fix(flask): align http.server.active_requests metric with semconv helper

[rite7sh]

Description

This PR fixes inconsistent initialization of the http.server.active_requests
metric in Flask instrumentation by aligning it with the semantic conventions
helper used by other instrumentations.

Depending on how Flask is instrumented, different units and descriptions were
being used for the same metric. This change ensures consistent behavior by
using the semconv helper everywhere.

Fixes #4093

Type of change

• Bug fix (non-breaking change which fixes an issue)

How Has This Been Tested?

• python -m ruff check instrumentation/opentelemetry-instrumentation-flask
• Existing Flask instrumentation tests
• Verified no lint or formatting issues after changes

Does This PR Require a Core Repo Change?

• No.

Checklist:

• Followed the style guidelines of this project
• Changelogs have been updated (not required for this fix)
• Unit tests have been added (existing coverage is sufficient)
• Documentation has been updated (no user-facing changes)

[JWinermaSplunk]

Left a small comment, but it looks simple enough if the helper functions suffice. Also, this change probably requires a changelog entry, otherwise LGTM.

[rite7sh]

@JWinermaSplunk I’ve pushed a small follow-up commit that replaces the hardcoded legacy metric name with a constant, keeping it consistent with the existing semantic convention helpers.
Please let me know if this approach looks good to you, and whether you’d like me to add a changelog entry as well, given that this doesn’t introduce any behavioral changes.

[JWinermaSplunk]

A changelog entry never hurts. It is simple enough and even the current PR name would be fine as an entry, but it isn't up to me in the end :)

[rite7sh]

@JWinermaSplunk you got a point, i surely will add a changelog update.
thanks for the suggestion.

[xrmx]

The issue here is that the two instrumentor methods we have in this instrumentation use two different semconv for the same metric: one as old as the default implementation and one that does follow the stable one. The PR is changing the older one to use the stable semconv too so both in the default older semconv and when output of the stable one is configured. That's a behaviour change. Since the difference is only on unit and description I'm not sure that's a big deal.

[rite7sh]

@xrmx Thanks for the clarification, I agree this is a small behavior change limited to unit and description, and I’m glad it aligns both paths on the same semconv, I’ve updated the branch and am happy to tweak anything if needed. :)

[xrmx]

Can we do something like

if newsemconv:
   active_requests_counter = create_http_server_active_requests(meter)
else:
   active_requests_counter = meter.create_up_down_counter(
            name=MetricInstruments.HTTP_SERVER_ACTIVE_REQUESTS,
            unit="requests",
            description="measures the number of concurrent HTTP requests that are currently in-flight",
        )

in both hunks so that the metrics matches the semconv?

[rite7sh]

@xrmx I’ve updated both paths to use the explicit semconv split as suggested. Let me know if this looks good now, thanks for suggestion.