feat: multi-pass type name resolution #2213
Open
+3,184
−305
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I've been meaning to use this approach for a long time, because the attempts at avoiding type collisions via structure suffixes or prefixes work sporadically, at best.
Conflict resolution is fundamentally a global problem, not a local problem when doing recursive traversal, so this PR splits the code generation into two parts. First, the OAPI document structure is traversed, and all the schemas that we generate are gathered up into a list of candidates, then we do global conflict resolution across the space of all schemas. This allows us to preserve the functionality of things affected by schema name -
$ref,requiredproperties, and so forth.This fixes issue #1474 (client response wrapper type colliding with a component schema of the same name and improves issue #200 handling (same name across schemas, parameters, responses, requestBodies, headers).
The new system is gated behind the existing
resolve-type-name-collisionsoutput option. When disabled, behavior is unchanged, oapi-codegen exits with an error. This flag is default false, so there is no behavior change to oapi-codegen unless it's specified. All current test files regenerate without any differences.Added a comprehensive test which reproduces the scenarios in all the PR's and Issues below, and adds a few more, to make sure that references to renamed targets are also correct..
Key changes:
Obsoletes issues:
Obsoletes PRs:
EOF
)