I have a utility that executes a query and send each Row though a channel and implements paging. But I experienced some segfault:

#0  core::str::run_utf8_validation () at src/libcore/str/mod.rs:1531
#1  0x000055dd508053ac in core::str::from_utf8 () at src/libcore/str/mod.rs:342
#2  0x000055dd506db27a in cassandra_cpp::cassandra::value::Value::get_str::{{closure}} ()
    at registry/src/github.com-1ecc6299db9ec823/cassandra-cpp-0.15.0/src/cassandra/value.rs:421
#3  core::result::Result<T,E>::and_then (self=..., op=...)
    at /rustc/5e7af4669f80e5f682141f050193ab679afdb4b1/src/libcore/result.rs:727
#4  cassandra_cpp::cassandra::value::Value::get_str (self=<optimized out>)
    at registry/src/github.com-1ecc6299db9ec823/cassandra-cpp-0.15.0/src/cassandra/value.rs:416
#5  0x000055dd506db3cc in cassandra_cpp::cassandra::value::Value::get_string (self=0x0)
    at registry/src/github.com-1ecc6299db9ec823/cassandra-cpp-0.15.0/src/cassandra/value.rs:428
#6  0x000055dd5050e758 in <cassandra_cpp::cassandra::row::Row as cassandra_cpp::cassandra::row::AsRustType<alloc::string::String>>::get_by_name (self=<optimized out>, name=...)
    at registry/src/github.com-1ecc6299db9ec823/cassandra-cpp-0.15.0/src/cassandra/row.rs:98
#7  <cassandra_cpp::cassandra::row::Row as services::cql::extensions::AsRustTypeExt<T>>::get_by_name (self=<optimized out>, name=...)
    at services/src/cql/extensions.rs:45

(gdb) select-frame 2
(gdb) info locals
slice = &[u8] {data_ptr: 0x0, length: 34}
message_length = <optimized out>
message_ptr = <optimized out>
slice = <optimized out>
err = <optimized out>
val = <optimized out>

It makes sense though, as iterating with cass_iterator_get_row invalidates the previous row. https://docs.datastax.com/en/developer/cpp-driver/1.0/api/struct.CassIterator/#function-cass_iterator_get_row

https://github.com/Metaswitch/cassandra-rs/blob/09d5bbbd05c9525e3600be8a15d3dcb2c03239b8/src/cassandra/result.rs#L187-L192

So, I suppose the Row type should not be Send or the iterator item should be a reference with a lifetime reflecting that, to prevent misuse like I did.

The cassandra cpp driver ignores the value length parameter and tries to find a null terminator in the passed string. Worked around using cass_statement_bind_bytes_by_name_n which is compatible and with string type and honor the value length parameter.

I tried using simple.rs to generate information and apply it. But when I generate a new field and specify the data type as list, the println! on the terminal is [<error>]. How do I display the correct list information on the terminal?

decimal by default pulls in various extra dependencies including the ancient rustc-serialize.

I can see that the project would like to reintroduce decimal support https://github.com/Metaswitch/cassandra-rs/issues/122 so I didnt outright remove this unused dep. But maybe you would prefer me to just remove the dep so that decimal support can start from scratch and be free to choose a more up to date decimal crate?

In these functions you use the unsafe keyword for almost the entrie function body.

We need to mark unsafe operations more precisely using unsafe keyword. Keeping unsafe blocks small can bring many benefits. For example, when mistakes happen, we can locate any errors related to memory safety within an unsafe block. This is the balance between Safe and Unsafe Rust. The separation is designed to make using Safe Rust as ergonomic as possible, but requires extra effort and care when writing Unsafe Rust.

Hope this PR can help you. Best regards. References https://doc.rust-lang.org/nomicon/safe-unsafe-meaning.html https://doc.rust-lang.org/book/ch19-01-unsafe-rust.html

currently this crate fails to compile on M1 / arm64 based machines due to this issue: https://github.com/remacs/remacs/issues/1393

This PR fixes this by casting to libc::c_char rather than i8

Session::execute_batch and Session::execute_batch_with_payloads now take &Batch rather than Batch.

This was apparently an oversight in the original code; the parallel execute takes &Statement already.

This is a breaking change, but it's a straightforward fix: just add & to your argument list.

While we're here, fix a couple of warnings in test code related to printing panics, and remove travis-cargo (which was intermittently broken, and only used for coverage which we don't look at).

Fixes #118

The following line in build.rs causes the problem: println!("cargo:rustc-flags=-l dylib=stdc++");

      ld: library not found for -lstdc++
      clang: error: linker command failed with exit code 1 (use -v to see invocation)

Can you please disable it, at least for macOS builds? Thanks.

On a Macbook Air with the M1 chip, attempting to run cargo build on a simple Rust project that simply imports from cassandra_cpp and it fails due to a mismatched raw pointer type

error[E0308]: mismatched types
  --> /Users/brandondail/.cargo/registry/src/github.com-1ecc6299db9ec823/cassandra-cpp-0.16.0/src/cassandra/schema/aggregate_meta.rs:52:22
   |
52 |             raw2utf8(name, name_length).expect("must be utf8")
   |                      ^^^^ expected `u8`, found `i8`
   |
   = note: expected raw pointer `*const u8`
              found raw pointer `*const i8`

error[E0308]: mismatched types
  --> /Users/brandondail/.cargo/registry/src/github.com-1ecc6299db9ec823/cassandra-cpp-0.16.0/src/cassandra/schema/aggregate_meta.rs:62:22
   |
62 |             raw2utf8(name, name_length).expect("must be utf8")
   |                      ^^^^ expected `u8`, found `i8`
   |
   = note: expected raw pointer `*const u8`
              found raw pointer `*const i8`

error[E0308]: mismatched types
   --> /Users/brandondail/.cargo/registry/src/github.com-1ecc6299db9ec823/cassandra-cpp-0.16.0/src/cassandra/schema/keyspace_meta.rs:150:22
    |
150 |             raw2utf8(name, name_length).expect("must be utf8")
    |                      ^^^^ expected `u8`, found `i8`
    |
    = note: expected raw pointer `*const u8`
               found raw pointer `*const i8`

error: aborting due to 3 previous errors

For more information about this error, try `rustc --explain E0308`.
error: could not compile `cassandra-cpp`

parking_lot::Mutex is faster than std::sync::Mutex, so let's use it for the critical lock in CassFuture.

This section of parking_lot's docs describes the difference: https://docs.rs/parking_lot/0.11.0/parking_lot/type.Mutex.html#differences-from-the-standard-library-mutex

being forked doesn't allow code search on GitHub which is very convenient. is it possible to convert this fork to a first-class repository?

also have you contacted the original author @tupshin about transfering the repo? that repo still shows up first in Google search from what I've seen

If a ResultIterator is dropped while a Row it returned is still alive then we hit undefined behavior. I have seen this exposed as garbage data and segfaults. Specifically cass_iterator_free is the underlying API call, its documentation does not say not to call it while children are alive, so not sure if its our bug or an upstream bug.

Specifically the driver version that exposed this issue to us was 0.16.2, the problem did not occur in 0.16.1. Its likely caused by one of these performance PRs:

• https://github.com/datastax/cpp-driver/pull/505
• https://github.com/datastax/cpp-driver/pull/510

I have a project that uses this crate but it fails because the linker cannot find 'cryoto.lib' I have installed correctly every dependencies. any idea ?

I can't find a way to get the number of parameters on a PreparedStatement. Also there is no way to get the actual type of the parameter by name, because the parameter_data_type_by_name returns a ConstDataType which is defined in a private module and not exposed in the public API.

Exposing such information would be useful in all the use-cases where the CQL string is provided by the user or built dynamically, so the types are not known at the time of writing the program and cannot be hardcoded.

I noticed the Decimal type related functionality of the driver is not exported in this crate. For example the bind_decimal method is commented out: https://github.com/Metaswitch/cassandra-rs/blob/master/src/cassandra/statement.rs#L682

What is the reason behind that? Is there a workaround clients can use? What would need to be solved so this crate can expose that functionality?

A prepared statement that has a SELECT * in it is unsafe.

More details here:

• https://docs.datastax.com/en/developer/java-driver/3.0/manual/statements/prepared/#avoid-preparing-select-queries
• https://issues.apache.org/jira/browse/CASSANDRA-10786

This is a very very bad footgun, and can cause your driver to return incorrect rows when a table is altered.

We should disallow these kind of queries from being prepared, as it can introduce soundness bugs in one's code.

Currently, the underlying C++ library defaults everything to a consistency level of ONE. (https://docs.datastax.com/en/developer/cpp-driver/2.15/api/struct.CassStatement/#function-cass_statement_set_consistency)

It doesn't allow you to specify a more sensible default. Since we're wrapping this library, we could very easily support setting a consistency level on a Cluster or Session level, and propagating that change down into the execution handler.

This has been a footgun for us, where we forget to set consistency level on a query to have it be doing ONE reads/writes instead of QUORUM read/writes. By providing an API that lets us set the default consistency level, this will let users of the library opt out of this nonsense default from the C++ driver, and not have to worry about setting the consistency level with each statement they execute. Perhaps even we should be opinionated here and say that the default consistency level should be LOCAL_QUORUM, because honestly, ONE is an insane default that can lead to data loss/inconsistency.

Our code right now has a LOT of x.set_consistency(cassandra::Consistency::LOCAL_QUORUM)?; and it's pretty dangerous if forgotten.