Fallback entrypoints#
This guide explains how to use fallback entrypoints, which can be useful when creating a proxy for your smart contract.
Preparation#
Make sure you have the Rust source code for a smart contract module you wish to add a fallback entrypoint to. You also need to have the following installed:
Rust
Cargo
The
wasm32-unknown-unknown
compiling targetcargo-concordium
See also
For instructions on how to install the developer tools, see Install tools for development.
What are fallback entrypoints#
A fallback entrypoint is a special kind of entrypoint for a contract. If defined for a given contract, it acts as a catchall for handling invocations of entrypoints that do not exist in the contract.
Consider the contract A
, which has two entrypoints, foo
and a fallback
entrypoint.
If you invoke foo
on A
, foo
is simply invoked.
But if you invoke any other entrypoint, for example bar
, then the fallback
entrypoint is invoked.
The fallback entrypoint will then have access to the parameter, amount, and,
notably, the name of the entrypoint you attempted to invoke (bar
).
Note, that if A
did not have a fallback entrypoint defined, invoking
bar
would result in a missing entrypoint error.
Using fallback entrypoints#
You can create fallback entrypoints by making a new entrypoint and adding
the fallback
attribute to it:
#[receive(contract = "MyContract", fallback)]
fn receive_fallback(
ctx: &ReceiveContext,
host: &Host<State>,
) -> ReceiveResult<MyReturnType> {
// ...
}
Note
For fallback entrypoints, you don’t (and cannot) use the name
attribute.
This is because a fallback entrypoints always have the empty string as the
name and that is, in fact, how the node recognizes it as a fallback entrypoint.
Also note that there can only ever be one fallback entrypoint per contract.
The named_entrypoint
method on HasReceiveContext
allows you to get the
name of the entrypoint used to the reach the fallback entrypoint.
#[receive(contract = "MyContract", fallback)]
fn receive_fallback(
ctx: &ReceiveContext,
host: &Host<State>,
) -> ReceiveResult<MyReturnType> {
let entrypoint_invoked = ctx.named_entrypoint();
}
In the example from the previous section, named_entrypoint
would return the
entrypoint "bar"
.
Note
The named_entrypoint
method is available in all receive methods, but it
is only useful in fallback receive methods. In regular receive methods, it
just returns the entrypoint name specified with the name
attribute.
See also
With fallback entrypoints, it is possible to create a proxy contract, which can act as the public contract address for your (proxied) smart contract. This allows you to upgrade the proxied contract, for example to fix bugs. An example proxy contract can be seen here.