feat(streaming): add shared stream topic constants

[BaoT1301]

IMPORTANT:

• PR must be opened from your personal branch -> dev
• You must tag @Team-Deepiri/support-team
• You must update Plaky to "Needs QA"
• Never move a feature/bug to "Done" (Done = production release only)

---

Description

Adds shared Python stream topic constants for the current Deepiri data-streaming contract so Python services can reference the same document.* and Cyrex-to-Helox training stream names without duplicating string literals.

Include:

• Related Issue number: N/A
• Plaky feature name: Deepiri data streaming / shared topic constants
• Component, feature, or system affected: deepiri-modelkit streaming topic registry
• Purpose of change: shared contract improvement

---

Changes

• Added document-routing topic constants:
  • DOCUMENT_VECTORIZE = "document.vectorize"
  • DOCUMENT_TRAINING = "document.training"
  • DOCUMENT_STRUCTURED = "document.structured"
  • DOCUMENT_ARTIFACTS = "document.artifacts"
• Added Cyrex-to-Helox training topic constants:
  • HELOX_TRAINING_RAW = "pipeline.helox-training.raw"
  • HELOX_TRAINING_STRUCTURED = "pipeline.helox-training.structured"
• Kept schema registration unchanged because LIS-owned document event schema classes do not exist yet.
• Merged the branch with dev so the PR includes the latest shared sidecar utility changes.

---

Related

• Issue: N/A
• Plaky: Deepiri data streaming / shared topic constants
• Related PRs: fix(streaming): harden Redis ACK handling and document topics deepiri-shared-utils#7, feat(pipeline): add durable Postgres producer path for Helox training diri-cyrex#79, feat(training): harden data ingestion and pipeline robustness diri-helox#54, feat(streaming): Redis Streams integration for real-time data ingestion deepiri-platform#184

---

Testing

Verified locally:

• python -m py_compile src/deepiri_modelkit/streaming/topics.py
• git diff --check

Additional testing details:

• This is a constants-only contract PR; document event schemas are intentionally left for the LIS producer/router implementation.

---

Important Notes (Optional)

• Known limitations: Does not add document event schema classes or update TOPIC_EVENT_SCHEMAS.
• Blockers: Plaky still needs to be moved to "Needs QA" externally.
• CI/CD issues unrelated to this PR: None known.
• Dependencies required for testing: Python 3.11 or compatible local Python.

---

Workflow Checklist (Required)

• Branch is up to date with dev
• PR is from your branch -> dev (no longer directly into main)
• PR title follows convention (feat:, fix:, refactor:, etc.)
• Plaky feature/bug name included above
• Tagged @Team-Deepiri/support-team
• Plaky feature moved to "Needs QA"

---

Review Requests

@Team-Deepiri/support-team

[jrb00013]

@BaoT1301 what's this for?

[jrb00013]

@BaoT1301 what’s this PR for?

[BaoT1301]

Hey @jrb00013 , this is the for Python-side shared constants, the stream names already exist in the plan and are currently used as local string literals in Cyrex/Helox. This PR puts those names into deepiri-modelkit so Python services can import the same contract instead of duplicating strings. It is not implementing routing, schemas or ingestion. It is just the shared Python topic registry, matching what shared-utils does for TypeScript.