| Age | Commit message (Collapse) | Author |
|
Pure formatting sweep (import ordering, line wrapping); the tree
had accumulated rustfmt drift. Separated from any semantic change
so it stays reviewable and revertable on its own.
|
|
Breaking: every success body is now {"data", "meta": {"state",
"took_ms", "request_id"}, "notes", "next"}; clients read `data`. Only
`/` (plain text) and `/openapi.json` are unwrapped.
* src/response.rs (Envelope, Meta, State, Ctx, RequestInfo): handlers
take Ctx first and build the response with ctx.envelope(data), so
timing and the request id are never filled in by hand. `state` is
`complete` or `partial`; notes are server-authored only; `next` are
concrete relative hrefs (path_segment percent-encodes names).
* src/routes/*.rs, src/lib.rs (version): every handler returns
Envelope<..> with next links; /search is `partial` when the locus
lookup fails or a sorted 1000-hit window may be truncated, and
/publications/{pubmed_id}/traits when datasets were left out for
authorization.
* src/error.rs (ApiError): add `took_ms`, an RFC 7807 extension member.
* src/auth.rs (AuthToken): the 401 now carries request_id and took_ms
(it never had a request_id); Ctx stores RequestInfo in the request
extensions for it.
* tests: common::data / envelope_data parse every success body through
the envelope; tests/envelope.rs and test/python/test_envelope.py check
the shape, errors, and that every `next` link resolves;
tests/envelope_partial.rs checks the `partial` search without a
database; tests/openapi.rs fails if a JSON success response is not an
Envelope_* schema.
* README.md, docs/SKILL.md, docs/gn-auth.md, RELEASE_NOTES.md: document
the envelope and took_ms on errors.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
|
|
Rename: POST /dev/token -> POST /get-test-token (src/routes/dev.rs,
TestToken replacing DevToken), matching tests/get_test_token.rs and
test/python/test_get_test_token.py (renamed from
tests/dev_token.rs/test_dev_token.py). Config::dev_token_enabled() and
the GN_AUTH_DEV_* config keys keep their names -- they're about the
dev-token feature generically, not tied to this one route's path.
Every endpoint now requires a valid token, not just the
dataset_probeset*/traits* family: species(_by_name/_datasets),
groups(_by_name/_strains/_traits*/_datasets) and dataset_by_name all
gain an AuthToken parameter. /, /version, /openapi.json and
/get-test-token itself are the only exceptions -- a health check, a
version check, the spec describing all this, and the endpoint that
hands out a token in the first place (which would be circular if it
needed one).
Rather than repeating security(("bearer_token" = [])) on every
#[utoipa::path], ApiDoc now declares it once as the document's default
security(...), and the three exceptions besides /get-test-token (/,
/version) override it with their own empty security() -- standard
OpenAPI semantics (an operation's own security, even empty, overrides
the document default; an absent one inherits it). error::Unauthorized,
a new #[derive(utoipa::IntoResponses)] type wrapping ApiError, is the
shared 401 every gated handler now references bare in its own
responses(...) list instead of writing out the same
(status = 401, description = "...", body = ApiError) tuple by hand.
Verified the resolved security per operation in a live /openapi.json.
tests/common::auth_token() (Rust) and the new
test/python/auth_helpers.auth_token() (Python, deduplicating what four
files previously each defined locally) now try this server's own
POST /get-test-token first, falling back to GN4_TEST_AUTH_TOKEN --
every previously-ungated test file gains a token check and switches to
get_with_token; the whole suite, both languages, now runs with zero
manually-provisioned tokens given GENENETWORK4_DEV_MODE=true
GENENETWORK4_BIND=127.0.0.1 for the test run. ServerTestCase gains
EXTRA_ENV so each Python test class can start its own server with dev
mode on. Verified both configured (auto-fetch) and unconfigured
(GN4_TEST_AUTH_TOKEN fallback, and full skip when neither is set, e.g.
the Guix build sandbox) paths.
README.md, docs/gn-auth.md and RELEASE_NOTES.md updated throughout for
the rename and the new blanket token requirement.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XDDpXvZBQat9uSgDeEMJ8d
|
|
* src/db.rs (TraitInfo): New struct. Joins PublishXRef, Phenotype and
Publication; LEFT JOIN on the latter two since a handful of
PublishXRef rows have no matching row. SQL column aliases are written
to match mysql_common's UpperCamelCase conversion exactly, so no field
needs a #[mysql(rename = "...")] override.
(get_group_trait_info): New function.
(pool): New function. mysql::Pool::new eagerly opens its default 10
connections before returning (about 7s against this database); every
query function here used to call it per request, paying that cost on
every single API call. Builds the Pool once behind a
OnceLock<Mutex<Option<Pool>>> and shares the cheap, Arc-backed handle
from then on (OnceLock::get_or_init can't be used alone since
Pool::new is fallible and get_or_try_init is still unstable).
(get_all_species, get_species_by_name, get_all_groups,
get_group_by_name, get_group_strains, get_group_trait_values,
get_group_trait_info): Use it instead of building their own Pool.
* src/lib.rs (GroupTraitLookup): Generalize over the found value (was
Vec<db::TraitValue>-specific), so group_trait_info can share it with
group_trait_values.
(get_group_trait_info_result): New function.
(group_trait_info): New handler for GET /groups/{group}/traits/{trait_id}.
(app): Register it.
* tests/common/mod.rs: New file. get() issues a GET against
genenetwork4::app() and warns on stderr if it takes over a second --
this is what surfaced the Pool::new cost above.
* tests/species.rs, tests/species_by_name.rs, tests/groups.rs,
tests/group_by_name.rs, tests/group_strains.rs,
tests/group_trait_values.rs: Use tests/common::get instead of building
each request by hand.
* tests/group_trait_info.rs, test/python/test_group_trait_info.py: New
files.
* tests/openapi.rs, test/python/test_openapi.py: Check the new path is
in the generated spec.
* README.md (GET /groups/{group}/traits/{trait_id}): Document it.
(Tests): List the new test files; document tests/common::get's timing
warnings and how to see them.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XDDpXvZBQat9uSgDeEMJ8d
|
|
Line-wrapping and canonical import ordering only (verified with
git diff -w). Establishes rustfmt's output as the canonical style
before the search/correlation work lands.
|
|
The old skip check connected to 127.0.0.1:3306, so tests ran (and
failed with 500) whenever a server was listening but the configured
credentials were rejected. tests/common/mod.rs now attempts a real
connection with the configured SQL_URI; tests skip only when the
database is genuinely unusable, and run for real otherwise.
|
|
* src/db.rs (Strain): New struct.
(get_group_strains): New function. Joins Strain and StrainXRef by
group Id, optionally filtered to strains used for mapping, matching
gn-guile's gn.data.strains:strain-id-names (same InbredSetId and
used-for-mapping? arguments), returning strain names as well as ids.
* src/lib.rs (GroupStrainsParams): New struct, the used_for_mapping
query parameter.
(group_strains): New handler for GET /groups/{group}/strains. Looks
the group up the same way group_by_name does, for the same 404 shape
on an unknown group.
(app): Register it.
* tests/group_strains.rs, test/python/test_group_strains.py: New
files.
* tests/openapi.rs, test/python/test_openapi.py: Check the new path
is in the generated spec.
* README.md (GET /groups/{group}/strains): Document it, including the
used_for_mapping filter excluding BXD's parent strains.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XDDpXvZBQat9uSgDeEMJ8d
|