golangLAKEHOUSE/cmd/ingestd/main_test.go

package main

import (
	"bytes"
	"io"
	"net/http"
	"net/http/httptest"
	"strings"
	"testing"

	"github.com/go-chi/chi/v5"

	"git.agentview.dev/profit/golangLAKEHOUSE/internal/catalogclient"
)

// Closes R-005 for ingestd: cmd-level tests for the cmd-shape
// of /ingest — name query param, body cap, multipart parsing,
// missing form file. CSV→Parquet logic is tested in internal/ingestd.

func newTestHandlers(t *testing.T) (*handlers, *httptest.Server) {
	t.Helper()
	// Stub catalogd so we can run end-to-end happy paths without the
	// real catalogd up. The stub returns a 200-shaped registerResponse.
	stub := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
		switch r.URL.Path {
		case "/catalog/register":
			w.Header().Set("Content-Type", "application/json")
			w.Write([]byte(`{"manifest":{"name":"x","dataset_id":"d","schema_fingerprint":"sha256:x","objects":[]},"existing":false}`))
		default:
			w.WriteHeader(http.StatusNotFound)
		}
	}))
	t.Cleanup(stub.Close)

	storaged := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
		w.WriteHeader(http.StatusOK)
		w.Write([]byte(`{"status":"ok"}`))
	}))
	t.Cleanup(storaged.Close)

	h := &handlers{
		storagedURL: strings.TrimRight(storaged.URL, "/"),
		catalogd:    catalogclient.New(stub.URL),
		hc:          &http.Client{},
		maxBytes:    256 << 20,
	}
	return h, stub
}

func mountedRouter(h *handlers) chi.Router {
	r := chi.NewRouter()
	h.register(r)
	return r
}

func TestRoutesMounted(t *testing.T) {
	h, _ := newTestHandlers(t)
	r := mountedRouter(h)
	found := false
	chi.Walk(r, func(method, route string, _ http.Handler, _ ...func(http.Handler) http.Handler) error {
		if method == "POST" && route == "/ingest" {
			found = true
		}
		return nil
	})
	if !found {
		t.Error("POST /ingest not mounted")
	}
}

func TestHandleIngest_MissingNameQueryParam(t *testing.T) {
	h, _ := newTestHandlers(t)
	r := mountedRouter(h)
	srv := httptest.NewServer(r)
	defer srv.Close()

	resp, err := http.Post(srv.URL+"/ingest",
		"multipart/form-data; boundary=x", strings.NewReader(""))
	if err != nil {
		t.Fatalf("POST: %v", err)
	}
	defer resp.Body.Close()
	if resp.StatusCode != http.StatusBadRequest {
		t.Errorf("expected 400 on missing name param, got %d", resp.StatusCode)
	}
	body, _ := io.ReadAll(resp.Body)
	if !strings.Contains(string(body), "name") {
		t.Errorf("error body should mention 'name', got %q", body)
	}
}

func TestHandleIngest_MalformedMultipart(t *testing.T) {
	h, _ := newTestHandlers(t)
	r := mountedRouter(h)
	srv := httptest.NewServer(r)
	defer srv.Close()

	resp, err := http.Post(srv.URL+"/ingest?name=test",
		"multipart/form-data; boundary=xyz", strings.NewReader("garbage not multipart"))
	if err != nil {
		t.Fatalf("POST: %v", err)
	}
	defer resp.Body.Close()
	if resp.StatusCode != http.StatusBadRequest {
		t.Errorf("expected 400 on malformed multipart, got %d", resp.StatusCode)
	}
}

func TestHandleIngest_MissingFormFile(t *testing.T) {
	h, _ := newTestHandlers(t)
	r := mountedRouter(h)
	srv := httptest.NewServer(r)
	defer srv.Close()

	// Valid multipart with no "file" field.
	body := bytes.NewReader([]byte(
		"--xyz\r\n" +
			"Content-Disposition: form-data; name=\"other\"\r\n" +
			"\r\n" +
			"value\r\n" +
			"--xyz--\r\n",
	))
	resp, err := http.Post(srv.URL+"/ingest?name=test",
		"multipart/form-data; boundary=xyz", body)
	if err != nil {
		t.Fatalf("POST: %v", err)
	}
	defer resp.Body.Close()
	if resp.StatusCode != http.StatusBadRequest {
		t.Errorf("expected 400 on missing form file, got %d", resp.StatusCode)
	}
}

func TestHandleIngest_BodyTooLarge(t *testing.T) {
	h, _ := newTestHandlers(t)
	h.maxBytes = 1024 // tiny cap so we hit it without huge upload
	r := mountedRouter(h)
	srv := httptest.NewServer(r)
	defer srv.Close()

	big := bytes.Repeat([]byte("x"), 4096)
	resp, err := http.Post(srv.URL+"/ingest?name=test",
		"multipart/form-data; boundary=xyz", bytes.NewReader(big))
	if err != nil {
		t.Fatalf("POST: %v", err)
	}
	defer resp.Body.Close()
	if resp.StatusCode < 400 || resp.StatusCode >= 500 {
		t.Errorf("expected 4xx on oversize body, got %d", resp.StatusCode)
	}
}

func TestEscapeKeyPath(t *testing.T) {
	cases := []struct {
		in   string
		want string
	}{
		{"a/b/c.parquet", "a/b/c.parquet"},
		{"data sets/x.parquet", "data%20sets/x.parquet"},
		{"O'Reilly/key", "O%27Reilly/key"},
		{"datasets/proof/abc.parquet", "datasets/proof/abc.parquet"},
		{"", ""},
		{"/", "/"},
	}
	for _, tc := range cases {
		t.Run(tc.in, func(t *testing.T) {
			got := escapeKeyPath(tc.in)
			if got != tc.want {
				t.Errorf("escapeKeyPath(%q) = %q, want %q", tc.in, got, tc.want)
			}
		})
	}
}

func TestParquetKeyPath_Format(t *testing.T) {
	// Lock the content-addressed key shape per scrum C-DRIFT.
	// Failure here means a dataset's parquet would land at an
	// unexpected key, breaking schema-drift idempotency.
	if !strings.Contains(parquetKeyPath, "%s") {
		t.Errorf("parquetKeyPath should be a fmt template, got %q", parquetKeyPath)
	}
	if !strings.HasPrefix(parquetKeyPath, "datasets/") {
		t.Errorf("parquetKeyPath should be under datasets/, got %q", parquetKeyPath)
	}
	if !strings.HasSuffix(parquetKeyPath, ".parquet") {
		t.Errorf("parquetKeyPath should end with .parquet, got %q", parquetKeyPath)
	}
}