The package `safe-event-emitter` has been updated to v2. This update
includes renaming the package to be scoped under `@metamask`, and it
includes a TypeScript migration.
The `network` store of the network controller crams two types of data
into one place. It roughly tracks whether we have enough information to
make requests to the network and whether the network is capable of
receiving requests, but it also stores the ID of the network (as
obtained via `net_version`).
Generally we shouldn't be using the network ID for anything, as it has
been completely replaced by chain ID, which all custom RPC endpoints
have been required to support for over a year now. However, as the
network ID is used in various places within the extension codebase,
removing it entirely would be a non-trivial effort. So, minimally, this
commit splits `network` into two stores: `networkId` and
`networkStatus`. But it also expands the concept of network status.
Previously, the network was in one of two states: "loading" and
"not-loading". But now it can be in one of four states:
- `available`: The network is able to receive and respond to requests.
- `unavailable`: The network is not able to receive and respond to
requests for unknown reasons.
- `blocked`: The network is actively blocking requests based on the
user's geolocation. (This is specific to Infura.)
- `unknown`: We don't know whether the network can receive and respond
to requests, either because we haven't checked or we tried to check
and were unsuccessful.
This commit also changes how the network status is determined —
specifically, how many requests are used to determine that status, when
they occur, and whether they are awaited. Previously, the network
controller would make 2 to 3 requests during the course of running
`lookupNetwork`.
* First, if it was an Infura network, it would make a request for
`eth_blockNumber` to determine whether Infura was blocking requests or
not, then emit an appropriate event. This operation was not awaited.
* Then, regardless of the network, it would fetch the network ID via
`net_version`. This operation was awaited.
* Finally, regardless of the network, it would fetch the latest block
via `eth_getBlockByNumber`, then use the result to determine whether
the network supported EIP-1559. This operation was awaited.
Now:
* One fewer request is made, specifically `eth_blockNumber`, as we don't
need to make an extra request to determine whether Infura is blocking
requests; we can reuse `eth_getBlockByNumber`;
* All requests are awaited, which makes `lookupNetwork` run fully
in-band instead of partially out-of-band; and
* Both requests for `net_version` and `eth_getBlockByNumber` are
performed in parallel to make `lookupNetwork` run slightly faster.
The network state is now passed to the TransactionController via a
getter function and a subscription function, instead of passing one of
the network controller stores directly.
This way of passing the state makes further refactoring easier, as we
don't have to change the input when the store is changed or replaced.
It's also more aligned with our conventions today.
This change was made as part of a larger refactor of the network
controller, as part of the effort to merge the mobile and extension
network controllers.
* we should not call normalize if we're failing the transaction
refactor out update history
we should fail if un update we get an error and the warning message is
error submitting.
Signed-off-by: Akintayo A. Olusegun <akintayo.segun@gmail.com>
* refactor _setTransactionStatus
Signed-off-by: Akintayo A. Olusegun <akintayo.segun@gmail.com>
ESLint rules have been added to enforce our JSDoc conventions. These
rules were introduced by updating `@metamask/eslint-config` to v9.
Some of the rules have been disabled because the effort to fix all lint
errors was too high. It might be easiest to enable these rules one
directory at a time, or one rule at a time.
Most of the changes in this PR were a result of running
`yarn lint:fix`. There were a handful of manual changes that seemed
obvious and simple to make. Anything beyond that and the rule was left
disabled.
* Fixes#12145
When transactions fail, preserve error message when presented with an error object. Fallback to stringified error for other cases.
* Perserve error stack when err.stack is present, else set stack to message.
* Use optional chaining
A few inconsistencies in JSDoc formatting have been fixed throughout
the project. Many issues remain; these were just the few things that
were easy to fix with a regular expression.
The changes include:
* Using lower-case for primitive types, but capitalizing non-primitive
types
* Separating the parameter identifier and the description with a dash
* Omitting a dash between the return type and the return description
* Ensuring the parameter type is first and the identifier is second (in
a few places it was backwards)
* Using square brackets to denote when a parameter is optional, rather
than putting "(optional)" in the parameter description
* Including a type and identifier with every parameter
* Fixing inconsistent spacing, except where it's used for alignment
* Remove incorrectly formatted `@deprecated` tags that reference non-
existent properties
* Remove lone comment block without accompanying function
Additionally, one parameter was renamed for clarity.
All transaction status updates were moved into a `setTimeout` callback
and wrapped in a `try...catch` block in #4131, apparently in an attempt
to prevent failures in event subscribers from interrupting the
transaction logic. The `try...catch` block did accomplish that, but by
putting the status update in a `setTimeout` callback the operation was
made asynchronous.
Transaction status updates now happen unpredictably, in some future
event loop from when they're triggered. This creates a race condition,
where the transaction status update may occur before or after
subsequent state changes. This also introduces a risk of accidentally
undoing a change to the transaction state, as the update made to the
transaction inside the `setTimeout` callback uses a reference to
`txMeta` obtained synchronously before the `setTimeout` call. Any
replacement of the `txMeta` between the `setTxStatus` call and the
execution of the timeout would be erased. Luckily the `txMeta` object
is more often than not mutated rather than replaced, which may explain
why we haven't seen this happen yet.
Everything seems to work correctly with the `setTimeout` call removed,
and now the transaction logic is easier to understand.
* Specify type before parameter name
Various JSDoc `@param` entries were specified as `name {type}` rather
than `{type} name`.
A couple of `@return` entries have been given types as well.
* Use JSDoc optional syntax rather than Closure syntax
* Use @returns rather than @return
* Use consistent built-in type capitalization
Primitive types are lower-case, and Object is upper-case.
* Separate param/return description with a dash
* Ensures the tx controller + tx-state-manager orders transactions in the order they are received
* Handle transaction ordering in cases where tx ids are off by more than 1 in tx-state-manager
* Add comment to addUnapprovedTransaction explaining calling _determineTransactionCategory after generateTxMeta
* Sort txes by timestamp of creation instead of id
* eslint: Check for unused function arguments
* eslint: Ignore unused '_' in argument list
Also allow any number of '_' e.g., '__' or '___' which is to be used sparingly
* Remove and rename unused arguments
