |
| 464 |
} |
| 465 |
} |
| 466 |
|
| 467 |
#[cfg(test)] |
| 468 |
mod tests { |
| 469 |
use super::*; |
| 470 |
|
| 471 |
const STR: &str = "Hello World Hello World Hello World Hello World Hello World \ |
| 472 |
Hello World Hello World Hello World Hello World Hello World \ |
| 473 |
Hello World Hello World Hello World Hello World Hello World \ |
| 474 |
Hello World Hello World Hello World Hello World Hello World \ |
| 475 |
Hello World Hello World Hello World Hello World Hello World"; |
| 476 |
|
| 477 |
#[test] |
| 478 |
fn decode_writer_one_chunk() { |
| 479 |
let mut e = GzEncoder::new(Vec::new(), Compression::default()); |
| 480 |
e.write_all(STR.as_ref()).unwrap(); |
| 481 |
let bytes = e.finish().unwrap(); |
| 482 |
|
| 483 |
let mut writer = Vec::new(); |
| 484 |
let mut decoder = GzDecoder::new(writer); |
| 485 |
let n = decoder.write(&bytes[..]).unwrap(); |
| 486 |
decoder.write_all(&bytes[n..]).unwrap(); |
| 487 |
decoder.try_finish().unwrap(); |
| 488 |
writer = decoder.finish().unwrap(); |
| 489 |
let return_string = String::from_utf8(writer).expect("String parsing error"); |
| 490 |
assert_eq!(return_string, STR); |
| 491 |
} |
| 492 |
|
| 493 |
#[test] |
| 494 |
fn decode_writer_partial_header() { |
| 495 |
let mut e = GzEncoder::new(Vec::new(), Compression::default()); |
| 496 |
e.write_all(STR.as_ref()).unwrap(); |
| 497 |
let bytes = e.finish().unwrap(); |
| 498 |
|
| 499 |
let mut writer = Vec::new(); |
| 500 |
let mut decoder = GzDecoder::new(writer); |
| 501 |
assert_eq!(decoder.write(&bytes[..5]).unwrap(), 5); |
| 502 |
let n = decoder.write(&bytes[5..]).unwrap(); |
| 503 |
if n < bytes.len() - 5 { |
| 504 |
decoder.write_all(&bytes[n + 5..]).unwrap(); |
| 505 |
} |
| 506 |
writer = decoder.finish().unwrap(); |
| 507 |
let return_string = String::from_utf8(writer).expect("String parsing error"); |
| 508 |
assert_eq!(return_string, STR); |
| 509 |
} |
| 510 |
|
| 511 |
#[test] |
| 512 |
fn decode_writer_partial_header_filename() { |
| 513 |
let filename = "test.txt"; |
| 514 |
let mut e = GzBuilder::new() |
| 515 |
.filename(filename) |
| 516 |
.read(STR.as_bytes(), Compression::default()); |
| 517 |
let mut bytes = Vec::new(); |
| 518 |
e.read_to_end(&mut bytes).unwrap(); |
| 519 |
|
| 520 |
let mut writer = Vec::new(); |
| 521 |
let mut decoder = GzDecoder::new(writer); |
| 522 |
assert_eq!(decoder.write(&bytes[..12]).unwrap(), 12); |
| 523 |
let n = decoder.write(&bytes[12..]).unwrap(); |
| 524 |
if n < bytes.len() - 12 { |
| 525 |
decoder.write_all(&bytes[n + 12..]).unwrap(); |
| 526 |
} |
| 527 |
assert_eq!( |
| 528 |
decoder.header().unwrap().filename().unwrap(), |
| 529 |
filename.as_bytes() |
| 530 |
); |
| 531 |
writer = decoder.finish().unwrap(); |
| 532 |
let return_string = String::from_utf8(writer).expect("String parsing error"); |
| 533 |
assert_eq!(return_string, STR); |
| 534 |
} |
| 535 |
|
| 536 |
#[test] |
| 537 |
fn decode_writer_partial_header_comment() { |
| 538 |
let comment = "test comment"; |
| 539 |
let mut e = GzBuilder::new() |
| 540 |
.comment(comment) |
| 541 |
.read(STR.as_bytes(), Compression::default()); |
| 542 |
let mut bytes = Vec::new(); |
| 543 |
e.read_to_end(&mut bytes).unwrap(); |
| 544 |
|
| 545 |
let mut writer = Vec::new(); |
| 546 |
let mut decoder = GzDecoder::new(writer); |
| 547 |
assert_eq!(decoder.write(&bytes[..12]).unwrap(), 12); |
| 548 |
let n = decoder.write(&bytes[12..]).unwrap(); |
| 549 |
if n < bytes.len() - 12 { |
| 550 |
decoder.write_all(&bytes[n + 12..]).unwrap(); |
| 551 |
} |
| 552 |
assert_eq!( |
| 553 |
decoder.header().unwrap().comment().unwrap(), |
| 554 |
comment.as_bytes() |
| 555 |
); |
| 556 |
writer = decoder.finish().unwrap(); |
| 557 |
let return_string = String::from_utf8(writer).expect("String parsing error"); |
| 558 |
assert_eq!(return_string, STR); |
| 559 |
} |
| 560 |
|
| 561 |
#[test] |
| 562 |
fn decode_writer_exact_header() { |
| 563 |
let mut e = GzEncoder::new(Vec::new(), Compression::default()); |
| 564 |
e.write_all(STR.as_ref()).unwrap(); |
| 565 |
let bytes = e.finish().unwrap(); |
| 566 |
|
| 567 |
let mut writer = Vec::new(); |
| 568 |
let mut decoder = GzDecoder::new(writer); |
| 569 |
assert_eq!(decoder.write(&bytes[..10]).unwrap(), 10); |
| 570 |
decoder.write_all(&bytes[10..]).unwrap(); |
| 571 |
writer = decoder.finish().unwrap(); |
| 572 |
let return_string = String::from_utf8(writer).expect("String parsing error"); |
| 573 |
assert_eq!(return_string, STR); |
| 574 |
} |
| 575 |
|
| 576 |
#[test] |
| 577 |
fn decode_writer_partial_crc() { |
| 578 |
let mut e = GzEncoder::new(Vec::new(), Compression::default()); |
| 579 |
e.write_all(STR.as_ref()).unwrap(); |
| 580 |
let bytes = e.finish().unwrap(); |
| 581 |
|
| 582 |
let mut writer = Vec::new(); |
| 583 |
let mut decoder = GzDecoder::new(writer); |
| 584 |
let l = bytes.len() - 5; |
| 585 |
let n = decoder.write(&bytes[..l]).unwrap(); |
| 586 |
decoder.write_all(&bytes[n..]).unwrap(); |
| 587 |
writer = decoder.finish().unwrap(); |
| 588 |
let return_string = String::from_utf8(writer).expect("String parsing error"); |
| 589 |
assert_eq!(return_string, STR); |
| 590 |
} |
| 591 |
|
| 592 |
// Two or more gzip files concatenated form a multi-member gzip file. MultiGzDecoder will |
| 593 |
// concatenate the decoded contents of all members. |
| 594 |
#[test] |
| 595 |
fn decode_multi_writer() { |
| 596 |
let mut e = GzEncoder::new(Vec::new(), Compression::default()); |
| 597 |
e.write_all(STR.as_ref()).unwrap(); |
| 598 |
let bytes = e.finish().unwrap().repeat(2); |
| 599 |
|
| 600 |
let mut writer = Vec::new(); |
| 601 |
let mut decoder = MultiGzDecoder::new(writer); |
| 602 |
let mut count = 0; |
| 603 |
while count < bytes.len() { |
| 604 |
let n = decoder.write(&bytes[count..]).unwrap(); |
| 605 |
assert!(n != 0); |
| 606 |
count += n; |
| 607 |
} |
| 608 |
writer = decoder.finish().unwrap(); |
| 609 |
let return_string = String::from_utf8(writer).expect("String parsing error"); |
| 610 |
let expected = STR.repeat(2); |
| 611 |
assert_eq!(return_string, expected); |
| 612 |
} |
| 613 |
|
| 614 |
// GzDecoder consumes one gzip member and then returns 0 for subsequent writes, allowing any |
| 615 |
// additional data to be consumed by the caller. |
| 616 |
#[test] |
| 617 |
fn decode_extra_data() { |
| 618 |
let compressed = { |
| 619 |
let mut e = GzEncoder::new(Vec::new(), Compression::default()); |
| 620 |
e.write_all(STR.as_ref()).unwrap(); |
| 621 |
let mut b = e.finish().unwrap(); |
| 622 |
b.push(b'x'); |
| 623 |
b |
| 624 |
}; |
| 625 |
|
| 626 |
let mut writer = Vec::new(); |
| 627 |
let mut decoder = GzDecoder::new(writer); |
| 628 |
let mut consumed_bytes = 0; |
| 629 |
loop { |
| 630 |
let n = decoder.write(&compressed[consumed_bytes..]).unwrap(); |
| 631 |
if n == 0 { |
| 632 |
break; |
| 633 |
} |
| 634 |
consumed_bytes += n; |
| 635 |
} |
| 636 |
writer = decoder.finish().unwrap(); |
| 637 |
let actual = String::from_utf8(writer).expect("String parsing error"); |
| 638 |
assert_eq!(actual, STR); |
| 639 |
assert_eq!(&compressed[consumed_bytes..], b"x"); |
| 640 |
} |
| 641 |
} |