Lint 6: duplicate-mutable-accounts #6
Conversation
|
Completed upgrade so lint can effectively handle cases where number of identical accounts > 2. All tests pass except for |
oslfmt
force-pushed
the
6-duplicate-mutable-accounts
branch
3 times, most recently
from
8d83daa
to
d0c2455
Compare
| /// represent the identifiers being compared. Following the previous example, this would be `(a, b)`. | ||
| fn generate_possible_expected_constraints( | ||
| identical_types: &[(Symbol, Span)], | ||
| ) -> Vec<((TokenStream, TokenStream), (Symbol, Symbol))> { |
There was a problem hiding this comment.
add closure argument to do comparison in function to not have to store generated streams
oslfmt
force-pushed
the
6-duplicate-mutable-accounts
branch
2 times, most recently
from
e42f644
to
12caca7
Compare
|
@smoelius lint 6 is completely done as it stands. Only small thing that is bugging me (and this is personal preference) is that when there are no key checks at all (not using anchor macro constraints, nor alternate constraints such as in function bodies), then the lint will print messages related to the v2 lint I wrote, which doesn't have as nice messages. This is due to some of the logic of the program, and its fairly involved to change it than one might think. Ideally, if no constraints are in the program at all, I would have liked for the lint to print the v1 messages, ie, the one that recommends adding the |
|
also if there's too many commits, I can squash |
|
|
|
@victor-wei126 Can you sign the CLA? |
| fn is_substream(stream: &TokenStream, other: &TokenStream) -> bool { | ||
| let other_len = other.len(); | ||
| for i in 0..stream.len() { | ||
| for (j, other_token) in other.trees().enumerate() { |
There was a problem hiding this comment.
I have a hunch this could be written more simply as:
if other.trees().enumerate().all(|(j, other_token)| { ... }) {
return true;
}
oslfmt
force-pushed
the
6-duplicate-mutable-accounts
branch
from
e8b0ab4
to
bbbb0a9
Compare
smoelius
force-pushed
the
6-duplicate-mutable-accounts
branch
2 times, most recently
from
94b6cf5
to
85a445b
Compare
smoelius
force-pushed
the
6-duplicate-mutable-accounts
branch
from
85a445b
to
e3d777b
Compare
| for next in current + 1..exprs.len() { | ||
| if values.check_key_constraint(exprs[current], exprs[next]) { | ||
| // if there is at least one alt constraint, set flag to false | ||
| self.no_alternate_constraints = false; |
There was a problem hiding this comment.
This could flip back to true if the program contains more than one function, right? (Or am I misunderstanding?)
I'm wondering if no_alternate_constraints and spans should be maintained per function, similar to how anchor_accounts is maintained per struct def.
Quick and somewhat-dirty implementation of lint 6. Tests pass, but may need further work/clean-up when encountering >2 duplicate accounts.