Getting size of deserialized bson document

[andrejlevkovitch]

Sometimes bson documents can be combined one after another (for example when asynchronously read from socket) and in this case they must be readed separately. First document can be readed as:

nlohmann::json::from_bson(body, false);

But now is a problem: how to read second document? Because we don't know haw match bytes already read. Of cause we can read first 4 bytes in body and get size of bson manually, but, I think, this will be a usefull feature, if function from_bson also return capacity of readed bytes (for example as 4-th argument). In this case reading combined bson documents will be match simpler:

size_t &length = 0;
nlohmann::json::from_bson(body, false, true, &length);

// and reading second document
nlohmann::json::from_bson(body.begin() + length, body.end());

What do you think about this?

Which version of the library did you use?

• latest release version 3.7.3
• other release - please state the version: ___
• the develop branch

[FrancoisChabot]

The correct (in my opinion) syntax for this would be:

auto current = body.begin();
auto first = nlohmann::json::from_bson(current, body.end(), false);
auto second = nlohmann::json::from_bson(current, body.end(), false);

This could be easily achieved with proper iterator-based input (which is also how I think #1813 should be tackled)

However, since bson, as a format, is prefixed with the document size, you can currently do:

std::uint32_t first_msg_size = 0;
assert(body.size() >= 4);
std::memcpy(&first_msg_size, body.data(), 4);

// insert potential endian flip here

assert(body.size() >= first_msg_size);
auto separator = std::next(body.begin(), first_msg_size);
auto first = nlohmann::json::from_bson(body.begin(), separator);
auto second = nlohmann::json::from_bson(separator , body.end());

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.