root c1d96b7b60 matrixd: multi-corpus retrieve+merge — SPEC §3.4 component 2 of 5
Lands the matrix indexer's first piece per docs/SPEC.md §3.4:
multi-corpus retrieve+merge with corpus attribution per result.
Future components (relevance filter, downgrade gate, learning-loop
integration) layer on top of this surface.

Architecture:
  - internal/matrix/retrieve.go — Retriever takes (query, corpora,
    k, per_corpus_k), parallel-fans across vectord indexes, merges
    by distance ascending, preserves corpus origin per hit
  - cmd/matrixd — HTTP service on :3217, fronts /v1/matrix/*
  - gateway proxy + [matrixd] config + lakehouse.toml entry
  - Either query_text (matrix calls embedd) or query_vector
    (caller pre-embedded) — vector takes precedence if both set

Error policy: fail-loud on any corpus error. Silent partial returns
would lie about coverage, defeating the matrix's whole purpose.
Bubbles vectord errors as 502 (upstream), validation as 400.

Smoke (scripts/matrix_smoke.sh, 6 assertions PASS first try):
  - /matrix/corpora lists indexes
  - Multi-corpus search returns hits from BOTH corpora
  - Top hit is the globally-closest across all corpora
    (b-near beats a-near at distance 0.05 vs 0.1 — proves merge)
  - Metadata round-trips through the merge
  - Distances ascending in result list
  - Negative paths: empty corpora → 400, missing corpus → 502,
    no query → 400

12-smoke regression sweep all green (D1-D6, G1, G1P, G2,
storaged_cap, pathway, matrix).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-29 18:39:17 -05:00

126 lines
3.6 KiB
Go

// matrixd is the matrix indexer service. Wraps internal/matrix's
// Retriever with HTTP routes per docs/SPEC.md §3.4.
//
// Routes:
// POST /matrix/search — multi-corpus retrieve+merge
// GET /matrix/corpora — list known vectord indexes (proxy)
//
// matrixd talks to embedd (for query-text embedding) and vectord
// (for per-corpus search) via HTTP. Both URLs come from
// [matrixd] config; gateway sets them to its own upstream URLs so
// matrixd inherits the same provider topology.
package main
import (
"encoding/json"
"errors"
"flag"
"log/slog"
"net/http"
"os"
"strings"
"github.com/go-chi/chi/v5"
"git.agentview.dev/profit/golangLAKEHOUSE/internal/matrix"
"git.agentview.dev/profit/golangLAKEHOUSE/internal/shared"
)
const maxRequestBytes = 4 << 20 // 4 MiB cap on request bodies
func main() {
configPath := flag.String("config", "lakehouse.toml", "path to TOML config")
flag.Parse()
cfg, err := shared.LoadConfig(*configPath)
if err != nil {
slog.Error("config", "err", err)
os.Exit(1)
}
if cfg.Matrixd.EmbeddURL == "" || cfg.Matrixd.VectordURL == "" {
slog.Error("matrixd: embedd_url and vectord_url required in [matrixd]")
os.Exit(1)
}
retriever := matrix.New(cfg.Matrixd.EmbeddURL, cfg.Matrixd.VectordURL)
h := &handlers{r: retriever}
if err := shared.Run("matrixd", cfg.Matrixd.Bind, h.register, cfg.Auth); err != nil {
slog.Error("server", "err", err)
os.Exit(1)
}
}
type handlers struct {
r *matrix.Retriever
}
func (h *handlers) register(r chi.Router) {
r.Post("/matrix/search", h.handleSearch)
r.Get("/matrix/corpora", h.handleCorpora)
}
func (h *handlers) handleSearch(w http.ResponseWriter, r *http.Request) {
var req matrix.SearchRequest
if !decodeJSON(w, r, &req) {
return
}
resp, err := h.r.Search(r.Context(), req)
if err != nil {
writeMatrixError(w, err)
return
}
writeJSON(w, http.StatusOK, resp)
}
func (h *handlers) handleCorpora(w http.ResponseWriter, r *http.Request) {
names, err := h.r.Corpora(r.Context())
if err != nil {
slog.Error("matrix corpora", "err", err)
http.Error(w, "vectord unavailable", http.StatusBadGateway)
return
}
writeJSON(w, http.StatusOK, map[string]any{"corpora": names, "count": len(names)})
}
func decodeJSON(w http.ResponseWriter, r *http.Request, v any) bool {
defer r.Body.Close()
r.Body = http.MaxBytesReader(w, r.Body, maxRequestBytes)
if err := json.NewDecoder(r.Body).Decode(v); err != nil {
var maxErr *http.MaxBytesError
if errors.As(err, &maxErr) || strings.Contains(err.Error(), "http: request body too large") {
http.Error(w, "body too large", http.StatusRequestEntityTooLarge)
return false
}
http.Error(w, "decode body: "+err.Error(), http.StatusBadRequest)
return false
}
return true
}
func writeJSON(w http.ResponseWriter, code int, v any) {
w.Header().Set("Content-Type", "application/json")
w.WriteHeader(code)
if err := json.NewEncoder(w).Encode(v); err != nil {
slog.Warn("matrix write json", "err", err)
}
}
// writeMatrixError maps internal/matrix sentinels to HTTP statuses.
// Corpus / embed failures bubble up as 502 (the upstream service is
// what's wrong); validation errors are 400.
func writeMatrixError(w http.ResponseWriter, err error) {
switch {
case errors.Is(err, matrix.ErrEmptyCorpora),
errors.Is(err, matrix.ErrEmptyQuery):
http.Error(w, err.Error(), http.StatusBadRequest)
case errors.Is(err, matrix.ErrCorpus),
errors.Is(err, matrix.ErrEmbed):
slog.Warn("matrix upstream", "err", err)
http.Error(w, err.Error(), http.StatusBadGateway)
default:
slog.Error("matrix", "err", err)
http.Error(w, "internal", http.StatusInternalServerError)
}
}