| 1 | use std::io; |
| 2 | use std::io::prelude::*; |
| 3 | |
| 4 | use super::bufread; |
| 5 | use super::{GzBuilder, GzHeader}; |
| 6 | use crate::bufreader::BufReader; |
| 7 | use crate::Compression; |
| 8 | |
| 9 | /// A gzip streaming encoder |
| 10 | /// |
| 11 | /// This structure implements a [`Read`] interface. When read from, it reads |
| 12 | /// uncompressed data from the underlying [`Read`] and provides the compressed data. |
| 13 | /// |
| 14 | /// [`Read`]: https://doc.rust-lang.org/std/io/trait.Read.html |
| 15 | /// |
| 16 | /// # Examples |
| 17 | /// |
| 18 | /// ``` |
| 19 | /// use std::io::prelude::*; |
| 20 | /// use std::io; |
| 21 | /// use flate2::Compression; |
| 22 | /// use flate2::read::GzEncoder; |
| 23 | /// |
| 24 | /// // Return a vector containing the GZ compressed version of hello world |
| 25 | /// |
| 26 | /// fn gzencode_hello_world() -> io::Result<Vec<u8>> { |
| 27 | /// let mut ret_vec = Vec::new(); |
| 28 | /// let bytestring = b"hello world" ; |
| 29 | /// let mut gz = GzEncoder::new(&bytestring[..], Compression::fast()); |
| 30 | /// gz.read_to_end(&mut ret_vec)?; |
| 31 | /// Ok(ret_vec) |
| 32 | /// } |
| 33 | /// ``` |
| 34 | #[derive (Debug)] |
| 35 | pub struct GzEncoder<R> { |
| 36 | inner: bufread::GzEncoder<BufReader<R>>, |
| 37 | } |
| 38 | |
| 39 | pub fn gz_encoder<R: Read>(inner: bufread::GzEncoder<BufReader<R>>) -> GzEncoder<R> { |
| 40 | GzEncoder { inner } |
| 41 | } |
| 42 | |
| 43 | impl<R: Read> GzEncoder<R> { |
| 44 | /// Creates a new encoder which will use the given compression level. |
| 45 | /// |
| 46 | /// The encoder is not configured specially for the emitted header. For |
| 47 | /// header configuration, see the `GzBuilder` type. |
| 48 | /// |
| 49 | /// The data read from the stream `r` will be compressed and available |
| 50 | /// through the returned reader. |
| 51 | pub fn new(r: R, level: Compression) -> GzEncoder<R> { |
| 52 | GzBuilder::new().read(r, lvl:level) |
| 53 | } |
| 54 | } |
| 55 | |
| 56 | impl<R> GzEncoder<R> { |
| 57 | /// Acquires a reference to the underlying reader. |
| 58 | pub fn get_ref(&self) -> &R { |
| 59 | self.inner.get_ref().get_ref() |
| 60 | } |
| 61 | |
| 62 | /// Acquires a mutable reference to the underlying reader. |
| 63 | /// |
| 64 | /// Note that mutation of the reader may result in surprising results if |
| 65 | /// this encoder is continued to be used. |
| 66 | pub fn get_mut(&mut self) -> &mut R { |
| 67 | self.inner.get_mut().get_mut() |
| 68 | } |
| 69 | |
| 70 | /// Returns the underlying stream, consuming this encoder |
| 71 | pub fn into_inner(self) -> R { |
| 72 | self.inner.into_inner().into_inner() |
| 73 | } |
| 74 | } |
| 75 | |
| 76 | impl<R: Read> Read for GzEncoder<R> { |
| 77 | fn read(&mut self, into: &mut [u8]) -> io::Result<usize> { |
| 78 | self.inner.read(buf:into) |
| 79 | } |
| 80 | } |
| 81 | |
| 82 | impl<R: Read + Write> Write for GzEncoder<R> { |
| 83 | fn write(&mut self, buf: &[u8]) -> io::Result<usize> { |
| 84 | self.get_mut().write(buf) |
| 85 | } |
| 86 | |
| 87 | fn flush(&mut self) -> io::Result<()> { |
| 88 | self.get_mut().flush() |
| 89 | } |
| 90 | } |
| 91 | |
| 92 | /// A decoder for a single member of a [gzip file]. |
| 93 | /// |
| 94 | /// This structure implements a [`Read`] interface. When read from, it reads |
| 95 | /// compressed data from the underlying [`Read`] and provides the uncompressed data. |
| 96 | /// |
| 97 | /// After reading a single member of the gzip data this reader will return |
| 98 | /// Ok(0) even if there are more bytes available in the underlying reader. |
| 99 | /// `GzDecoder` may have read additional bytes past the end of the gzip data. |
| 100 | /// If you need the following bytes, wrap the `Reader` in a `std::io::BufReader` |
| 101 | /// and use `bufread::GzDecoder` instead. |
| 102 | /// |
| 103 | /// To handle gzip files that may have multiple members, see [`MultiGzDecoder`] |
| 104 | /// or read more |
| 105 | /// [in the introduction](../index.html#about-multi-member-gzip-files). |
| 106 | /// |
| 107 | /// [gzip file]: https://www.rfc-editor.org/rfc/rfc1952#page-5 |
| 108 | /// |
| 109 | /// # Examples |
| 110 | /// |
| 111 | /// ``` |
| 112 | /// use std::io::prelude::*; |
| 113 | /// use std::io; |
| 114 | /// # use flate2::Compression; |
| 115 | /// # use flate2::write::GzEncoder; |
| 116 | /// use flate2::read::GzDecoder; |
| 117 | /// |
| 118 | /// # fn main() { |
| 119 | /// # let mut e = GzEncoder::new(Vec::new(), Compression::default()); |
| 120 | /// # e.write_all(b"Hello World" ).unwrap(); |
| 121 | /// # let bytes = e.finish().unwrap(); |
| 122 | /// # println!("{}" , decode_reader(bytes).unwrap()); |
| 123 | /// # } |
| 124 | /// # |
| 125 | /// // Uncompresses a Gz Encoded vector of bytes and returns a string or error |
| 126 | /// // Here &[u8] implements Read |
| 127 | /// |
| 128 | /// fn decode_reader(bytes: Vec<u8>) -> io::Result<String> { |
| 129 | /// let mut gz = GzDecoder::new(&bytes[..]); |
| 130 | /// let mut s = String::new(); |
| 131 | /// gz.read_to_string(&mut s)?; |
| 132 | /// Ok(s) |
| 133 | /// } |
| 134 | /// ``` |
| 135 | #[derive (Debug)] |
| 136 | pub struct GzDecoder<R> { |
| 137 | inner: bufread::GzDecoder<BufReader<R>>, |
| 138 | } |
| 139 | |
| 140 | impl<R: Read> GzDecoder<R> { |
| 141 | /// Creates a new decoder from the given reader, immediately parsing the |
| 142 | /// gzip header. |
| 143 | pub fn new(r: R) -> GzDecoder<R> { |
| 144 | GzDecoder { |
| 145 | inner: bufread::GzDecoder::new(BufReader::new(inner:r)), |
| 146 | } |
| 147 | } |
| 148 | } |
| 149 | |
| 150 | impl<R> GzDecoder<R> { |
| 151 | /// Returns the header associated with this stream, if it was valid. |
| 152 | pub fn header(&self) -> Option<&GzHeader> { |
| 153 | self.inner.header() |
| 154 | } |
| 155 | |
| 156 | /// Acquires a reference to the underlying reader. |
| 157 | /// |
| 158 | /// Note that the decoder may have read past the end of the gzip data. |
| 159 | /// To prevent this use [`bufread::GzDecoder`] instead. |
| 160 | pub fn get_ref(&self) -> &R { |
| 161 | self.inner.get_ref().get_ref() |
| 162 | } |
| 163 | |
| 164 | /// Acquires a mutable reference to the underlying stream. |
| 165 | /// |
| 166 | /// Note that mutation of the stream may result in surprising results if |
| 167 | /// this decoder continues to be used. |
| 168 | /// |
| 169 | /// Note that the decoder may have read past the end of the gzip data. |
| 170 | /// To prevent this use [`bufread::GzDecoder`] instead. |
| 171 | pub fn get_mut(&mut self) -> &mut R { |
| 172 | self.inner.get_mut().get_mut() |
| 173 | } |
| 174 | |
| 175 | /// Consumes this decoder, returning the underlying reader. |
| 176 | /// |
| 177 | /// Note that the decoder may have read past the end of the gzip data. |
| 178 | /// Subsequent reads will skip those bytes. To prevent this use |
| 179 | /// [`bufread::GzDecoder`] instead. |
| 180 | pub fn into_inner(self) -> R { |
| 181 | self.inner.into_inner().into_inner() |
| 182 | } |
| 183 | } |
| 184 | |
| 185 | impl<R: Read> Read for GzDecoder<R> { |
| 186 | fn read(&mut self, into: &mut [u8]) -> io::Result<usize> { |
| 187 | self.inner.read(buf:into) |
| 188 | } |
| 189 | } |
| 190 | |
| 191 | impl<R: Read + Write> Write for GzDecoder<R> { |
| 192 | fn write(&mut self, buf: &[u8]) -> io::Result<usize> { |
| 193 | self.get_mut().write(buf) |
| 194 | } |
| 195 | |
| 196 | fn flush(&mut self) -> io::Result<()> { |
| 197 | self.get_mut().flush() |
| 198 | } |
| 199 | } |
| 200 | |
| 201 | /// A gzip streaming decoder that decodes a [gzip file] that may have multiple members. |
| 202 | /// |
| 203 | /// This structure implements a [`Read`] interface. When read from, it reads |
| 204 | /// compressed data from the underlying [`Read`] and provides the uncompressed |
| 205 | /// data. |
| 206 | /// |
| 207 | /// A gzip file consists of a series of *members* concatenated one after another. |
| 208 | /// MultiGzDecoder decodes all members of a file and returns Ok(0) once the |
| 209 | /// underlying reader does. |
| 210 | /// |
| 211 | /// To handle members separately, see [GzDecoder] or read more |
| 212 | /// [in the introduction](../index.html#about-multi-member-gzip-files). |
| 213 | /// |
| 214 | /// [gzip file]: https://www.rfc-editor.org/rfc/rfc1952#page-5 |
| 215 | /// |
| 216 | /// # Examples |
| 217 | /// |
| 218 | /// ``` |
| 219 | /// use std::io::prelude::*; |
| 220 | /// use std::io; |
| 221 | /// # use flate2::Compression; |
| 222 | /// # use flate2::write::GzEncoder; |
| 223 | /// use flate2::read::MultiGzDecoder; |
| 224 | /// |
| 225 | /// # fn main() { |
| 226 | /// # let mut e = GzEncoder::new(Vec::new(), Compression::default()); |
| 227 | /// # e.write_all(b"Hello World" ).unwrap(); |
| 228 | /// # let bytes = e.finish().unwrap(); |
| 229 | /// # println!("{}" , decode_reader(bytes).unwrap()); |
| 230 | /// # } |
| 231 | /// # |
| 232 | /// // Uncompresses a Gz Encoded vector of bytes and returns a string or error |
| 233 | /// // Here &[u8] implements Read |
| 234 | /// |
| 235 | /// fn decode_reader(bytes: Vec<u8>) -> io::Result<String> { |
| 236 | /// let mut gz = MultiGzDecoder::new(&bytes[..]); |
| 237 | /// let mut s = String::new(); |
| 238 | /// gz.read_to_string(&mut s)?; |
| 239 | /// Ok(s) |
| 240 | /// } |
| 241 | /// ``` |
| 242 | #[derive (Debug)] |
| 243 | pub struct MultiGzDecoder<R> { |
| 244 | inner: bufread::MultiGzDecoder<BufReader<R>>, |
| 245 | } |
| 246 | |
| 247 | impl<R: Read> MultiGzDecoder<R> { |
| 248 | /// Creates a new decoder from the given reader, immediately parsing the |
| 249 | /// (first) gzip header. If the gzip stream contains multiple members all will |
| 250 | /// be decoded. |
| 251 | pub fn new(r: R) -> MultiGzDecoder<R> { |
| 252 | MultiGzDecoder { |
| 253 | inner: bufread::MultiGzDecoder::new(BufReader::new(inner:r)), |
| 254 | } |
| 255 | } |
| 256 | } |
| 257 | |
| 258 | impl<R> MultiGzDecoder<R> { |
| 259 | /// Returns the current header associated with this stream, if it's valid. |
| 260 | pub fn header(&self) -> Option<&GzHeader> { |
| 261 | self.inner.header() |
| 262 | } |
| 263 | |
| 264 | /// Acquires a reference to the underlying reader. |
| 265 | pub fn get_ref(&self) -> &R { |
| 266 | self.inner.get_ref().get_ref() |
| 267 | } |
| 268 | |
| 269 | /// Acquires a mutable reference to the underlying stream. |
| 270 | /// |
| 271 | /// Note that mutation of the stream may result in surprising results if |
| 272 | /// this decoder is continued to be used. |
| 273 | pub fn get_mut(&mut self) -> &mut R { |
| 274 | self.inner.get_mut().get_mut() |
| 275 | } |
| 276 | |
| 277 | /// Consumes this decoder, returning the underlying reader. |
| 278 | pub fn into_inner(self) -> R { |
| 279 | self.inner.into_inner().into_inner() |
| 280 | } |
| 281 | } |
| 282 | |
| 283 | impl<R: Read> Read for MultiGzDecoder<R> { |
| 284 | fn read(&mut self, into: &mut [u8]) -> io::Result<usize> { |
| 285 | self.inner.read(buf:into) |
| 286 | } |
| 287 | } |
| 288 | |
| 289 | impl<R: Read + Write> Write for MultiGzDecoder<R> { |
| 290 | fn write(&mut self, buf: &[u8]) -> io::Result<usize> { |
| 291 | self.get_mut().write(buf) |
| 292 | } |
| 293 | |
| 294 | fn flush(&mut self) -> io::Result<()> { |
| 295 | self.get_mut().flush() |
| 296 | } |
| 297 | } |
| 298 | |
| 299 | #[cfg (test)] |
| 300 | mod tests { |
| 301 | use std::io::{Cursor, ErrorKind, Read, Result, Write}; |
| 302 | |
| 303 | use super::GzDecoder; |
| 304 | |
| 305 | //a cursor turning EOF into blocking errors |
| 306 | #[derive (Debug)] |
| 307 | pub struct BlockingCursor { |
| 308 | pub cursor: Cursor<Vec<u8>>, |
| 309 | } |
| 310 | |
| 311 | impl BlockingCursor { |
| 312 | pub fn new() -> BlockingCursor { |
| 313 | BlockingCursor { |
| 314 | cursor: Cursor::new(Vec::new()), |
| 315 | } |
| 316 | } |
| 317 | |
| 318 | pub fn set_position(&mut self, pos: u64) { |
| 319 | return self.cursor.set_position(pos); |
| 320 | } |
| 321 | } |
| 322 | |
| 323 | impl Write for BlockingCursor { |
| 324 | fn write(&mut self, buf: &[u8]) -> Result<usize> { |
| 325 | return self.cursor.write(buf); |
| 326 | } |
| 327 | fn flush(&mut self) -> Result<()> { |
| 328 | return self.cursor.flush(); |
| 329 | } |
| 330 | } |
| 331 | |
| 332 | impl Read for BlockingCursor { |
| 333 | fn read(&mut self, buf: &mut [u8]) -> Result<usize> { |
| 334 | //use the cursor, except it turns eof into blocking error |
| 335 | let r = self.cursor.read(buf); |
| 336 | match r { |
| 337 | Err(ref err) => { |
| 338 | if err.kind() == ErrorKind::UnexpectedEof { |
| 339 | return Err(ErrorKind::WouldBlock.into()); |
| 340 | } |
| 341 | } |
| 342 | Ok(0) => { |
| 343 | //regular EOF turned into blocking error |
| 344 | return Err(ErrorKind::WouldBlock.into()); |
| 345 | } |
| 346 | Ok(_n) => {} |
| 347 | } |
| 348 | return r; |
| 349 | } |
| 350 | } |
| 351 | |
| 352 | #[test ] |
| 353 | fn blocked_partial_header_read() { |
| 354 | // this is a reader which receives data afterwards |
| 355 | let mut r = BlockingCursor::new(); |
| 356 | let data = vec![1, 2, 3]; |
| 357 | |
| 358 | match r.write_all(&data) { |
| 359 | Ok(()) => {} |
| 360 | _ => { |
| 361 | panic!("Unexpected result for write_all" ); |
| 362 | } |
| 363 | } |
| 364 | r.set_position(0); |
| 365 | |
| 366 | // this is unused except for the buffering |
| 367 | let mut decoder = GzDecoder::new(r); |
| 368 | let mut out = Vec::with_capacity(7); |
| 369 | match decoder.read(&mut out) { |
| 370 | Err(e) => { |
| 371 | assert_eq!(e.kind(), ErrorKind::WouldBlock); |
| 372 | } |
| 373 | _ => { |
| 374 | panic!("Unexpected result for decoder.read" ); |
| 375 | } |
| 376 | } |
| 377 | } |
| 378 | } |
| 379 | |