feat[lang]: auto-export events in ABI

[charles-cooper]

What I did

depends #3786

after this, the standard token demo looks like this:

from snekmate.tokens import ERC20 as base_token

implements: base_token.IERC20
initializes: base_token

exports: (
    base_token.transfer,
    base_token.transferFrom,
    base_token.approve,
    base_token.allowance,
    base_token.totalSupply,
    base_token.balanceOf,
)

@deploy
def __init__():
    base_token.__init__("My Token", "TOK", 100_000_000, "My DApp", "v0.0.1")

How I did it

How to verify it

Commit message

this commit modifies ABI json generation so that it automatically
exports events.

since this additionally obviates the need for `implements: <interface>`
to check that event declarations conform with the interface, the event
conformance check has been removed from the interface implements check.

the methodology used here is to export only those events which are
detected as used, as opposed to any event imported in the import graph.
this is intended to minimize namespace pollution, since if multiple
imported modules define the same event, they can result in a namespace
collision when computing the interface. this restriction could be
reconsidered in the future.

this commit also implements some refactors to the analysis pipeline:
- move ModuleT computation to *after* call graph analysis, so that
  ModuleT can compute all events used in the call graph
- additionally, move ModuleT computation to after getter expansion,
  so that `exposed_functions` can return `FunctionDef`s so that
  downstream code doesn't need to do the work to resolve the expanded
  getter from `VariableDecl`.
- move call graph analysis outside of `ModuleAnalyzer.analyze()`
- rename `ModuleAnalyzer.analyze()` to `.analyze_module_body()` to
  better reflect what it does

Description for the changelog

Cute Animal Picture

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 91.76471% with 7 lines in your changes are missing coverage. Please review.

Project coverage is 84.87%. Comparing base (a91af13) to head (f8b27ee).
Report is 1 commits behind head on master.

Files	Patch %	Lines
vyper/semantics/analysis/module.py	90.00%	3 Missing and 1 partial ⚠️
vyper/semantics/types/module.py	90.32%	2 Missing and 1 partial ⚠️

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #3808      +/-   ##
==========================================
- Coverage   85.08%   84.87%   -0.21%     
==========================================
  Files          92       92              
  Lines       13875    13907      +32     
  Branches     3113     3116       +3     
==========================================
- Hits        11805    11804       -1     
- Misses       1576     1607      +31     
- Partials      494      496       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[cyberthirst]

Name collisions with imported events are performed based on used_events, thus NamespaceCollision isn't thrown in the cases like the following one:

# ITest
event Foo:
    a: uint256

import ITest

implements: ITest

event Foo:
    a: address

@external
def foo(u: uint256):
    log Foo(empty(address))
    #log ITest.Foo(u)

Testing with an attribute access to i throws IndexError: list index out of range:

import ITest

implements: ITest

event Foo:
    a: address

@external
def foo(u: uint256):
    log Foo(empty(address))
    log i.Foo(empty(address)

[charles-cooper]

Name collisions with imported events are performed based on used_events, thus NamespaceCollision isn't thrown in the cases like the following one:

# ITest
event Foo:
    a: uint256

import ITest

implements: ITest

event Foo:
    a: address

@external
def foo(u: uint256):
    log Foo(empty(address))
    #log ITest.Foo(u)

this is intended, it's an attempt to reduce namespace pollution (instead of just exporting all events found in all imports)

the other option is just to export all events in all imports, but that leads to some weird interactions with the namespace, e.g. you import ERC721 and import ERC20, we can either let there be a namespace collision (which is terrible UX here!) or we do some "refactoring", where the Transfer event is exposed in the ABI, but not in the namespace inside of vyper.

Testing with an attribute access to i throws IndexError: list index out of range:

import ITest

implements: ITest

event Foo:
    a: address

@external
def foo(u: uint256):
    log Foo(empty(address))
    log i.Foo(empty(address)

weird, will investigate

[charles-cooper]

after investigating the parser issue i think it's an unrelated, latent error. it should be fixed, but reviewed in a separate PR