homelab-codex-ws/services/kb-query/app/links.py

73 lines
2.8 KiB
Python
Raw Normal View History

"""Per-source result shaping for `/search` -- module 5 phase 4 (kb/phases/kb-m5-faza4.md,
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
§4 response shape, §6/decision 3 links). `kb_retrieval.retrieval` only ever touches
`document_chunk`/`document_summary`; the join to `envelope` (for `source` and link/metadata
fields) is new HTTP-layer code that belongs to kb-query, not the shared retrieval package.
Paperless link format verified live 2026-07-22: `curl -I https://paper.kapala.org/documents/1/details`
-> 302 to `/accounts/login/?next=/documents/1/details` (unauthenticated, but confirms the route
exists and is handled by the Angular app, not a 404) -- the plan's candidate format.
Gmail has nothing to link to (plan §2 decision 3: `envelope_id` **is** the Message-ID already,
nothing to extract) -- `mail_ui_url` stays a reserved `null` until kb-00's future mail-UI.
"""
from __future__ import annotations
from typing import Optional
PAPERLESS_LINK_TEMPLATE = "https://paper.kapala.org/documents/{doc_id}/details"
def _extract_headers(entities: list) -> dict:
for entity in entities or []:
if isinstance(entity, dict) and entity.get("type") == "headers":
return entity
return {}
def _format_from(headers: dict) -> Optional[str]:
frm = headers.get("from")
if not isinstance(frm, dict):
return None
name = frm.get("name")
address = frm.get("address")
if name and address:
return f"{name} <{address}>"
return address or name
def _paperless_link(envelope_id: str) -> str:
# envelope_id is "paperless:<id>" (paperless_adapter.py:126) -- the raw Paperless doc id
# is everything after the first colon.
doc_id = envelope_id.split(":", 1)[1] if ":" in envelope_id else envelope_id
return PAPERLESS_LINK_TEMPLATE.format(doc_id=doc_id)
def build_result(chunk: dict, envelope: Optional[dict]) -> dict:
"""One `/search` result item: retrieval hit (`chunk`) + its envelope metadata (`envelope`,
`None` if the envelope vanished between chunk-embedding and this query -- shouldn't happen
on a read-only corpus, but the join must not crash on it)."""
source = (envelope or {}).get("source", "unknown")
result = {
"envelope_id": chunk["envelope_id"],
"source": source,
"dist": chunk["dist"],
"chunk_index": chunk["chunk_index"],
"text": chunk["text"],
}
if source == "paperless":
result["link"] = _paperless_link(chunk["envelope_id"])
return result
if source == "gmail":
headers = _extract_headers((envelope or {}).get("entities", []))
result["subject"] = headers.get("subject")
result["from"] = _format_from(headers)
result["date"] = headers.get("date_raw")
result["link"] = None
result["mail_ui_url"] = None
return result
result["link"] = None
return result