fix(test): stop pinning the tracer scope name to a development build

[aryanmehrotra]

Description:

Test_TracerContract_InstrumentationScopeName asserts the instrumentation scope name equals the literal "gofr-dev". The middleware builds that name as "gofr-" + version.Framework (tracer.go:64), and version.Framework is "dev" only on a development build — a release branch sets it to the version being cut.

So the test passes on development and fails on every release branch:

    tracer_test.go:436
    Error:      Not equal:
                expected: "gofr-dev"
                actual  : "gofr-v1.60.0"
--- FAIL: Test_TracerContract_InstrumentationScopeName (0.00s)
FAIL	gofr.dev/pkg/gofr/http/middleware

Found while cutting v1.60.0 (#4107), where it failed all three PKG Unit Testing shards — v1.24, v1.25 and v1.26. It is the only failure in ./pkg/....

This is a same-cycle regression, not inherited debt. The assertion arrived with #3770 (7cbb41a3), which is itself part of the v1.60.0 release:

	v1.59.0	v1.60.0 (development)
"gofr-dev" literal in tracer_test.go	absent	present
version.Framework referenced by that test	never	yes

tracer_test.go predates this cycle, but nothing in it depended on the version until #3770, so v1.59.0's release branch could not have hit this. v1.60.0 is the first release that would fail — and so would every release after it, if this shipped as is. Fixing it at the source rather than patching the release branch keeps the next cut clean.

Why the literal was never load-bearing

The two assertions were:

assert.Equal(t, "gofr-dev", tracerCharScopeName(),
    "version.Framework changed; the scope name below moves with it")
assert.Equal(t, tracerCharScopeName(), spans[0].InstrumentationScope().Name)

with tracerCharScopeName() defined as "gofr-" + version.Framework. The second line already pinned the contract — the span's scope name is the gofr prefix plus the framework version. The first only asserted that this particular build happens to be a development build, which is a property of the branch, not of the middleware.

Its comment shows the intent was a tripwire for the name moving. But the name is designed to move with every release, so the tripwire fires on exactly the occasions when nothing is wrong.

The change

The expected value is derived the way production derives it, and a second assertion pins that the version is actually appended rather than resolving to an empty string — the one failure mode a derived expectation could otherwise hide:

assert.Equal(t, "gofr-"+version.Framework, spans[0].InstrumentationScope().Name,
    "scope name is the gofr- prefix plus version.Framework")
assert.NotEqual(t, "gofr-", spans[0].InstrumentationScope().Name,
    "the version must actually be appended, not an empty string")

tracerCharScopeName is removed; it had no other caller.

Breaking Changes (if applicable):

None. Test-only change; no production file is touched.

Additional Information:

The assertion was checked in both directions rather than assumed to work.

Mutation-verified against the production expression in tracer.go:

production expression	result
"gofrx-" + version.Framework	FAIL
"gofr-" + ""	FAIL
unmodified	PASS

(A third mutation — deleting version.Framework outright — was discarded because it leaves the import unused and fails to build, so it tests nothing.)

Verified under both builds, which is the point of the change:

• Framework = "dev" → go test ./pkg/gofr/http/middleware/ -run Test_TracerContract green.
• Framework = "v1.60.0" → green (fails on development today).

go vet ./pkg/gofr/http/middleware/ is clean.

golangci-lint run ./pkg/gofr/http/middleware/... reports 5 findings, none in the changed file — 2 goconst in logger.go and 3 noctx in apikey_auth_test.go / auth_test.go, all pre-existing in files this PR does not touch (the diff is tracer_test.go alone, +8/-7). CI runs with only-new-issues, and Code Quality passes.

On the full local go test ./pkg/...: one package fails on my machine, pkg/gofr/datasource/pubsub/mqtt, which hangs to a test timed out panic at both the default and a 5m timeout. It talks to a public MQTT broker, this PR does not touch it (the diff is tracer_test.go alone), and it passes in CI. I am citing CI as the authority rather than my own run: all three PKG Unit Testing shards pass on this branch, which is precisely the check that was red on the release branch.

Checklist:

• I have formatted my code using goimport and golangci-lint.
• All new code is covered by unit tests. — the change is the test; verified by mutation in both directions
• This PR does not decrease the overall code coverage.
• I have reviewed the code comments and documentation for clarity.