Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

refactor(x/staking)!: removes the use of Accounts String() method #19742

Merged
merged 15 commits into from
Mar 19, 2024

Conversation

JulianToledano
Copy link
Contributor

@JulianToledano JulianToledano commented Mar 13, 2024

Description

ref:
#13140
#7448

This PR removes the use of Address (AccAddress, ValAddress, ConsAddress) String method in x/staknigmodule.


Author Checklist

All items are required. Please add a note to the item if the item is not applicable and
please add links to any relevant follow up issues.

I have...

  • included the correct type prefix in the PR title
  • confirmed ! in the type prefix if API or client breaking change
  • targeted the correct branch (see PR Targeting)
  • provided a link to the relevant issue or specification
  • reviewed "Files changed" and left comments if necessary
  • included the necessary unit and integration tests
  • added a changelog entry to CHANGELOG.md
  • updated the relevant documentation or specification, including comments for documenting Go code
  • confirmed all CI checks have passed

Reviewers Checklist

All items are required. Please add a note if the item is not applicable and please add
your handle next to the items reviewed if you only reviewed selected items.

I have...

  • confirmed the correct type prefix in the PR title
  • confirmed all author checklist items have been addressed
  • reviewed state machine logic, API design and naming, documentation is accurate, tests and test coverage

Summary by CodeRabbit

  • New Features
    • Enhanced address and validator codecs for improved address conversions and operations across multiple modules.
  • Bug Fixes
    • Resolved issues with address conversion and delegation operations in tests and simulations.
  • Documentation
    • Updated test cases and simulations to align with changes in address and validator codec usage.
  • Refactor
    • Updated simulation state structures to incorporate new codec fields for addresses and validators.
    • Improved message simulation functions to accommodate new codec parameters, enhancing error handling and message creation.

Copy link
Contributor

coderabbitai bot commented Mar 13, 2024

Walkthrough

The updates aim to enhance the Cosmos SDK by integrating address and validator codecs across various modules and tests. These changes ensure consistent handling of address conversions and validations, improving the SDK's robustness and functionality. By adding codec parameters to function calls and struct initializations, the modifications facilitate more flexible and error-resistant address handling within the Cosmos ecosystem.

Changes

Files Change Summary
simapp/sim_bench_test.go, simapp/sim_test.go Updated function calls to include codec parameters for AuthKeeper and StakingKeeper.
tests/integration/evidence/keeper/infraction_test.go, tests/integration/slashing/keeper/keeper_test.go Modified functions for address conversions and operations related to staking and slashing modules.
tests/sims/gov/operations_test.go, x/auth/simulation/genesis_test.go, x/auth/simulation/proposals.go, x/auth/simulation/proposals_test.go Added address.Codec parameter in function signatures and updated simulation and genesis tests.
testutil/sims/simulation_helpers.go, types/module/simulation.go, types/simulation/types.go Included AddressCodec and ValidatorCodec in simulation-related structures and functions.
x/auth/simulation/genesis_test.go, x/authz/client/cli/tx.go, x/authz/msgs_test.go, x/authz/simulation/genesis_test.go Addressed codec integration for authentication and authorization modules.
x/bank/simulation/genesis_test.go, x/distribution/simulation/genesis_test.go, x/gov/simulation/genesis_test.go, x/group/simulation/genesis_test.go, x/mint/simulation/genesis_test.go, x/nft/simulation/genesis_test.go, x/protocolpool/simulation/proposals.go, x/slashing/simulation/genesis_test.go Updated simulation and genesis tests across various modules to include codec options.
x/staking/simulation/genesis_test.go, x/staking/simulation/operations.go, x/staking/simulation/proposals.go, x/staking/simulation/proposals_test.go Enhanced staking module by adding address conversion and codec parameters in various functions.
x/staking/testutil/helpers.go, x/staking/testutil/validator.go, x/staking/types/authz.go, x/staking/types/authz_test.go, x/staking/types/delegation_test.go, x/staking/types/msg_test.go, x/staking/types/validator_test.go Adjusted staking module functions and tests for address conversions and codec integrations.
CHANGELOG.md Documented the changes made in the update.

Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media?

Share

Tips

Chat

There are 3 ways to chat with CodeRabbit:

  • Review comments: Directly reply to a review comment made by CodeRabbit. Example:
    • I pushed a fix in commit <commit_id>.
    • Generate unit-tests for this file.
    • Open a follow-up GitHub issue for this discussion.
  • Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
    • @coderabbitai generate unit tests for this file.
    • @coderabbitai modularize this function.
  • PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
    • @coderabbitai generate interesting stats about this repository and render them as a table.
    • @coderabbitai show all the console.log statements in this repository.
    • @coderabbitai read src/utils.ts and generate unit tests.
    • @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (invoked as PR comments)

  • @coderabbitai pause to pause the reviews on a PR.
  • @coderabbitai resume to resume the paused reviews.
  • @coderabbitai review to trigger a review. This is useful when automatic reviews are disabled for the repository.
  • @coderabbitai resolve resolve all the CodeRabbit review comments.
  • @coderabbitai help to get help.

Additionally, you can add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.

CodeRabbit Configration File (.coderabbit.yaml)

  • You can programmatically configure CodeRabbit by adding a .coderabbit.yaml file to the root of your repository.
  • The JSON schema for the configuration file is available here.
  • If your editor has YAML language server enabled, you can add the path at the top of this file to enable auto-completion and validation: # yaml-language-server: $schema=https://coderabbit.ai/integrations/coderabbit-overrides.v2.json

CodeRabbit Discord Community

Join our Discord Community to get help, request features, and share feedback.

Copy link
Contributor

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review Status

Actionable comments generated: 4

Configuration used: .coderabbit.yml

Commits Files that changed from the base of the PR and between cdc3291 and 20469b4.
Files selected for processing (56)
  • simapp/sim_bench_test.go (1 hunks)
  • simapp/sim_test.go (5 hunks)
  • tests/integration/evidence/keeper/infraction_test.go (2 hunks)
  • tests/integration/slashing/keeper/keeper_test.go (3 hunks)
  • tests/sims/gov/operations_test.go (2 hunks)
  • testutil/sims/simulation_helpers.go (1 hunks)
  • testutil/sims/state_helpers.go (6 hunks)
  • types/module/simulation.go (2 hunks)
  • types/simulation/types.go (2 hunks)
  • x/auth/simulation/genesis_test.go (2 hunks)
  • x/auth/simulation/proposals.go (3 hunks)
  • x/auth/simulation/proposals_test.go (2 hunks)
  • x/authz/client/cli/tx.go (1 hunks)
  • x/authz/msgs_test.go (3 hunks)
  • x/authz/simulation/genesis_test.go (1 hunks)
  • x/bank/simulation/genesis_test.go (2 hunks)
  • x/bank/simulation/proposals.go (3 hunks)
  • x/bank/simulation/proposals_test.go (2 hunks)
  • x/distribution/simulation/genesis_test.go (2 hunks)
  • x/distribution/simulation/proposals.go (3 hunks)
  • x/distribution/simulation/proposals_test.go (2 hunks)
  • x/evidence/simulation/genesis_test.go (2 hunks)
  • x/feegrant/simulation/genesis_test.go (1 hunks)
  • x/gov/simulation/genesis_test.go (2 hunks)
  • x/gov/simulation/operations.go (1 hunks)
  • x/gov/simulation/proposals.go (2 hunks)
  • x/gov/simulation/proposals_test.go (2 hunks)
  • x/group/simulation/genesis_test.go (1 hunks)
  • x/mint/simulation/genesis_test.go (1 hunks)
  • x/mint/simulation/proposals.go (3 hunks)
  • x/mint/simulation/proposals_test.go (2 hunks)
  • x/nft/simulation/genesis_test.go (1 hunks)
  • x/protocolpool/simulation/proposals.go (3 hunks)
  • x/protocolpool/simulation/proposals_test.go (2 hunks)
  • x/slashing/simulation/genesis_test.go (2 hunks)
  • x/slashing/simulation/proposals.go (3 hunks)
  • x/slashing/simulation/proposals_test.go (2 hunks)
  • x/staking/keeper/alias_functions.go (1 hunks)
  • x/staking/keeper/cons_pubkey_test.go (2 hunks)
  • x/staking/keeper/delegation_test.go (31 hunks)
  • x/staking/keeper/keeper_test.go (10 hunks)
  • x/staking/keeper/msg_server_test.go (61 hunks)
  • x/staking/keeper/validator_test.go (3 hunks)
  • x/staking/migrations/v5/migrations_test.go (3 hunks)
  • x/staking/simulation/genesis.go (2 hunks)
  • x/staking/simulation/genesis_test.go (2 hunks)
  • x/staking/simulation/operations.go (4 hunks)
  • x/staking/simulation/proposals.go (3 hunks)
  • x/staking/simulation/proposals_test.go (3 hunks)
  • x/staking/testutil/helpers.go (3 hunks)
  • x/staking/testutil/validator.go (1 hunks)
  • x/staking/types/authz.go (4 hunks)
  • x/staking/types/authz_test.go (25 hunks)
  • x/staking/types/delegation_test.go (4 hunks)
  • x/staking/types/msg_test.go (2 hunks)
  • x/staking/types/validator_test.go (4 hunks)
Check Runs (11)
tests (03) completed (1)
tests (01) completed (1)
test-x-mint completed (1)
test-x-bank completed (1)
test-x-authz completed (1)
test-x-slashing completed (1)
test-x-gov completed (1)
test-x-auth in_progress (1)
test-x-upgrade completed (1)
test-x-params completed (1)
test-x-group completed (1)
Additional comments: 132
x/staking/testutil/validator.go (1)
  • 18-19: The use of require.NoError(tb, err) for error handling in tests is generally acceptable as it will halt the test execution if an error is encountered. However, consider if there's a need for more nuanced error handling or logging in this context, especially for more complex test setups where you might want to clean up resources or provide more detailed error messages.
x/bank/simulation/proposals.go (2)
  • 6-6: The addition of the coreaddress import is appropriate for the changes made in the SimulateMsgUpdateParams function.
  • 34-34: The modification to include an additional parameter addressCodec coreaddress.Codec and changing the return type to (sdk.Msg, error) in the SimulateMsgUpdateParams function enhances error handling and aligns with the PR's objectives for improved address handling.
x/protocolpool/simulation/proposals.go (2)
  • 6-6: The addition of the coreaddress import is appropriate for the changes made in the SimulateMsgCommunityPoolSpend function.
  • 31-31: The modification to include an additional parameter addressCodec coreaddress.Codec and changing the return type to (sdk.Msg, error) in the SimulateMsgCommunityPoolSpend function enhances error handling and aligns with the PR's objectives for improved address handling.
x/distribution/simulation/proposals.go (2)
  • 6-6: The addition of the coreaddress import is appropriate for the changes made in the SimulateMsgUpdateParams function.
  • 35-35: The modification to include an additional parameter addressCodec coreaddress.Codec and changing the return type to (sdk.Msg, error) in the SimulateMsgUpdateParams function enhances error handling and aligns with the PR's objectives for improved address handling.
x/bank/simulation/proposals_test.go (2)
  • 12-12: The addition of the codectestutil import is appropriate for the changes made in the TestProposalMsgs function, providing necessary codec utilities for testing.
  • 36-36: The modification to include an additional argument codectestutil.CodecOptions{}.GetAddressCodec() in the TestProposalMsgs function aligns with the PR's objectives for improved address handling and ensures consistency between the main codebase and tests.
x/distribution/simulation/proposals_test.go (2)
  • 13-13: The addition of the codectestutil import is appropriate for the changes made in the TestProposalMsgs function, providing necessary codec utilities for testing.
  • 37-37: The modification to include an additional argument codectestutil.CodecOptions{}.GetAddressCodec() in the TestProposalMsgs function aligns with the PR's objectives for improved address handling and ensures consistency between the main codebase and tests. The addition of an error check for message creation is a good practice for robust testing.
x/feegrant/simulation/genesis_test.go (1)
  • 25-37: The updates to the SimulationState struct initialization in the TestRandomizedGenState function, including the AddressCodec and ValidatorCodec fields, align with the PR's objectives for improved address handling. This ensures consistency between the main codebase and simulation tests.
x/protocolpool/simulation/proposals_test.go (2)
  • 12-12: The addition of the codectestutil import is appropriate for the changes made in the TestProposalMsgs function, providing necessary codec utilities for testing.
  • 36-36: The modification to include an additional argument codectestutil.CodecOptions{}.GetAddressCodec() in the TestProposalMsgs function aligns with the PR's objectives for improved address handling and ensures consistency between the main codebase and tests.
x/evidence/simulation/genesis_test.go (2)
  • 26-26: The introduction of cdcOpts as testutil.CodecOptions{} is a good practice for managing codec options. This ensures that any codec-related configurations are centralized and easily modifiable.
  • 33-34: Adding AddressCodec and ValidatorCodec to the SimulationState struct is crucial for handling address conversions more robustly. This change aligns with the PR's objective to improve address handling across the SDK.
x/authz/simulation/genesis_test.go (1)
  • 32-33: The inclusion of AddressCodec and ValidatorCodec in the SimulationState struct is a significant improvement. It ensures that address handling is consistent and error-resistant across different modules of the SDK.
x/nft/simulation/genesis_test.go (2)
  • 23-24: The introduction of cdcOpts and its use in MakeTestEncodingConfig is a good practice. It allows for more flexible and configurable codec options, which can be beneficial for testing different encoding scenarios.
  • 32-33: The assignment of AddressCodec and ValidatorCodec using cdcOpts is a positive change. It ensures that address handling is consistent and robust, aligning with the PR's objectives to improve the SDK's code quality.
x/auth/simulation/proposals_test.go (1)
  • 36-37: The addition of codectestutil.CodecOptions{}.GetAddressCodec() as an argument in the MsgSimulatorFn call and the introduction of error handling are significant improvements. They ensure that address handling is consistent with the new codec approach and that potential errors are properly managed.
x/auth/simulation/proposals.go (2)
  • 6-6: The addition of the coreaddress package import is necessary for utilizing the new codec approach for address handling. This aligns with the PR's objectives to improve address handling across the SDK.
  • 31-37: > 📝 NOTE

This review was outside the diff hunks and was mapped to the diff hunk with the greatest overlap. Original lines [34-48]

Modifying the SimulateMsgUpdateParams function to include an addressCodec coreaddress.Codec parameter and changing the return type to (sdk.Msg, error) are crucial changes. They ensure that the function is compatible with the new address handling approach and that potential errors are properly managed.

x/gov/simulation/proposals.go (2)
  • 6-6: The addition of the coreaddress package import is a necessary step for utilizing the new codec approach for address handling within the government simulation proposals. This aligns with the broader objectives of improving address handling across the SDK.
  • 30-31: The modification of the SimulateTextProposal function to include an addressCodec coreaddress.Codec parameter, even though it currently returns nil, nil, prepares the function for future enhancements where address handling might be required.
x/slashing/simulation/proposals_test.go (1)
  • 38-39: The addition of codectestutil.CodecOptions{}.GetAddressCodec() as an argument in the MsgSimulatorFn call and the introduction of error handling are important improvements. They ensure that address handling is consistent with the new codec approach and that potential errors are properly managed.
x/slashing/simulation/proposals.go (2)
  • 7-7: The addition of the coreaddress package import is necessary for utilizing the new codec approach for address handling within the slashing module. This aligns with the broader objectives of improving address handling across the SDK.
  • 33-39: > 📝 NOTE

This review was outside the diff hunks and was mapped to the diff hunk with the greatest overlap. Original lines [36-50]

Modifying the SimulateMsgUpdateParams function to include an addressCodec coreaddress.Codec parameter and changing the return type to (sdk.Msg, error) are crucial changes. They ensure that the function is compatible with the new address handling approach and that potential errors are properly managed.

x/mint/simulation/proposals_test.go (2)
  • 13-13: The addition of the codectestutil package import is appropriate for the changes made in this file, specifically for handling address encoding/decoding.
  • 37-37: The modification to include codectestutil.CodecOptions{}.GetAddressCodec() as an additional parameter in the MsgSimulatorFn call aligns with the PR's objectives to enhance address handling. Ensure that all related function calls across the codebase are updated to include this new parameter.
x/mint/simulation/proposals.go (2)
  • 6-6: The addition of the coreaddress package import supports the PR's goal of improving address handling through enhanced codec utilities.
  • 35-35: The update to SimulateMsgUpdateParams to include an addressCodec coreaddress.Codec parameter and return (sdk.Msg, error) is a positive change, enhancing address handling and error management. Ensure that all calls to this function are updated to handle the new return type properly.
x/staking/simulation/proposals.go (2)
  • 7-7: The import of coreaddress is appropriate for the changes in this file, facilitating improved address handling through codec utilities.
  • 36-36: The modifications to SimulateMsgUpdateParams, including the addition of an addressCodec coreaddress.Codec parameter and improved error handling, align with the PR's objectives for better address management. Ensure proper error handling is implemented wherever this function is called.
x/staking/simulation/proposals_test.go (2)
  • 14-14: The addition of the codectestutil package import is appropriate for the changes in this file, specifically for handling address encoding/decoding.
  • 27-27: The initialization of addressCodec and its use in the TestProposalMsgs function align with the PR's objectives to enhance address handling through codec utilities. Ensure that address handling is consistent across all tests.
x/staking/types/msg_test.go (2)
  • 12-12: The addition of the codectestutil package import supports the PR's goal of enhancing codec handling, specifically for validator codec utilities in this test.
  • 38-38: The introduction of vAddr1 and its use in the NewMsgCreateValidator call within the TestMsgDecode function align with the PR's objectives for improved address handling through codec utilities. Ensure consistent address handling is maintained in all relevant tests.
x/group/simulation/genesis_test.go (1)
  • 25-39: The updates to the SimulationState struct in the TestRandomizedGenState function, including the addition of AddressCodec and ValidatorCodec fields, align with the PR's objectives to standardize address handling using codec utilities. Ensure consistent codec handling is applied across all simulation tests.
x/auth/simulation/genesis_test.go (1)
  • 27-40: The inclusion of AddressCodec and ValidatorCodec fields in the SimulationState struct during its initialization in the TestRandomizedGenState function supports the PR's goal of enhancing codec handling for address and validator information. Ensure that codec handling is consistently applied across all simulation tests.
x/distribution/simulation/genesis_test.go (1)
  • 26-39: The updates to the SimulationState struct in the TestRandomizedGenState function, including the addition of AddressCodec and ValidatorCodec fields, align with the PR's objectives to standardize codec handling across the SDK. Ensure consistent codec handling is maintained in all simulation tests.
simapp/sim_bench_test.go (1)
  • 69-69: The addition of AuthKeeper.AddressCodec() and StakingKeeper.ValidatorAddressCodec() as parameters to AppStateFn aligns with the refactor's goal to improve address handling. Ensure the implementation and usage of these codecs are consistent and error-free across the SDK.
x/slashing/simulation/genesis_test.go (1)
  • 27-27: The initialization of AddressCodec and ValidatorCodec using cdcOpts in the SimulationState struct is a good practice for standardizing address handling. Verify the correct usage of these codecs in simulations and tests to ensure consistency and prevent issues.
x/bank/simulation/genesis_test.go (1)
  • 28-28: The creation of cdcOpts to handle codec options for testing, and its use in initializing AddressCodec and ValidatorCodec in the SimulationState struct, is a positive step towards improving address handling. Ensure the correct usage of these codecs in the context of bank module simulations and tests.
x/mint/simulation/genesis_test.go (1)
  • 25-25: The introduction of cdcOpts for CodecOptions and its use in initializing AddressCodec and ValidatorCodec in the SimulationState struct is a good practice for standardizing address handling within the mint module. Verify the correct usage of these codecs in simulations and tests to ensure consistency and prevent issues.
x/gov/simulation/genesis_test.go (1)
  • 29-29: The initialization of cdcOpts for codec options and its use in setting AddressCodec and ValidatorCodec in the SimulationState struct is a good practice for standardizing address handling within the gov module. Verify the correct usage of these codecs in simulations and tests to ensure consistency and prevent issues.
x/staking/simulation/genesis.go (2)
  • 79-82: Converting ValAddress to strings using ValidatorCodec and introducing ValStrAddress for use in NewValidator and NewDelegation calls aligns with efforts to standardize address handling in the staking module. Ensure the correct implementation and usage of these conversions to prevent issues.
  • 100-104: Similarly, converting AccAddress to strings using AddressCodec for use in NewDelegation is a good practice. Verify the correct implementation and usage of these conversions to ensure consistency and prevent issues in the staking module.
x/staking/keeper/alias_functions.go (1)
  • 62-66: Using validatorAddressCodec.BytesToString(address) for converting validator addresses to strings in IterateBondedValidatorsByPower improves error handling by providing more informative error messages. Ensure the correct implementation and usage of this method across the keeper functions to maintain consistency and prevent issues.
x/staking/migrations/v5/migrations_test.go (1)
  • 93-105: > 📝 NOTE

This review was outside the diff hunks and was mapped to the diff hunk with the greatest overlap. Original lines [84-102]

The introduction of codecOpts for codec options and its use in converting validator and account addresses to strings in migration tests aligns with efforts to standardize address handling within the staking module. Verify the correct implementation and usage of these codecs in the context of staking module migrations and tests to ensure consistency and prevent issues.

x/staking/simulation/genesis_test.go (2)
  • 29-29: The introduction of cdcOpts for codec options is a good practice for managing codec configurations. This change enhances the flexibility and maintainability of codec usage within tests.
  • 37-38: Using AddressCodec and ValidatorCodec within the SimulationState struct is a significant improvement. It ensures that address handling is consistent and error-resistant across the simulation tests. This change aligns with the PR's objective to improve address handling within the Cosmos SDK.
x/staking/testutil/helpers.go (4)
  • 56-58: Converting sdk.ValAddress to string using the ValidatorAddressCodec and handling potential errors with require.NoError is a robust approach. This ensures that address conversions are handled safely and errors are properly checked during tests.
  • 69-71: Similar to the previous comment, the use of ValidatorAddressCodec for address conversion and error checking with require.NoError in the createValidator function is a good practice. It maintains consistency and reliability in address handling.
  • 84-86: The changes in the Delegate function, including the direct use of stakingtypes.NewMsgDelegate without address conversion, seem appropriate given the context. However, it's important to ensure that the delegator and val parameters are correctly formatted as strings elsewhere in the code.
  • 102-104: The Undelegate function's modifications follow the same pattern as the Delegate function, focusing on message creation without direct address conversion in this context. Consistency in handling these operations is crucial for maintainability.
x/staking/types/delegation_test.go (6)
  • 21-25: Introducing codecOpts and utilizing it for address and validator codec operations in tests is a commendable approach. It ensures that address conversions are handled in a standardized and error-resistant manner, aligning with the PR's objectives.
  • 32-34: The use of codecOpts for address conversion in the TestDelegationEqual function is consistent with the overall improvements in address handling. This enhances the reliability of tests by ensuring proper address formatting.
  • 42-47: Applying codecOpts in the TestDelegationString function for address conversions further demonstrates the PR's commitment to improving address handling within the SDK. This change contributes to the maintainability and readability of the code.
  • 59-61: Utilizing codecOpts for address conversion in the TestUnbondingDelegationEqual function is in line with the PR's goals. It ensures that address handling is consistent and reduces the potential for errors.
  • 99-103: The introduction of codecOpts in the TestDelegationResponses function and its use for address conversions is a positive change. It standardizes address handling across tests, contributing to the code's overall quality.
  • 134-145: In the TestRedelegationResponses function, the addition of codecOpts and its application for address and validator codec operations is a significant improvement. It ensures that address conversions are handled in a consistent and error-proof manner.
types/simulation/types.go (2)
  • 11-12: Adding the cosmossdk.io/core/address import is necessary for the subsequent changes in the file. This ensures that the address codec functionality is available for use in simulation types, aligning with the broader goals of improving address handling.
  • 44-44: Modifying the signature of the MsgSimulatorFn function to include an additional parameter cdc address.Codec is a strategic improvement. It allows for more flexible and error-resistant address handling within simulation functions, contributing to the overall code quality.
x/authz/msgs_test.go (2)
  • 58-58: The addition of codectestutil and the use of valAddressCodec in the NewStakeAuthorization function call represent a thoughtful approach to improving the testability and correctness of authorization logic. This ensures that validator addresses are handled correctly in the context of authorization tests.
  • 77-77: Incorporating valAddressCodec into the NewStakeAuthorization function enhances the robustness of the authorization logic by ensuring that validator addresses are converted and handled properly. This change aligns with the PR's objectives to improve address handling.
x/staking/keeper/cons_pubkey_test.go (3)
  • 32-34: Converting validator addresses from bytes to string using the ValidatorAddressCodec and handling potential errors with require.NoError in the TestConsPubKeyRotationHistory function is a good practice. It ensures that address conversions are handled safely and errors are properly checked during tests.
  • 161-163: Similar to previous comments, the use of ValidatorAddressCodec for address conversion and error checking with require.NoError in the setValidators function is a robust approach. It maintains consistency and reliability in address handling.
  • 171-174: The creation of NewDelegation objects with modified parameters, including the conversion of validator addresses, in the setValidators function is consistent with the PR's objectives. This change enhances the maintainability and readability of the code.
x/gov/simulation/proposals_test.go (2)
  • 11-11: The addition of codectestutil import is aligned with the PR's objective to improve address handling by using codec utilities. This change supports the modifications in the TestProposalMsgs function to use the AddressCodec.
  • 34-35: The modification in the TestProposalMsgs function to include codectestutil.CodecOptions{}.GetAddressCodec() as an argument in the MsgSimulatorFn call is a direct implementation of the PR's objective to avoid direct string conversions of addresses. This change, along with the added error handling, enhances the robustness of the address handling in tests.

However, it's important to ensure that the error handling pattern is consistent across the test suite. The use of assert.NilError(t, err) directly after the function call that could produce an error is a good practice, ensuring that tests fail fast on unexpected errors.

One minor suggestion is to add a comment explaining the use of codectestutil.CodecOptions{}.GetAddressCodec() for future maintainers or contributors who might not be familiar with the context of this change.

Consider adding a brief comment explaining the use of codectestutil.CodecOptions{}.GetAddressCodec() to improve code readability and maintainability.

x/staking/types/authz.go (2)
  • 20-21: The addition of the valAddressCodec address.Codec parameter in the NewStakeAuthorization function is a significant improvement towards achieving the PR's objective of enhancing address handling. This change ensures that address conversions within the function are handled more robustly and error-proof.

It's important to ensure that all calls to NewStakeAuthorization throughout the codebase have been updated to include the new valAddressCodec parameter. This change might require modifications in other parts of the Cosmos SDK where NewStakeAuthorization is used.

Verification successful

The verification process confirms that all calls to NewStakeAuthorization throughout the codebase have been updated to include the new valAddressCodec parameter. This ensures compatibility with the updated function signature and aligns with the objective of enhancing address handling. The changes are consistent across tests, client CLI commands, and other usages, indicating a comprehensive update.

Scripts Executed

The following scripts were executed for the analysis:


Script:

#!/bin/bash
# Search for calls to NewStakeAuthorization to ensure the new parameter is included.
rg --type go 'NewStakeAuthorization\('

Length of output: 3003

* 167-197: > 📝 **NOTE** > This review was outside the diff hunks and was mapped to the diff hunk with the greatest overlap. Original lines [157-194]

The modifications in the validateAllowAndDenyValidators function to include the valAddressCodec address.Codec parameter and use it for address conversions are well-aligned with the PR's objectives. This change enhances the robustness of address handling by utilizing the codec for conversions instead of relying on direct string conversions.

The error handling for address conversion failures (sdkerrors.ErrInvalidRequest.Wrap("could not convert validator address")) is appropriate and ensures that the function fails gracefully in case of conversion issues.

One suggestion is to ensure that the error messages are as informative as possible. For instance, including the invalid address in the error message could help with debugging.

Consider including the invalid address in the error message for conversion failures to improve debuggability.

- sdkerrors.ErrInvalidRequest.Wrap("could not convert validator address")
+ sdkerrors.ErrInvalidRequest.Wrapf("could not convert validator address: %s", validator.String())

Note: Ensure that validator.String() does not reintroduce the direct string conversion issue the PR aims to solve.

testutil/sims/simulation_helpers.go (1)
  • 57-61: The addition of AddressCodec and ValidatorCodec to the SimulationState struct and their usage in authtx.NewTxConfig is a positive change. It aligns with the PR's goal to improve address handling by using codecs instead of direct string conversions. This should enhance maintainability and reduce potential errors.
types/module/simulation.go (2)
  • 9-9: The import of the cosmossdk.io/core/address package is necessary for utilizing the address.Codec type in the SimulationState struct. This is a good practice as it leverages the address package for more robust address handling.
  • 149-150: Adding AddressCodec and ValidatorCodec of type address.Codec to the SimulationState struct is a significant improvement. It ensures that address handling is consistent and error-resistant across simulation operations, aligning with the PR's objectives.
x/authz/client/cli/tx.go (3)
  • 214-214: Adding clientCtx.ValidatorAddressCodec as an argument to the NewStakeAuthorization function call for the delegate case is a good practice. It ensures that address handling is consistent and leverages the new codec approach.
  • 216-216: Similarly, for the unbond case, the inclusion of clientCtx.ValidatorAddressCodec in the NewStakeAuthorization function call aligns with the PR's objectives to improve address handling.
  • 218-218: For the redelegate case, adding clientCtx.ValidatorAddressCodec to the NewStakeAuthorization function call is consistent with the refactor's goal. It ensures that address handling uses the new codec approach.
testutil/sims/state_helpers.go (5)
  • 15-15: The import of the cosmossdk.io/core/address package is necessary for the new addressCodec and validatorCodec parameters introduced in various functions. This aligns with the PR's goal of improving address handling.
  • 40-41: The addition of addressCodec and validatorCodec parameters to the AppStateFn function signature is a positive change. It ensures that address handling is consistent and leverages the new codec approach across simulation helpers.
  • 48-53: Similarly, the inclusion of addressCodec and validatorCodec in the AppStateFnWithExtendedCb function signature aligns with the refactor's goal. It ensures that address handling uses the new codec approach.
  • 106-116: > 📝 NOTE

This review was outside the diff hunks and was mapped to the diff hunk with the greatest overlap. Original lines [64-113]

The changes in AppStateFnWithExtendedCbs and AppStateRandomizedFn to include addressCodec and validatorCodec parameters, and the update to the SimulationState struct to include these codecs, are significant improvements. They ensure that address handling is consistent and error-resistant across simulation operations.

  • 230-246: > 📝 NOTE

This review was outside the diff hunks and was mapped to the diff hunk with the greatest overlap. Original lines [201-243]

The instantiation of SimulationState with AddressCodec and ValidatorCodec in AppStateRandomizedFn is a crucial update. It aligns with the PR's objectives to improve address handling by using codecs instead of direct string conversions, enhancing maintainability and reducing potential errors.

x/staking/types/validator_test.go (2)
  • 18-18: The addition of codectestutil import is appropriate for enhancing testing capabilities related to encoding and decoding. This aligns well with the PR's objectives to improve address handling.
  • 338-346: > 📝 NOTE

This review was outside the diff hunks and was mapped to the diff hunk with the greatest overlap. Original lines [140-356]

The changes to use codectestutil for encoding and decoding addresses in tests are well-aligned with the PR's objectives to avoid direct string conversions of addresses. This approach enhances consistency and reduces the potential for errors, which is a good practice.

simapp/sim_test.go (4)
  • 89-89: The inclusion of additional parameters for address codecs in the AppStateFn function call within TestFullAppSimulation is a positive change. It ensures more comprehensive codec handling for addresses, aligning with the PR's objectives to improve address handling.
  • 137-137: Similar to TestFullAppSimulation, the update in TestAppImportExport to include additional parameters for address codecs in the AppStateFn function call is well-justified. It enhances the simulation setup's ability to handle addresses more effectively.
  • 258-258: The changes in TestAppSimulationAfterImport, including additional parameters for address codecs in the AppStateFn function call, are consistent with the PR's objectives. This ensures a more robust simulation setup for address handling.
  • 389-389: Including additional parameters for address codecs in the AppStateFn function call within TestAppStateDeterminism aligns with the PR's goals to improve address handling. This is a commendable update to the simulation tests.
x/staking/types/authz_test.go (3)
  • 28-40: The addition of valAddressToString and accAddressToString helper functions improves code readability and maintainability by centralizing address conversion logic. Ensure that the codec utilities (GetValidatorCodec and GetAddressCodec) used here are correctly implemented and available in the test context.
  • 46-46: The introduction of valAddressCodec for address conversion in tests aligns with the PR's objectives. Ensure that it is consistently used across tests and correctly initialized.
  • 25-84: > 📝 NOTE

This review was outside the diff hunks and was mapped to the diff hunk with the greatest overlap. Original lines [48-372]

Updates to test cases to use the new helper functions for address conversion and the valAddressCodec parameter are consistent with the PR's objectives. Verify that these changes do not alter the intended behavior of the tests and that all necessary cases are covered.

tests/integration/evidence/keeper/infraction_test.go (2)
  • 264-268: Modifications to TestHandleDoubleSign to use address conversion before calling tstaking.Undelegate align with the PR's objectives. Ensure that the address codecs used are correctly implemented and available in the test context.
  • 431-435: Modifications to TestHandleDoubleSignAfterRotation to use address conversion before calling tstaking.Undelegate are consistent with the PR's objectives. Verify the correct implementation and availability of the address codecs used.
x/staking/keeper/keeper_test.go (7)
  • 72-73: The conversion from ModuleAddress to string using BytesToString is a good approach to avoid direct string manipulation. However, ensure that the authority variable is used consistently and securely throughout the codebase.
  • 97-101: The addressToString function correctly utilizes the accountKeeper's AddressCodec for converting addresses to strings. This is a safer approach compared to direct string conversions.
  • 103-107: Similarly, the valAddressToString function uses the stakingKeeper's ValidatorAddressCodec for address conversion. This consistency in handling address conversions is commendable.
  • 364-365: The use of addressToString and valAddressToString within the TestUnbondingDelegationsMigrationToColls test function for setting DelegatorAddress and ValidatorAddress fields in UnbondingDelegation instances is a good practice. It ensures that address conversions are handled through the newly introduced functions, enhancing code maintainability and readability.

Also applies to: 389-390

  • 451-451: In the TestValidatorsMigrationToColls test function, the conversion of validator addresses using valAddressToString before setting the OperatorAddress field in Validator instances is consistent with the PR's objective to standardize address handling. This change improves the code's robustness against potential errors in direct string conversions.

Also applies to: 477-477

  • 509-509: The conversion of validator addresses to strings using valAddressToString in the TestValidatorQueueMigrationToColls test function before appending them to the addrs slice is a good practice. It aligns with the overall goal of the PR to standardize and secure address handling across the codebase.

Also applies to: 526-526

  • 549-551: The use of addressToString and valAddressToString for converting delegator and validator addresses in the TestRedelegationQueueMigrationToColls test function is consistent with the PR's objectives. This approach ensures that address conversions are handled in a standardized manner, reducing the risk of errors.

Also applies to: 572-574

tests/sims/gov/operations_test.go (2)
  • 12-12: The addition of "cosmossdk.io/core/address" is aligned with the PR's objective to improve address handling by using more robust methods instead of direct string conversions. This import is necessary for utilizing address.Codec in the modified function signature.
  • 59-60: The modification of the MsgSimulatorFn function signature to include an additional parameter address.Codec is a direct implementation of the PR's objective to avoid direct string conversions of addresses. This change ensures that address handling within simulations leverages the address.Codec for more robust and error-resistant conversions. However, it's important to ensure that all calls to this function across the codebase are updated to pass the necessary address.Codec parameter.
Verification successful

The verification process confirms that the MsgSimulatorFn function signature change, which includes an additional parameter address.Codec, has been correctly implemented and used across the codebase. All calls to this function now pass the necessary address.Codec parameter, aligning with the PR's objective to enhance address handling within simulations. This ensures that the modification is consistent with the intended improvements in address conversion robustness and error resistance.

Scripts Executed

The following scripts were executed for the analysis:


Script:

#!/bin/bash
# Search for MsgSimulatorFn usage across the codebase to ensure the new signature is correctly used.
rg --type go 'MsgSimulatorFn\('

Length of output: 1687

x/gov/simulation/operations.go (1)
  • 158-161: The addition of error handling for the msgSim function call is a positive change, enhancing the robustness of the simulation logic. It's good practice to handle errors explicitly rather than ignoring them, which this change accomplishes.

However, consider adding more context to the error message returned on line 160 to help with debugging. For instance, prefixing the error with "SimulateMsgSubmitProposal: failed to generate proposal message:" could provide clearer insights into the source of the error when it occurs.

tests/integration/slashing/keeper/keeper_test.go (3)
  • 198-202: The conversion of validator addresses using AddressCodec and ValidatorAddressCodec is a key part of the refactor. Ensure that the BytesToString method correctly handles the address formats specific to the Cosmos SDK. This change enhances readability and maintainability by using more descriptive methods for address handling.
  • 227-232: The delegation logic to meet minimum self-delegation is correctly updated to use the new address conversion methods. This ensures consistency in address handling throughout the test, aligning with the PR's objectives.
  • 441-445: The use of AddressCodec and ValidatorAddressCodec for address conversions in the context of delegating with power is correctly implemented. This change is consistent with the PR's goal of improving address handling and reduces the potential for errors associated with direct string conversions.
x/staking/simulation/operations.go (4)
  • 191-194: In SimulateMsgCreateValidator, the conversion of address from bytes to string using k.ValidatorAddressCodec().BytesToString(address) is correctly implemented. However, ensure that the error handling for this conversion is robust and logs sufficient information for debugging purposes.
  • 351-354: In SimulateMsgDelegate, the conversion of simAccount.Address from bytes to string using ak.AddressCodec().BytesToString(simAccount.Address) is correctly implemented. It's important to verify that the AddressCodec is properly initialized and configured to handle the address formats used in the Cosmos SDK.
  • 564-567: In SimulateMsgCancelUnbondingDelegate, the conversion of simAccount.Address from bytes to string using ak.AddressCodec().BytesToString(simAccount.Address) follows the intended refactor. Ensure that the AddressCodec is consistently used across the SDK for address conversions to maintain uniformity and reduce potential errors.
  • 777-787: In SimulateMsgRotateConsPubKey, the conversion of consensus addresses from bytes to string using k.ConsensusAddressCodec().BytesToString(cons) and the subsequent comparison to ensure the new pubkey is different from the current one are correctly implemented. This change enhances the clarity and correctness of address handling. However, it's crucial to ensure that the ConsensusAddressCodec is thoroughly tested to handle all edge cases in address conversion.
x/staking/keeper/msg_server_test.go (10)
  • 73-74: The conversion of DelegatorAddress and ValidatorAddress using s.addressToString(Addr) and s.valAddressToString(ValAddr) respectively is a positive change, aligning with the PR's objective to avoid direct string conversions of addresses. This approach enhances code readability and maintainability by utilizing helper functions for address conversion.
  • 93-94: Similar to previous comments, using s.addressToString(Addr) for DelegatorAddress and s.addressToString([]byte("invalid")) for ValidatorAddress in test cases is consistent with the PR's goal. However, using a byte slice directly with "invalid" for ValidatorAddress is a good way to simulate an invalid address scenario in tests.
  • 113-114: Again, the use of helper functions for address conversion in test cases is noted and aligns with the PR's objectives. This consistency across test cases is beneficial for maintainability.
  • 133-134: The consistent use of helper functions for address conversion is observed throughout the test cases, which is a good practice. This approach ensures that the code adheres to the PR's objectives and improves code quality.
  • 153-154: Utilizing helper functions for address conversion in test cases is consistently applied here, which is in line with the PR's goals. This consistency is appreciated and enhances the code's maintainability.
  • 173-174: The use of helper functions for address conversion is consistently observed, which aligns with the PR's objectives. This practice is beneficial for code readability and maintainability.
  • 193-194: The consistent application of helper functions for address conversion in test cases is noted. This approach aligns with the PR's goals and is beneficial for code quality.
  • 213-214: Utilizing helper functions for address conversion in test cases is consistently observed, aligning with the PR's objectives. This consistency is beneficial for code quality and maintainability.
  • 233-234: The use of helper functions for address conversion in test cases is consistently applied, which aligns with the PR's objectives. This practice enhances code readability and maintainability.
  • 257-258: Consistent use of helper functions for address conversion in test cases is observed, aligning with the PR's objectives. This approach is beneficial for code quality.
x/staking/keeper/delegation_test.go (11)
  • 47-47: The usage of s.addressToString and s.valAddressToString for address conversion in test cases has been updated. Ensure that these helper functions correctly handle address conversions in line with the PR's objective to improve address handling.
  • 67-71: The creation of multiple delegations using updated address conversion functions is consistent with the PR's goal. Verify that these changes do not affect the logic of the tests and that the address conversions are correctly implemented.
  • 117-117: The use of s.valAddressToString for obtaining the validator's operator address in assertions is noted. Ensure this conversion aligns with the intended address handling improvements and does not introduce any discrepancies in the test outcomes.
  • 184-184: The delegation test has been updated to use the new address conversion functions. Confirm that the delegation logic remains unaffected by these changes and that the address conversions are accurately performed.
  • 382-382: In the TestUnbondDelegation test, the delegation creation uses updated address conversion methods. It's important to verify that these changes are in line with the PR's objectives and do not alter the test's expected behavior.
  • 802-802: The TestRedelegateToSameValidator test checks for an error when attempting to redelegate to the same validator. This test uses the updated address conversion functions. Confirm that the logic of detecting redelegation to the same validator is not impacted by these changes.
  • 824-824: In the TestRedelegationMaxEntries test, the delegation setup uses the new address conversion methods. Ensure that the test's logic for checking maximum redelegation entries remains valid and that the address conversions are correctly implemented.
  • 879-879: The TestRedelegateSelfDelegation test uses updated address conversion functions for setting up self-delegation. Verify that these changes align with the PR's goals and do not affect the test's logic for redelegating self-delegation.
  • 928-928: In the TestRedelegateFromUnbondingValidator test, the delegation setup with updated address conversion methods is noted. Confirm that the test's logic for redelegating from an unbonding validator remains accurate and that the address conversions are properly handled.
  • 997-997: The TestRedelegateFromUnbondedValidator test has been updated to use the new address conversion functions. It's crucial to verify that these changes support the PR's objectives without altering the test's intended behavior.
  • 1196-1196: In the TestUndelegateWithDustShare test, the delegation creation using updated address conversion methods is observed. Ensure that these changes are consistent with the PR's goals and do not impact the test's logic for handling dust shares during undelegation.

x/staking/keeper/validator_test.go Outdated Show resolved Hide resolved
x/staking/keeper/validator_test.go Outdated Show resolved Hide resolved
x/staking/keeper/validator_test.go Outdated Show resolved Hide resolved
x/staking/keeper/validator_test.go Outdated Show resolved Hide resolved
Copy link
Member

@tac0turtle tac0turtle left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

could we get a changelog for the api break, otherwise lgtm

Copy link
Contributor

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review Status

Actionable comments generated: 0

Configuration used: .coderabbit.yml

Commits Files that changed from the base of the PR and between 20469b4 and 45a3a3c.
Files selected for processing (1)
  • x/staking/keeper/validator_test.go (3 hunks)
Check Runs (11)
tests (03) completed (1)
tests (02) completed (1)
tests (01) completed (1)
test-x-mint completed (1)
test-x-upgrade completed (1)
test-x-staking completed (1)
test-x-slashing completed (1)
test-x-params completed (1)
test-x-group completed (1)
test-x-gov completed (1)
test-x-authz completed (1)
Files skipped from review as they are similar to previous changes (1)
  • x/staking/keeper/validator_test.go

x/auth/simulation/proposals.go Outdated Show resolved Hide resolved
@@ -39,7 +41,7 @@ type WeightedProposalMsg interface {
MsgSimulatorFn() MsgSimulatorFn // msg simulator function
}

type MsgSimulatorFn func(r *rand.Rand, ctx sdk.Context, accs []Account) sdk.Msg
type MsgSimulatorFn func(r *rand.Rand, ctx sdk.Context, accs []Account, cdc address.Codec) (sdk.Msg, error)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Breaking this is quite disruptive for all modules implementing simulation. Is there a way around it? After a quick skim I couldn't find for which simulation it is actually used.

If not, we definitely need a section under Modules > All > Simulation in the UPGRADING.md

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I couldn't find a way to avoid it other than using a global codec for simulation. In the end, either the prefix or the codec must be passed all the way through the stack.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

but for which simulation is that param used? I cannot find it 😅

Copy link
Contributor Author

@JulianToledano JulianToledano Mar 13, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

right now just for the staking https://github.com/cosmos/cosmos-sdk/pull/19742/files#diff-ccc155cefb86edd2f89c6436d48cde5df2e5734fa9e7df73fd543a29d318b92dR36-R52

But It'll be updated in coming PRs as the rest of MsgSimulatorFn makes use of the String method.

@JulianToledano
Copy link
Contributor Author

could we get a changelog for the api break, otherwise lgtm

Added a changelog under x/staking. Should add another for simulation?

@julienrbrt

Copy link
Contributor

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review Status

Actionable comments generated: 54

Configuration used: .coderabbit.yml

Commits Files that changed from the base of the PR and between 45a3a3c and d10707f.
Files selected for processing (9)
  • UPGRADING.md (1 hunks)
  • x/auth/simulation/proposals.go (3 hunks)
  • x/bank/simulation/proposals.go (3 hunks)
  • x/distribution/simulation/proposals.go (3 hunks)
  • x/gov/simulation/proposals.go (2 hunks)
  • x/mint/simulation/proposals.go (3 hunks)
  • x/protocolpool/simulation/proposals.go (3 hunks)
  • x/slashing/simulation/proposals.go (3 hunks)
  • x/staking/CHANGELOG.md (1 hunks)
Check Runs (1)
labeler completed (1)
Files skipped from review as they are similar to previous changes (7)
  • x/auth/simulation/proposals.go
  • x/bank/simulation/proposals.go
  • x/distribution/simulation/proposals.go
  • x/gov/simulation/proposals.go
  • x/mint/simulation/proposals.go
  • x/protocolpool/simulation/proposals.go
  • x/slashing/simulation/proposals.go
Additional comments: 9
x/staking/CHANGELOG.md (2)
  • 44-44: The entry for PR #19742 is clear and concise. However, consider adding a brief explanation of the impact of this change for better clarity.
  • 41-47: > 📝 NOTE

This review was outside the diff hunks, and no overlapping diff hunk was found. Original lines [96-96]

The term "key_rotation_fee" is correctly used in the context of staking module changes. However, ensure that this parameter is clearly documented elsewhere to avoid confusion.

UPGRADING.md (7)
  • 143-158: > 📝 NOTE

This review was outside the diff hunks, and no overlapping diff hunk was found. Original lines [4-4]

There seems to be a spelling mistake.

Consider changing "SimApp" to "Sim App" if referring to a simulated application, unless "SimApp" is a specific term used in the Cosmos SDK documentation.

  • 143-158: > 📝 NOTE

This review was outside the diff hunks, and no overlapping diff hunk was found. Original lines [15-15]

This sentence could be improved for clarity.

Consider rephrasing to start with an uppercase letter and ensure proper punctuation for better readability.

  • 143-158: > 📝 NOTE

This review was outside the diff hunks, and no overlapping diff hunk was found. Original lines [19-19]

There's a missing space after a period, which affects readability.

Please add a space after the period for proper sentence separation.

  • 143-158: > 📝 NOTE

This review was outside the diff hunks, and no overlapping diff hunk was found. Original lines [30-30]

The term "depinject" might be unclear to some readers.

Verify that "depinject" is a well-understood term within the Cosmos SDK community or consider adding a brief explanation.

  • 143-158: > 📝 NOTE

This review was outside the diff hunks, and no overlapping diff hunk was found. Original lines [41-41]

There's an extra space before "For depinject users," which might be a typographical error.

Consider removing the extra space for consistency in document formatting.

  • 143-158: > 📝 NOTE

This review was outside the diff hunks, and no overlapping diff hunk was found. Original lines [45-45]

The term "depinject" appears again without explanation.

Ensure that "depinject" is clearly defined or widely understood in the context of Cosmos SDK documentation.

  • 143-158: > 📝 NOTE

This review was outside the diff hunks, and no overlapping diff hunk was found. Original lines [57-57]

The merging of decorators is mentioned without further explanation.

Consider adding a brief explanation or reference to documentation that explains the impact of merging GasConsumptionDecorator and IncreaseSequenceDecorator with SigVerificationDecorator.

x/staking/CHANGELOG.md Outdated Show resolved Hide resolved
UPGRADING.md Show resolved Hide resolved
UPGRADING.md Show resolved Hide resolved
Copy link
Contributor

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review Status

Actionable comments generated: 1

Configuration used: .coderabbit.yml

Commits Files that changed from the base of the PR and between d10707f and ec42726.
Files selected for processing (1)
  • CHANGELOG.md (1 hunks)
Check Runs (1)
labeler completed (1)
Additional comments: 2
CHANGELOG.md (2)
  • 100-100: The description of the change to MakeTestEncodingConfig is clear and accurate.
  • 100-100: The explanation for the removal of the basic manager and related functions is clear and provides a good rationale for the change.

CHANGELOG.md Outdated Show resolved Hide resolved
Copy link
Contributor

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review Status

Actionable comments generated: 4

Configuration used: .coderabbit.yml

Commits Files that changed from the base of the PR and between ec42726 and 91be717.
Files selected for processing (4)
  • CHANGELOG.md (1 hunks)
  • UPGRADING.md (1 hunks)
  • simapp/sim_test.go (5 hunks)
  • x/staking/CHANGELOG.md (1 hunks)
Files skipped from review as they are similar to previous changes (2)
  • CHANGELOG.md
  • simapp/sim_test.go
Additional comments: 2
UPGRADING.md (2)
  • 150-156: The MsgSimulatorFn type has been updated to include an address.Codec parameter. This change is crucial for simulations and should be clearly understood by developers.

Ensure to update simulation functions across your codebase to accommodate this change.

  • 160-160: The introduction of appmodule.Environment represents a significant architectural change. It's important to understand how to effectively use this new interface for fetching different services within the application.

Consider reviewing examples or additional documentation to fully leverage the appmodule.Environment interface.

x/staking/CHANGELOG.md Show resolved Hide resolved
x/staking/CHANGELOG.md Show resolved Hide resolved
Copy link
Contributor

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review Status

Actionable comments generated: 0

Configuration used: .coderabbit.yml

Commits Files that changed from the base of the PR and between 91be717 and ebb185c.
Files selected for processing (1)
  • CHANGELOG.md (1 hunks)
Files skipped from review as they are similar to previous changes (1)
  • CHANGELOG.md

Copy link
Contributor

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review Status

Actionable comments generated: 0

Configuration used: .coderabbit.yml

Commits Files that changed from the base of the PR and between ebb185c and 5b5d080.
Files selected for processing (1)
  • CHANGELOG.md (1 hunks)
Files skipped from review as they are similar to previous changes (1)
  • CHANGELOG.md

Copy link
Contributor

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review Status

Actionable comments generated: 0

Configuration used: .coderabbit.yml

Commits Files that changed from the base of the PR and between 5b5d080 and 7f804f2.
Files selected for processing (1)
  • x/staking/keeper/cons_pubkey_test.go (4 hunks)
Files skipped from review as they are similar to previous changes (1)
  • x/staking/keeper/cons_pubkey_test.go

Copy link
Member

@tac0turtle tac0turtle left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

it is api breaking, but not as many people as we think use the simulator

Copy link
Contributor

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review Status

Actionable comments generated: 1

Configuration used: .coderabbit.yml

Commits Files that changed from the base of the PR and between 7f804f2 and 479c7c8.
Files selected for processing (3)
  • CHANGELOG.md (1 hunks)
  • UPGRADING.md (1 hunks)
  • x/gov/simulation/genesis_test.go (2 hunks)
Files skipped from review as they are similar to previous changes (2)
  • CHANGELOG.md
  • x/gov/simulation/genesis_test.go
Additional comments: 2
UPGRADING.md (2)
  • 150-156: The update to MsgSimulatorFn is a critical change.

Highlight the importance of updating simulation functions to accommodate the new address.Codec parameter, and provide examples if possible.

  • 160-160: The introduction of appmodule.Environment is a notable architectural change.

Consider providing examples or additional documentation to help developers understand how to use the new appmodule.Environment interface effectively.

UPGRADING.md Show resolved Hide resolved
Copy link
Member

@julienrbrt julienrbrt left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ACK, but still not a fan of the api break. I almost think, if we break MsgSimulatorFn we should break it better and remove the arguments that are never used (like sdk.Context), so that it is still an api cleanup instead of just a breakage.

@JulianToledano JulianToledano added this pull request to the merge queue Mar 19, 2024
Merged via the queue into main with commit 65ab253 Mar 19, 2024
71 of 72 checks passed
@JulianToledano JulianToledano deleted the julian/staking-accString-removal branch March 19, 2024 09:24
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

5 participants