Reject malformed search bodies and report unranked candidates
CI / ci (push) Successful in 1m35s
CI / ci (pull_request) Successful in 1m35s

SearXNG now requires a results array and Brave requires "type": "search"
with a results array in any web object, so `{}` is ErrSearchResponse
instead of an empty search. Rank no longer drops failed candidates: it
returns the ranked ones with ErrCandidatesUnranked joined with each
unranked candidate's URL and reason, or ErrNoUsableRecipe when none rank.
Adds an HTTPS ranking test through the SSRF-safe Fetcher.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
2026-10-05 13:27:07 -04:00
co-authored by Claude Opus 5.5
parent 26928b538b
commit 01523f634b
5 changed files with 177 additions and 22 deletions
+98 -2
View File
@@ -4,6 +4,8 @@ import (
"context"
"errors"
"fmt"
"net/http"
"strings"
"sync"
"testing"
)
@@ -66,8 +68,11 @@ func TestRankOrder(t *testing.T) {
{URL: "https://f.example/toodeep"},
}
got, err := Rank(context.Background(), f, "Chicken curry", cands)
if err != nil {
t.Fatal(err)
if !errors.Is(err, ErrCandidatesUnranked) || !errors.Is(err, ErrNoImage) || !errors.Is(err, ErrNoIngredients) {
t.Errorf("err = %v, want ErrCandidatesUnranked with ErrNoImage and ErrNoIngredients", err)
}
if errors.Is(err, ErrNoUsableRecipe) {
t.Errorf("err = %v, want not ErrNoUsableRecipe when some are ranked", err)
}
want := []struct {
url string
@@ -119,6 +124,29 @@ func TestRankTiesKeepSearchOrder(t *testing.T) {
}
}
// TestRankReportsFetchFailure checks a failed fetch is reported with the
// candidates that were ranked, never swallowed.
func TestRankReportsFetchFailure(t *testing.T) {
f := &fakeFetcher{pages: map[string][]byte{
"https://a.example/1": recipePage("Beef Stew", true, true, true, true),
}}
cands := []Candidate{{URL: "https://a.example/1"}, {URL: "https://b.example/missing"}}
got, err := Rank(context.Background(), f, "beef stew", cands)
if len(got) != 1 || got[0].Candidate.URL != "https://a.example/1" {
t.Errorf("ranked = %+v, want only https://a.example/1", got)
}
if !errors.Is(err, ErrCandidatesUnranked) {
t.Errorf("err = %v, want ErrCandidatesUnranked", err)
}
var se *StatusError
if !errors.As(err, &se) || se.Status != 404 || se.URL != "https://b.example/missing" {
t.Errorf("err = %v, want the 404 StatusError for https://b.example/missing", err)
}
if err != nil && !strings.Contains(err.Error(), "https://b.example/missing") {
t.Errorf("err = %v, want it to name the failed candidate", err)
}
}
func TestRankAllRejected(t *testing.T) {
f := &fakeFetcher{pages: map[string][]byte{
"https://a.example/noimg": recipePage("Chicken Curry", true, false, true, true),
@@ -157,3 +185,71 @@ func TestRankContextCancelled(t *testing.T) {
t.Errorf("err = %v, want context.Canceled", err)
}
}
// TestRankOverHTTP ranks fixture candidates served over HTTPS through the
// SSRF-safe Fetcher: every host, a subdomain the test server's certificate
// covers, resolves to a public test address the test
// dialer routes to one httptest server, which serves pages by host and path.
func TestRankOverHTTP(t *testing.T) {
t.Parallel()
pages := map[string][]byte{
"best.example.com/curry": recipePage("Chicken Curry", true, true, true, true),
"easy.example.com/curry": recipePage("Easy Chicken Curry", true, true, true, false),
"thai.example.com/curry": recipePage("Thai Green Curry", true, true, true, true),
"noimg.example.com/curry": recipePage("Chicken Curry", true, false, true, true),
}
env := newTestEnv(t, http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
body, ok := pages[r.Host+r.URL.Path]
if !ok {
http.Error(w, "broken", http.StatusInternalServerError)
return
}
w.Header().Set("Content-Type", "text/html")
w.Write(body)
}), map[string][][]string{
"best.example.com": {{publicIP}},
"easy.example.com": {{publicIP}},
"thai.example.com": {{publicIP}},
"noimg.example.com": {{publicIP}},
"broken.example.com": {{publicIP}},
})
cands := []Candidate{
{URL: "https://thai.example.com/curry"},
{URL: "https://broken.example.com/curry"},
{URL: "https://easy.example.com/curry"},
{URL: "https://noimg.example.com/curry"},
{URL: "https://best.example.com/curry"},
}
got, err := Rank(t.Context(), env.f, "Chicken curry", cands)
want := []struct {
url string
score float64
}{
{"https://best.example.com/curry", 2},
{"https://easy.example.com/curry", 1.75},
{"https://thai.example.com/curry", 1.25},
}
if len(got) != len(want) {
t.Fatalf("got %d ranked, want %d: %+v (err %v)", len(got), len(want), got, err)
}
for i, w := range want {
if got[i].Candidate.URL != w.url || got[i].Score != w.score {
t.Errorf("rank %d = %s (%.2f), want %s (%.2f)", i, got[i].Candidate.URL, got[i].Score, w.url, w.score)
}
}
if got[0].Recipe.Title != "Chicken Curry" || got[0].Recipe.ImageURL != "https://img.example.com/dish.jpg" {
t.Errorf("top recipe = %+v", got[0].Recipe)
}
if !errors.Is(err, ErrCandidatesUnranked) || !errors.Is(err, ErrNoImage) {
t.Errorf("err = %v, want ErrCandidatesUnranked with ErrNoImage", err)
}
var se *StatusError
if !errors.As(err, &se) || se.Status != http.StatusInternalServerError {
t.Errorf("err = %v, want the 500 StatusError for broken.example.com", err)
}
if n := len(env.dials()); n != len(cands) {
t.Errorf("dialed %d times, want %d", n, len(cands))
}
}