9273f356e6
- BuildQueries(reg, source) dedups keywords and formats per-source syntax - github/gist use 'keyword' in:file; others use bare keyword - SourcesConfig placeholder struct for Wave 2 plans to depend on - RegisterAll no-op stub (Plan 10-09 will fill)
125 lines
3.2 KiB
Go
125 lines
3.2 KiB
Go
package sources
|
|
|
|
import (
|
|
"reflect"
|
|
"testing"
|
|
|
|
"github.com/salvacybersec/keyhunter/pkg/providers"
|
|
)
|
|
|
|
func testRegistry() *providers.Registry {
|
|
return providers.NewRegistryFromProviders([]providers.Provider{
|
|
{
|
|
Name: "openai",
|
|
Keywords: []string{"sk-proj-", "sk-"},
|
|
},
|
|
{
|
|
Name: "anthropic",
|
|
Keywords: []string{"sk-ant-", "sk-"}, // shared "sk-" for dedup test
|
|
},
|
|
{
|
|
Name: "empty-kw",
|
|
Keywords: []string{},
|
|
},
|
|
})
|
|
}
|
|
|
|
func TestBuildQueries_GitHub(t *testing.T) {
|
|
got := BuildQueries(testRegistry(), "github")
|
|
want := []string{
|
|
`"sk-" in:file`,
|
|
`"sk-ant-" in:file`,
|
|
`"sk-proj-" in:file`,
|
|
}
|
|
if !reflect.DeepEqual(got, want) {
|
|
t.Fatalf("github queries mismatch:\n got=%v\nwant=%v", got, want)
|
|
}
|
|
}
|
|
|
|
func TestBuildQueries_Gist(t *testing.T) {
|
|
got := BuildQueries(testRegistry(), "gist")
|
|
if len(got) != 3 {
|
|
t.Fatalf("expected 3 queries, got %d: %v", len(got), got)
|
|
}
|
|
if got[0] != `"sk-" in:file` {
|
|
t.Fatalf("gist should use in:file syntax, got %q", got[0])
|
|
}
|
|
}
|
|
|
|
func TestBuildQueries_GitLab(t *testing.T) {
|
|
got := BuildQueries(testRegistry(), "gitlab")
|
|
want := []string{"sk-", "sk-ant-", "sk-proj-"}
|
|
if !reflect.DeepEqual(got, want) {
|
|
t.Fatalf("gitlab queries mismatch:\n got=%v\nwant=%v", got, want)
|
|
}
|
|
}
|
|
|
|
func TestBuildQueries_HuggingFace(t *testing.T) {
|
|
got := BuildQueries(testRegistry(), "huggingface")
|
|
want := []string{"sk-", "sk-ant-", "sk-proj-"}
|
|
if !reflect.DeepEqual(got, want) {
|
|
t.Fatalf("huggingface queries mismatch:\n got=%v\nwant=%v", got, want)
|
|
}
|
|
}
|
|
|
|
func TestBuildQueries_UnknownSourceDefault(t *testing.T) {
|
|
got := BuildQueries(testRegistry(), "mystery-source")
|
|
want := []string{"sk-", "sk-ant-", "sk-proj-"}
|
|
if !reflect.DeepEqual(got, want) {
|
|
t.Fatalf("default queries mismatch:\n got=%v\nwant=%v", got, want)
|
|
}
|
|
}
|
|
|
|
func TestBuildQueries_NilRegistry(t *testing.T) {
|
|
if got := BuildQueries(nil, "github"); got != nil {
|
|
t.Fatalf("expected nil for nil registry, got %v", got)
|
|
}
|
|
}
|
|
|
|
func TestBuildQueries_DedupAcrossProviders(t *testing.T) {
|
|
// "sk-" is shared by openai and anthropic — should appear exactly once.
|
|
got := BuildQueries(testRegistry(), "gitlab")
|
|
count := 0
|
|
for _, q := range got {
|
|
if q == "sk-" {
|
|
count++
|
|
}
|
|
}
|
|
if count != 1 {
|
|
t.Fatalf("expected dedup of 'sk-' to 1 occurrence, got %d", count)
|
|
}
|
|
}
|
|
|
|
func TestBuildQueries_EmptyKeywordsSkipped(t *testing.T) {
|
|
reg := providers.NewRegistryFromProviders([]providers.Provider{
|
|
{Name: "only-empty", Keywords: []string{"", ""}},
|
|
{Name: "real", Keywords: []string{"xyz"}},
|
|
})
|
|
got := BuildQueries(reg, "gitlab")
|
|
want := []string{"xyz"}
|
|
if !reflect.DeepEqual(got, want) {
|
|
t.Fatalf("empty keywords should be skipped:\n got=%v\nwant=%v", got, want)
|
|
}
|
|
}
|
|
|
|
func TestRegisterAll_NilEngineNoPanic(t *testing.T) {
|
|
defer func() {
|
|
if r := recover(); r != nil {
|
|
t.Fatalf("RegisterAll(nil, ...) panicked: %v", r)
|
|
}
|
|
}()
|
|
RegisterAll(nil, SourcesConfig{})
|
|
}
|
|
|
|
func TestRegisterAll_EmptyCfgNoPanic(t *testing.T) {
|
|
defer func() {
|
|
if r := recover(); r != nil {
|
|
t.Fatalf("RegisterAll panicked with empty cfg: %v", r)
|
|
}
|
|
}()
|
|
// Use a non-nil engine so the nil-guard early-return doesn't mask issues.
|
|
// Import cycle prevention: pkg/recon is imported by register.go.
|
|
eng := newTestEngine()
|
|
RegisterAll(eng, SourcesConfig{})
|
|
}
|