refactor(P2): extract ingestion-summary renderer into reporting.ConsoleRenderer

[LukasWodka]

Summary

Structural refactor — phase P2 (epic backend#796): pull human-facing presentation out of the ingestion logic. This is the first structural phase, now unblocked since #242–#245 merged.

The ingestion-summary box (success rate, per-channel counts, status banner) was ~114 lines of print + ANSI + emoji living inside the 1,196-line BaseIngestor god class.

Changes

• New tracebloc_ingestor/reporting.py with ConsoleRenderer.render_summary — the single home for that presentation. The body moved verbatim, so the customer-facing output is byte-for-byte identical.
• BaseIngestor._log_summary becomes a thin delegate (kept as a method so existing callers/tests that reference BaseIngestor._log_summary keep working). Net: base.py −119 / +10.
• Drops the now-unused BLUE import from base.py.
• IngestionSummary stays the pure data object the renderer consumes; the renderer type-hints it via TYPE_CHECKING to avoid a runtime import cycle.

Payoff

The presentation is now unit-testable without a DB or a full ingest — tests/test_reporting.py (8 cases, reporting.py 100% covered) locks the output contract, including #234's behavior that dropped records count toward the failure total and disqualify the success banner.

Behaviour preservation (the gate)

• Summary code moved verbatim; output identical.
• Full unit suite: 1026 passed, 1 xfailed, coverage 97.3%.
• The characterization harness goldens (#247: DB cells / DEST_PATH manifest / backend payloads) are unaffected — they don't assert the summary box. The CI e2e job re-confirms against real MySQL (Docker was unavailable locally this run, so the renderer was instead verified directly via the new unit tests; the data path is untouched).

Scope note

This is P2 step 1 — the largest single presentation block. The remainder of P2 (strip ANSI from raised exception messages; print → logging in validators) is a focused follow-up, kept separate so each PR is small and reviewable. No conflict with any open work.

🤖 Generated with Claude Code

[LukasWodka]

👋 Heads-up — Code review queue is at 14 / 8

Above the WIP limit. The team convention is to review existing PRs before opening new work.

Open PRs currently in Code review (oldest first):

• averaging-service#106 — test(integration): pin XGBoost/LightGBM/CatBoost end-to-end averaging round-trips · author: @LukasWodka · no reviewer assigned
• averaging-service#111 — fix(averaging): correctness round-2 sweep — N5/N7/N8/N9/N10 + load_model except (Sync develop → master for v0.3.0 release #109) · author: @LukasWodka · no reviewer assigned
• client#254 — Sync develop → main for v1.7.1 chart release (egress-enforcement preflight helm test, inert) · author: @saadqbal · no reviewer assigned
• client-runtime#105 — chore(ci): drop unused mysqlclient dep; report test coverage (Phase 0) · author: @LukasWodka · no reviewer assigned
• client-runtime#106 — fix(schema): sync vendored ingest.v1.json — restore token_classification · author: @LukasWodka · no reviewer assigned
• design-system#47 — feat(table): Table compound organism (Figma 959:6061 + 986:9822) · author: @waqaskhanroghani · no reviewer assigned
• tracebloc-py-package#217 — fix(sdk): accept NumPy integer counts; framework-aware custom-loss header · author: @divyasinghds · no reviewer assigned
• tracebloc-py-package#218 — fix(upload): output-shape validation + hard-fail class-count check for text_classification · author: @shujaatTracebloc · reviewer: @aptracebloc
• tracebloc-py-package#219 — fix(plan): reject non-dict training_classes input cleanly · author: @divyasinghds · no reviewer assigned
• tracebloc-py-package#220 — fix(plan): recover from failed image reform; roll back shape/type · author: @divyasinghds · no reviewer assigned

Pull from review before opening new work. (This is a nudge from the kanban WIP check, not a block.)