1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610
use std::{
cmp::{max, min},
fmt,
io::{self, Write},
str,
};
use crate::{
pool::{Checkout, Pool},
pool_crate::Reset,
};
#[derive(Debug, PartialEq, Eq, Clone)]
pub enum InputElement {
/// length in the stream
Slice(usize),
Splice(usize), // x bytes copied in kernel
}
#[derive(Debug, PartialEq, Eq, Clone)]
pub enum OutputElement {
/// length in the stream
Slice(usize),
Delete(usize),
Insert(Vec<u8>),
Splice(usize), // should copy x bytes from kernel to socket
}
/// The BufferQueue has two roles: holding incoming data, and indicating
/// which data will go out. When new data arrives, it is added at the
/// end of the internal buffer. This new data is then eventually parsed or
/// handled in some way by external code. The external code then adds
/// element to the queue, indicating what to do with the data:
/// - copy a subset of the input data (and advance if needed)
/// - insert external data, like a HTTP header
/// - splice out of the kernel some data that was spliced in
///
/// position is the index in the stream of data already handled.
/// it corresponds to the beginning of available data in the Buffer
/// a Slice(begin, end) would point to buffer.data()[begin-position..end-position]
/// (in the easiest case)
///
/// unparsed_position is the index in the stream of data that was
/// not parsed yet
///
/// The buffer's available data may be smaller than `end - begin`.
/// It can happen if the parser indicated we need to copy more data than is available,
/// like with a content length
///
/// should the buffer queue indicate how much data it needs?
pub struct BufferQueue {
/// position of buffer start in stream
pub buffer_position: usize,
pub parsed_position: usize,
pub start_parsing_position: usize,
pub buffer: Checkout,
/// Vec<(start, length)>
pub input_queue: Vec<InputElement>,
pub output_queue: Vec<OutputElement>,
}
impl BufferQueue {
pub fn with_buffer(buffer: Checkout) -> BufferQueue {
BufferQueue {
buffer_position: 0,
parsed_position: 0,
start_parsing_position: 0,
input_queue: Vec::with_capacity(8),
output_queue: Vec::with_capacity(8),
buffer,
}
}
pub fn invariant(&self) {
debug_assert!(
self.buffer_position <= self.parsed_position,
"buffer_position {} should be smaller than parsed_position {}",
self.buffer_position,
self.parsed_position
);
debug_assert!(
self.parsed_position <= self.start_parsing_position,
"parsed_position {} should be smaller than start_parsing_position {}",
self.parsed_position,
self.start_parsing_position
);
}
pub fn available_input_data(&self) -> usize {
self.input_queue.iter().fold(0, |acc, el| {
acc + match el {
&InputElement::Slice(sz) | &InputElement::Splice(sz) => sz,
}
})
}
pub fn sliced_input(&mut self, count: usize) {
let needed = self.start_parsing_position - self.parsed_position;
if needed > 0 {
if count > needed {
self.parsed_position = self.start_parsing_position;
self.input_queue.push(InputElement::Slice(count - needed));
} else if count <= needed {
self.parsed_position += count;
}
} else if count > 0 {
self.input_queue.push(InputElement::Slice(count));
}
self.invariant();
//println!("sliced_input: buffer size: {}, parsed_position: {} start_parsing_position: {}, input_queue: {:?}, output_queue: {:?}",
// self.buffer.available_data(), self.parsed_position, self.start_parsing_position,
// self.input_queue, self.output_queue);
}
pub fn spliced_input(&mut self, count: usize) {
//FIXME: do the same thing with needed data as in sliced_input
if count > 0 {
self.input_queue.push(InputElement::Splice(count));
}
}
pub fn needs_input(&self) -> bool {
self.start_parsing_position > self.parsed_position
}
pub fn can_restart_parsing(&self) -> bool {
self.start_parsing_position == self.buffer_position
}
pub fn empty(&self) -> bool {
self.input_queue.is_empty() && self.output_queue.is_empty() && self.buffer.empty()
}
pub fn merge_input_slices(&self) -> usize {
let mut acc = 0usize;
for el in self.input_queue.iter() {
match *el {
InputElement::Splice(_) => break,
InputElement::Slice(sz) => acc += sz,
}
}
assert!(
acc <= self.buffer.available_data(),
"the merged input slices can't be larger than current data in buffer"
);
acc
}
// same as available_input_data, TODO: delete one of them?
pub fn input_data_size(&self) -> usize {
let mut acc = 0usize;
for el in self.input_queue.iter() {
match *el {
InputElement::Splice(sz) => acc += sz,
InputElement::Slice(sz) => acc += sz,
}
}
acc
}
pub fn unparsed_data(&self) -> &[u8] {
let largest_size = self.merge_input_slices();
//println!("buffer: {}, parsed: {}", self.buffer_position, self.parsed_position);
let start = self.parsed_position - self.buffer_position;
if largest_size == 0 || start >= self.buffer.available_data() {
return &self.buffer.data()[0..0];
}
//println!("available buffer data: {}, buffer position: {}, parsed_position: {}, start: {}, merged slices size: {}",
// self.buffer.available_data(), self.buffer_position,
//self.parsed_position, start, largest_size);
let end = max(self.buffer.available_data(), start + largest_size);
&self.buffer.data()[start..end]
}
/// should only be called with a count inferior to self.input_data_size()
pub fn consume_parsed_data(&mut self, size: usize) {
//FIXME: to_consume must contain unparsed_position - parsed_position ?
let mut to_consume = size;
while to_consume > 0 {
let new_first_element = match self.input_queue.first() {
None => {
//assert!(to_consume == 0, "no more element in queue, we should not ask to consume {} more bytes", to_consume);
break;
}
Some(&InputElement::Slice(sz)) => {
if to_consume >= sz {
to_consume -= sz;
None
} else {
let new_element = InputElement::Slice(sz - to_consume);
to_consume = 0;
Some(new_element)
}
}
Some(&InputElement::Splice(sz)) => {
if to_consume >= sz {
to_consume -= sz;
None
} else {
panic!("we should not start parsing from inside a splicing buffer. But what if consume_parsed_data was called during a parsing loop? Should only call consume_parsed_data after the parsing loop finished");
}
}
};
match new_first_element {
None => {
self.input_queue.remove(0);
}
Some(el) => {
self.input_queue[0] = el;
}
};
}
self.parsed_position += size - to_consume;
self.start_parsing_position += size;
self.invariant();
}
pub fn slice_output(&mut self, count: usize) {
self.output_queue.push(OutputElement::Slice(count));
}
pub fn delete_output(&mut self, count: usize) {
self.output_queue.push(OutputElement::Delete(count));
}
pub fn splice_output(&mut self, count: usize) {
self.output_queue.push(OutputElement::Splice(count));
}
pub fn insert_output(&mut self, v: Vec<u8>) {
self.output_queue.push(OutputElement::Insert(v));
}
pub fn has_output_data(&self) -> bool {
!self.output_queue.is_empty()
}
pub fn output_data_size(&self) -> usize {
let mut acc = 0usize;
let mut available_buffer_size = self.buffer.available_data();
for el in self.output_queue.iter() {
match *el {
OutputElement::Splice(sz) => acc += sz,
OutputElement::Slice(sz) => {
if available_buffer_size >= sz {
acc += sz;
available_buffer_size -= sz;
} else {
let advance = sz - available_buffer_size;
acc += advance;
return acc;
}
}
OutputElement::Insert(ref v) => acc += v.len(),
OutputElement::Delete(sz) => {
if available_buffer_size >= sz {
available_buffer_size -= sz;
} else {
return acc;
}
}
}
}
acc
}
pub fn merge_output_slices(&self) -> usize {
let mut acc = 0usize;
for el in self.output_queue.iter() {
match *el {
OutputElement::Slice(sz) => acc += sz,
_ => break,
}
}
assert!(
acc <= self.buffer.available_data(),
"the merged output slices can't be larger than current data in buffer"
);
acc
}
pub fn merge_output_deletes(&self) -> usize {
let mut acc = 0usize;
for el in self.output_queue.iter() {
match *el {
OutputElement::Delete(sz) => acc += sz,
_ => break,
}
}
assert!(
acc <= self.buffer.available_data(),
"the merged output deletes can't be larger than current data in buffer"
);
acc
}
pub fn next_output_data(&self) -> &[u8] {
let it = self.output_queue.iter();
//first, calculate how many bytes we need to jump
let mut start = 0usize;
let mut largest_size = 0usize;
let mut delete_ended = false;
//println!("NEXT OUTPUT DATA:\nqueue:\n{:?}\nbuffer:\n{}", self.output_queue, self.buffer.data().to_hex(16));
for el in it {
//println!("start={}, length={}, el = {:?}", start, largest_size, el);
if !delete_ended {
match *el {
OutputElement::Delete(sz) => start += sz,
_ => {
delete_ended = true;
match el {
OutputElement::Slice(sz) => largest_size += *sz,
OutputElement::Insert(vec) => return vec,
_ => break,
}
}
}
} else {
match *el {
OutputElement::Slice(sz) => largest_size += sz,
_ => break,
}
}
}
//println!("buffer data: {:?}", self.buffer.data());
//println!("calculated start={}, length={}", start, largest_size);
//FIXME: should not be larger than the buffer
let length = self.buffer.available_data();
if start > length {
&self.buffer.data()[0..0]
} else {
let end = min(start + largest_size, length);
&self.buffer.data()[start..end]
}
}
pub fn as_ioslice(&self) -> Vec<std::io::IoSlice> {
let mut res = Vec::new();
let it = self.output_queue.iter();
//first, calculate how many bytes we need to jump
let mut start = 0usize;
let length = self.buffer.available_data();
//println!("NEXT OUTPUT DATA:\nqueue:\n{:?}\nbuffer:\n{}", self.output_queue, self.buffer.data().to_hex(16));
for el in it {
match *el {
OutputElement::Delete(sz) => start += sz,
OutputElement::Slice(sz) => {
//println!("Slice({})", sz);
if sz == 0 {
continue;
}
let end = min(start + sz, length);
let i = std::io::IoSlice::new(&self.buffer.data()[start..end]);
//println!("iovec size: {}", i.len());
res.push(i);
start = end;
if end == length {
break;
}
}
OutputElement::Insert(ref v) => {
if v.is_empty() {
continue;
}
let i = std::io::IoSlice::new(&v[..]);
//println!("got Insert with {} bytes", v.len());
res.push(i);
}
OutputElement::Splice(_sz) => {
unimplemented!("splice not used in ioslice")
}
}
}
//println!("returning iovec: {:?}", res);
//println!("returning iovec with {} bytes", complete_size);
res
}
/// should only be called with a count inferior to self.input_data_size()
pub fn consume_output_data(&mut self, size: usize) {
let mut to_consume = size;
while to_consume > 0 {
let new_first_element = match self.output_queue.first() {
None => {
assert!(
to_consume == 0,
"no more element in queue, we should not ask to consume {to_consume} more bytes"
);
break;
}
Some(&OutputElement::Slice(sz)) => {
if to_consume >= sz {
to_consume -= sz;
self.buffer_position += sz;
self.buffer.consume(sz);
None
} else {
let new_element = OutputElement::Slice(sz - to_consume);
self.buffer_position += to_consume;
self.buffer.consume(to_consume);
to_consume = 0;
Some(new_element)
}
}
Some(&OutputElement::Delete(sz)) => {
self.buffer_position += sz;
//FIXME: what if we can't delete that much data?
self.buffer.consume(sz);
None
}
Some(&OutputElement::Splice(sz)) => {
if to_consume >= sz {
to_consume -= sz;
None
} else {
let new_element = OutputElement::Splice(sz - to_consume);
to_consume = 0;
Some(new_element)
}
}
Some(OutputElement::Insert(v)) => {
if to_consume >= v.len() {
to_consume -= v.len();
None
} else {
let new_element = OutputElement::Insert(Vec::from(&v[to_consume..]));
to_consume = 0;
Some(new_element)
}
}
};
match new_first_element {
None => {
self.output_queue.remove(0);
}
Some(el) => {
self.output_queue[0] = el;
}
};
}
self.invariant();
}
pub fn print_unparsed(&self) {
println!("{:?}", str::from_utf8(self.unparsed_data()));
}
pub fn print_and_consume_output(&mut self) {
while self.output_data_size() > 0 {
println!("{:?}", str::from_utf8(self.next_output_data()));
let len = self.next_output_data().len();
self.consume_output_data(len);
}
}
}
impl Write for BufferQueue {
fn write(&mut self, buf: &[u8]) -> io::Result<usize> {
match self.buffer.write(buf) {
Err(e) => Err(e),
Ok(sz) => {
if sz > 0 {
self.input_queue.push(InputElement::Slice(sz));
}
Ok(sz)
}
}
}
fn flush(&mut self) -> io::Result<()> {
Ok(())
}
}
impl Reset for BufferQueue {
fn reset(&mut self) {
self.parsed_position = 0;
self.buffer_position = 0;
self.start_parsing_position = 0;
self.buffer.reset();
self.input_queue.clear();
self.output_queue.clear();
}
}
impl fmt::Debug for BufferQueue {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
//let b: &Buffer = &self.buffer;
write!(f, "BufferQueue {{\nbuffer_position: {},\nparsed_position: {},\nstart_parsing_position: {},\ninput_queue: {:?},\noutput_queue:{:?},\nbuffer: {:?}\n}}",
self.buffer_position, self.parsed_position, self.start_parsing_position,
self.input_queue, self.output_queue, /*b*/ ())
}
}
pub fn buf_with_capacity(capacity: usize) -> (Pool, BufferQueue) {
let mut pool = Pool::with_capacity(1, capacity, 16384);
let b = BufferQueue::with_buffer(pool.checkout().unwrap());
(pool, b)
}
#[cfg(test)]
mod tests {
use super::*;
use crate::sozu_command::buffer::fixed::Buffer;
use nom::HexDisplay;
use std::io::Write;
#[test]
#[cfg(target_pointer_width = "64")]
fn size_test() {
assert_size!(BufferQueue, 88);
assert_size!(Buffer, 16);
}
#[test]
fn consume() {
let (_pool, mut b) = buf_with_capacity(10);
b.buffer.write(&b"ABCDEFGHIJ"[..]).unwrap();
b.buffer.fill(10);
b.input_queue.push(InputElement::Slice(10));
/*let mut b = BufferQueue {
parsed_position: 0,
buffer_position: 0,
start_parsing_position: 0,
buffer: Buffer::from_slice(b"ABCDEFGHIJ"),
input_queue: vec!(InputElement::Slice(10)),
output_queue: vec!()
};*/
// the pool will align the buffer to 16 bytes so there are trailing zeroes
assert_eq!(b.unparsed_data(), &b"ABCDEFGHIJ\0\0\0\0\0\0\0\0\0\0"[..]);
b.consume_parsed_data(4);
assert_eq!(b.parsed_position, 4);
assert_eq!(b.start_parsing_position, 4);
assert_eq!(b.input_queue, vec!(InputElement::Slice(6)));
println!("TEST[{}]", line!());
assert_eq!(b.unparsed_data(), &b"EFGHIJ\0\0\0\0\0\0\0\0\0\0"[..]);
println!("TEST[{}]", line!());
b.slice_output(4);
assert_eq!(b.output_queue, vec!(OutputElement::Slice(4)));
b.insert_output(Vec::from(&b"test"[..]));
assert_eq!(
b.output_queue,
vec!(
OutputElement::Slice(4),
OutputElement::Insert(Vec::from(&b"test"[..]))
)
);
assert_eq!(b.next_output_data(), &b"ABCD"[..]);
println!("before consume: {b:?}");
b.consume_output_data(2);
println!("after consume: {b:?}");
println!("next output data: {}", b.next_output_data().to_hex(8));
assert_eq!(b.next_output_data(), &b"CD"[..]);
println!("TEST[{}]", line!());
b.consume_parsed_data(8);
assert_eq!(b.parsed_position, 10);
assert_eq!(b.start_parsing_position, 12);
assert_eq!(b.input_queue, vec!());
println!("TEST[{}]", line!());
assert_eq!(b.unparsed_data(), &b""[..]);
println!("TEST[{}]", line!());
println!("**test**");
b.consume_output_data(2);
assert_eq!(b.next_output_data(), &b"test"[..]);
b.consume_output_data(2);
assert_eq!(b.next_output_data(), &b"st"[..]);
b.delete_output(2);
b.slice_output(4);
assert_eq!(
b.output_queue,
vec!(
OutputElement::Insert(Vec::from(&b"st"[..])),
OutputElement::Delete(2),
OutputElement::Slice(4)
)
);
b.consume_output_data(2);
assert_eq!(
b.output_queue,
vec!(OutputElement::Delete(2), OutputElement::Slice(4))
);
assert_eq!(b.next_output_data(), &b"GHIJ"[..]);
b.consume_output_data(1);
assert_eq!(b.output_queue, vec!(OutputElement::Slice(3)));
assert_eq!(b.next_output_data(), &b"HIJ"[..]);
b.write(&b"KLMNOP"[..]).unwrap();
}
}