talksik/uWebSockets
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Greatly improved chunk-encoded parser and test

Browse Source
This commit is contained in:
Alex Hultman
2022-10-08 04:32:59 +02:00
parent 5062696f63
commit e4e04940c5
2 changed files with 83 additions and 37 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/ChunkedEncoding.h
+69 -16
View File
@@ -52,7 +52,6 @@ namespace uWS {
}
/* Now we stand on \n so consume it and enable size */
if (data.length()) {
//printf("size of chunk is %d\n", state);
state += 2; // include the two last /r/n
state |= STATE_HAS_SIZE;
data.remove_prefix(1);
@@ -71,13 +70,8 @@ namespace uWS {
return state & STATE_HAS_SIZE;
}
// bättre interface
std::optional<std::string_view> getNextChunk(std::string_view data, unsigned int &state);
/* Consumes as much as possible, emitting chunks using cb. Changes passed state for later resumption. Returns number of bytes consumed. */
inline unsigned int consumeChunkedEncoding(std::string_view data, unsigned int &state, MoveOnlyFunction<void(std::string_view)> cb) {
std::string_view originalData = data;
/* Returns next chunk (empty or not), or if all data was consumed, nullopt is returned. */
std::optional<std::string_view> getNextChunk(std::string_view &data, unsigned int &state) {
while (data.length()) {
@@ -85,12 +79,10 @@ namespace uWS {
if ((state & STATE_IS_CHUNKED) && hasChunkSize(state) && chunkSize(state)) {
while(data.length() && chunkSize(state)) {
//printf("dropping 1 byte of trailer\n");
data.remove_prefix(1);
decChunkSize(state, 1);
if (chunkSize(state) == 0) {
//printf("dropped the whole trailer\n");
state = 0;
}
}
@@ -104,7 +96,7 @@ namespace uWS {
// set trailer state and increase size to 4
state = 4 | STATE_IS_CHUNKED | STATE_HAS_SIZE;
cb(std::string_view(nullptr, 0));
return std::string_view(nullptr, 0);
}
continue;
}
@@ -113,29 +105,90 @@ namespace uWS {
if (data.length() >= chunkSize(state)) {
// emit all but 2 bytes then reset state to 0 and goto beginning
// not fin
std::string_view emitSoon;
bool shouldEmit = false;
if (chunkSize(state) > 2) {
cb(std::string_view(data.data(), chunkSize(state) - 2));
emitSoon = std::string_view(data.data(), chunkSize(state) - 2);
shouldEmit = true;
}
data.remove_prefix(chunkSize(state));
state = 0;
if (shouldEmit) {
return emitSoon;
}
continue;
} else {
/* We will consume all our input data */
std::string_view emitSoon;
if (chunkSize(state) > 2) {
unsigned int maximalAppEmit = chunkSize(state) - 2;
if (data.length() > maximalAppEmit) {
cb(data.substr(0, maximalAppEmit));
emitSoon = data.substr(0, maximalAppEmit);
} else {
cb(data);
//cb(data);
emitSoon = data;
}
}
decChunkSize(state, data.length());
return originalData.length();
// new: decrease data by its size (bug)
data.remove_prefix(data.length()); // ny bug fix för getNextChunk
if (emitSoon.length()) {
return emitSoon;
} else {
return std::nullopt;
}
}
}
return originalData.length() - data.length();
return std::nullopt;
}
/* This is really just a wrapper for convenience */
struct ChunkIterator {
std::string_view data;
std::optional<std::string_view> chunk;
unsigned int *state;
ChunkIterator(std::string_view data, unsigned int *state) : data(data), state(state) {
chunk = uWS::getNextChunk(this->data, *state);
if (!chunk && this->data.length()) {
std::abort();
}
}
ChunkIterator() {
}
ChunkIterator begin() {
return *this;
}
ChunkIterator end() {
return ChunkIterator();
}
std::string_view operator*() {
if (!chunk.has_value()) {
std::abort();
}
return chunk.value();
}
bool operator!=(const ChunkIterator &other) const {
return other.chunk.has_value() != chunk.has_value();
}
ChunkIterator &operator++() {
chunk = uWS::getNextChunk(data, *state);
if (!chunk && this->data.length()) {
std::abort();
}
return *this;
}
};
}
#endif // UWS_CHUNKEDENCODING_H
tests/ChunkedEncoding.cpp
+14 -21
View File
@@ -21,32 +21,30 @@ void runTest(unsigned int maxConsume) {
/* Encode them in chunked encoding */
std::stringstream ss;
for (std::string_view chunk : chunks) {
// if length is 0 then append trailer also
/* Generic chunked encoding format */
ss << std::hex << chunk.length() << "\r\n" << chunk << "\r\n";
// every null chunk is followed by an empty trailer
/* Every null chunk is followed by an empty trailer */
if (chunk.length() == 0) {
ss << "\r\n";
}
}
/* Consume them, checking that we get what we expect */
std::string buffer = ss.str();
/* Begin with a clear state and the full data */
unsigned int state = 0;
unsigned int chunkOffset = 0;
std::string_view chunkEncoded = buffer;
while (chunkEncoded.length()) {
/* Parse a small part of the given data */
std::string_view data = chunkEncoded.substr(0, std::min<size_t>(maxConsume, chunkEncoded.length()));
chunkEncoded.remove_prefix(data.length());
unsigned int consumed = UINT_MAX;
int chunkOffset = 0;
while (consumed) {
/* Consume up to maxConsume */
std::string_view indata = chunkEncoded.substr(0, std::min<size_t>(maxConsume, chunkEncoded.length()));
consumed = uWS::consumeChunkedEncoding(indata, state, [&](std::string_view chunk) {
/* Print for logging */
std::cout << "<" << chunk << ">";
/* Whatever chunk we emit, or part of chunk, it must match the expected one */
for (auto chunk : uWS::ChunkIterator(data, &state)) {
std::cout << "<" << chunk << ">" << std::endl;
/* Run check here */
if (!chunk.length() && chunks[chunkOffset].length()) {
std::cout << "We got emitted an empty chunk but expected a non-empty one" << std::endl;
std::abort();
@@ -61,17 +59,12 @@ void runTest(unsigned int maxConsume) {
std::cerr << "Chunk does not match! Should be <" << chunks[chunkOffset] << ">" << std::endl;
std::abort();
}
});
if (consumed != indata.length()) {
std::cerr << "Chunk parser did not consume exactly the bytes passed!" << std::endl;
std::abort();
}
chunkEncoded.remove_prefix(consumed);
}
}
int main() {
for (int i = 0; i < 1000; i++) {
for (int i = 1; i < 1000; i++) {
runTest(i);
}