Fix FFDH CryptographyDeprecationWarning raised when importing spiffe - #446
Merged
maxlambrecht merged 1 commit intoAug 8, 2026
Merged
Conversation
c4rlo
changed the title
`certificate_utils.PRIVATE_KEY_TYPES` was a hand-rolled `Union` that duplicated cryptography's `PrivateKeyTypes`. It was added as `_PRIVATE_KEY_TYPES` in 77537c6, and `dh.DHPrivateKey` was added to it later in 95b2e32. That copy had two problems: - It referenced `dh.DHPrivateKey`. cryptography 50.0.0 deprecated Diffie-Hellman over finite fields (FFDH) (https://cryptography.io/en/latest/changelog/#v50-0-0), so touching that attribute now emits a `CryptographyDeprecationWarning` at module import time. Because that warning subclasses `UserWarning` (not `DeprecationWarning`), a strict `filterwarnings = error` policy turns importing `spiffe` into an ImportError before any code runs. This became reachable once the cryptography upper bound was removed in 429c165, relaxing the `>=46,<48` pin from 8deed17 to `>=46` and letting cryptography 50 resolve. - The copy had drifted from cryptography and no longer matched what the loaders return (e.g. it predated the ML-DSA/ML-KEM key types). Alias `cryptography.hazmat.primitives.asymmetric.types.PrivateKeyTypes` directly. That is exactly what `load_der_private_key`/`load_pem_private_key` are declared to return: both are annotated `-> PrivateKeyTypes` in cryptography's type stub `hazmat/bindings/_rust/openssl/keys.pyi` (https://github.com/pyca/cryptography/blob/50.0.0/src/cryptography/hazmat/bindings/_rust/openssl/keys.pyi#L13-L26). The `parse_*_private_key` wrappers just return those calls, so they now annotate precisely what they produce. The alias also avoids the deprecated attribute reference and stays in sync with cryptography. `PRIVATE_KEY_TYPES` is only used as a type annotation, so there is no runtime behaviour change. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Signed-off-by: Carlo Teubner <cteubner1@bloomberg.net>
c4rlo
force-pushed
the
fix-private-key-types-ffdh-warning
branch
from
August 7, 2026 11:40
8873d12 to
40e5ed7
Compare
Collaborator
|
Hi @c4rlo, thanks for this. Is it ready for review? |
Contributor
Author
|
Hi, yes it's ready for review, thanks! I've still got it in draft because I still need to get internal signoff to contribute code to this project, which I hope won't take long to arrive. So please don't merge it just yet, if that's okay. |
Collaborator
Sounds good, thanks! I'll review it and hold off on merging. |
maxlambrecht
approved these changes
Aug 7, 2026
maxlambrecht
left a comment
maxlambrecht
left a comment
Collaborator
There was a problem hiding this comment.
LGTM, thanks @c4rlo!
c4rlo
marked this pull request as ready for review
August 7, 2026 16:59
Contributor
Author
|
Thank you, good to merge from my end! I'd much appreciate it if this could make its way into a new release soon. |
Collaborator
|
Released in v0.3.1. Thanks again @c4rlo! |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
Summary
certificate_utils.PRIVATE_KEY_TYPESis a hand-rolledUnionthat duplicatescryptography'sPrivateKeyTypes. It was added as_PRIVATE_KEY_TYPESin 77537c6 (#41), anddh.DHPrivateKeywas added to it later in 95b2e32 (#107).cryptography50.0.0 deprecated Diffie-Hellman over finite fields (FFDH), so merely touchingdh.DHPrivateKeynow emits aCryptographyDeprecationWarningat module import time:Because
CryptographyDeprecationWarningsubclassesUserWarning(notDeprecationWarning), any downstream project that runs under a strictfilterwarnings = errorpolicy fails to importspiffeat all — the warning is raised as an error during import, before any code runs.This became reachable once the
cryptographyupper bound was removed in 429c165 (>=46,<48→>=46, relaxing the pin from 8deed17), allowingcryptography50 to resolve.The copied union had also drifted from
cryptography: it no longer matches what the key loaders can return (e.g. it predates the ML-DSA / ML-KEM key types).Change
Alias
cryptography's canonical type instead of hand-maintaining a copy:This is exactly what
load_der_private_key/load_pem_private_keyare declared to return — both are annotated-> PrivateKeyTypesincryptography's type stubhazmat/bindings/_rust/openssl/keys.pyi— andparse_*_private_keysimply returns those calls, so the annotations now match precisely what is produced. It also removes the deprecated attribute reference and stays in sync withcryptographygoing forward.PRIVATE_KEY_TYPESis only ever used as a type annotation (no runtimeisinstance), so there is no runtime behaviour change. Lint (ruff/mypy/pyright, 100% type completeness) and the full unit suite (376 passed) are green.🤖 Generated with Claude Code